AudioDateDownLeftRightUpIconClosefacebookReportGallerySettingsGiftLanguageGridListMapMenunoAudionoPhotoPhotoPlayPlusSearchStartwitterUserVideo

American Redstart

Setophaga ruticilla

Order:
Passeriformes
Family:
Parulidae
Sections

Demography and Populations

Welcome to the Birds of North America Online!

You are currently viewing one of the free species accounts available in our complimentary tour of BNA. In this courtesy review, you can access all the life history articles and the multimedia galleries associated with this species.

For complete access to all species accounts, a subscription is required. Subscriptions are available for as little as $5 for 30 days of complete access! If you would like to subscribe to BNA, please visit the Cornell Lab of Ornithology E-Store or call us at 877-873-2626 (M-F, 8:00-4:00 ET).

Explore American Redstart distribution in BirdVis

Use our distribution analysis tool to view predicted relative abundance in specific areas of the American Redstart range.

Enlarge
Figure 4. Summer distribution map of American Redstart; data from the Breeding Bird Survey.

Relative abundance of American Redstart in U.S. and southern Canada, 2007–2013, based solely on data from Breeding Bird Survey. Numbers shown are average number of individuals detected per route per year. See Sauer et al. (2014) for details.

Enlarge
Figure 5. Regional trends in American Redstart populations, 1966-2013; data from the Breeding Bird Survey.

Data show best estimates of population change for the species over its range. Trend shown for the period 1966-2013 (Sauer et al. 2014).

Enlarge
Figure 6. Migratory connectivity of American Redstart populations between summer and winter.

The distribution of the most likely breeding region (NW, Northwest; MW, Midwest; NE, Northeast; CE, Central-east; SE, Southeast) for individuals at each wintering region (M, Mexico; C, Central America; W, Western Greater Antilles; E, Eastern Greater Antilles; L, Lesser Antilles/South America). Black dots indicate sampling locations and bars indicate the proportion of individuals assigned to each breeding region (rounded to the nearest 5%). Figure from Martin et al. 2007, adapted from Norris et al. 2006a.

Measures of Breeding Activity

Age At First Breeding

Males are sexually mature, on the basis of enlarged gonads at 1 yr of age (TWS), despite their female-like plumage. At Hubbard Brook Experimental Forest, New Hampshire, > 50% of such yearling males fail to reproduce, primarily because of shortage of mates (TWS; see Behavior: Sexual Behavior). Thus, many males reproduce successfully for first time as 2-yr-olds. Females are mature by 1 yr of age, and have not been reported to forgo reproductive effort at that age.

Clutch

Mean clutch size is 3.82 eggs at Hubbard Brook (n = 359 nests; TWS), 3.56 eggs in Ontario (n = 172 nests; Peck and James 1987), and 4.37 eggs in Alberta (n = 172 nests; M.-A. Villard and S. Hannon, unpublished data). Throughout geographic range, modal clutch size is 4 eggs (range 1–5; Baker 1944a, Sturm 1945, Bent 1953b, Peck and James 1987, TWS). Small clutches (1–2 eggs) are unusual and tend to occur toward end of breeding season (TWS). Five-egg clutches are not rare, but tend to occur early in season (29 May–11 June at Hubbard Brook), which may help explain why some observers (e.g., Baker 1944a, Bent 1953b) have not reported 5-egg clutches. In Ontario study, 5-egg clutches accounted for 5.2% of total, 4-egg clutches (64.5%), 3-egg clutches (16.3%), 2-egg clutches (9.3%), and 1-egg clutches (4.7%) (n = 172 nests; Peck and James 1987). Similarly, percentages of 5-egg, 4-egg, 3-egg, 2-egg, and 1-egg clutches at Hubbard Brook were 5.6%, 72.4%, 20.3%, 1.4%, and 0.3%, respectively (n = 359 nests; TWS). In Alberta, by contrast, 5-egg clutches were relatively frequent (43%), 4-egg clutches (51%), and 3-egg clutches (6%) (n = 172 nests; M.-A. Villard and S. Hannon, unpublished data), consistent with general latitudinal increase in avian clutch sizes. Latitudinal change in clutch size reinforced by data from nests in southernmost breeding sites, Pearl River floodplain in Louisiana, where average clutch size was 3.61 eggs (based on 14 four-egg clutches and 9 three-egg clutches; TWS, unpublished data). Thus, redstarts clutch size differs by almost 1 egg on average from Alberta (4.37 average) to Louisiana (see Breeding: Nest for details on latitudinal variation in nest and nest microhabitat).

At Hubbard Brook, clutch size decreases significantly over course of season (TWS), accounting for observed clutch size decreases of individual birds that replace lost clutches or broods (Baker 1944a, Bent 1953b).

Corresponding with declining clutch size over course of season, number of young fledged also declines linearly from almost 3 fledged offspring from initial nests to about 1 fledged young 25 days later (McKellar et al. 2013a).

At Hubbard Brook, average clutch size differs significantly by age of male: 3.86 eggs for females mated to older males (n = 287 nests) versus 3.63 eggs for females mated to yearlings (n = 72 nests; effect of female age per se on clutch size is not known). The smaller average clutch associated with yearling males at Hubbard Brook results from later clutch completion dates, since age-specific clutch sizes were indistinguishable when examined by clutch completion date. Clutch size also differs by age in New Brunswick: Mean clutch sizes 2.9 eggs (n = 106 nests) and 2.1 eggs (n = 60 nests) for older male versus yearling male nests, respectively (Lemon et al. 1992).

Annual And Lifetime Reproductive Success

Lifetime reproductive success has not been measured, but annual (season-long) reproductive success has been estimated in various studies. Annual seasonal productivity was quantified at Hubbard Brook from 1982–1995 within a 34-ha area; an attempt was made to find all nests, including renesting attempts, and most individuals were color-banded and could thus be followed season-long and between years. At this site, annual reproductive success varied significantly from summer to summer, largely due to changing nest predator abundance (Sherry et al. 2015). Females fledged an average of 1.1 young per year, but this did not differ with respect to female age. In contrast, the mean seasonal productivity of males did change with age, with 0.6, 1.0, 1.4, and 1.2 young fledged per year by yearlings, 2-yr-olds, 3-yr-olds, and 4-yr-olds, respectively; TWS and RTH, unpublished data). These estimates for male reproductive success do not include any extra-pair young sired (see Other: Life-history Traits: Delayed Plumage Maturation and Sexual Selection, and Behavior: Sexual Behavior: Extra-pair Copulations).

In a bottomland hardwoods habitat in Maryland, mean annual number of fledglings was 1.37 ± 1.48 SD (range = 0–6, n = 260 individuals, 2009–2012), and the frequency distribution was distinctly bimodal with high frequency of zero young fledged (resulting primarily from the presence of unmated yearling males) and a secondary mode of 3 young fledged for males that obtained mates (Rushing et al. 2016). In this same study, the mean annual numbers of young fledged by yearling versus older males were 0.43 ± 0.08 SD and 2.05 ± 0.26 SD, respectively; and corresponding statistics for females were 1.75 ± 0.29 SD and 2.38 ± 0.24 SD. These numbers varied by year for both males and females (Rushing et al. 2016), as observed by Sherry et al. 2015. This Maryland study was able to examine the effects on reproductive success of prior experience breeding at this site (i.e., by individuals who had reproduced previously at site) versus no previous experience at site by newly immigrating individuals (determined using stable-isotope ratio of feathers produced during molt the previous summer). This study found no effect of immigration status (local versus immigrant from a distant site) on number of young fledged by either females or males, although the number of individuals reproducing for the first time in the Maryland site was small (6–14%, n = 260; Rushing et al. 2015a, Rushing et al. 2016). The authors conclude that neither natal or breeding dispersal (any time after natal) affects reproductive success (Rushing et al. 2016), suggesting that prior experience at a site does not confer a reproductive advantage.

In the Driftless Area Ecoregion of Wisconsin, Minnesota, and Iowa, annual reproductive success of redstart population modeled to include brood parasitism effects estimated at 1.56 ± 1.31 SD young fledged per female (Powell and Knutson 2006). In same study, high brood parasitism affected productivity; each 20% increase in parasitism probability decreased productivity by about 0.08 fledglings.

The number of young fledged declines with calendar date and in parallel with a seasonal decline in clutch size (see Breeding: Eggs: Clutch Size), although other factors are involved. Based on an 11-year study in Ontario, the number of young fledged declined linearly from around 3 to 1, over a 25-day period (McKellar et al. 2013a). Hypotheses to explain this seasonal decline in reproductive success were experimentally tested by delaying the effective arrival date of older males (McKellar et al. 2013a). Experimental treatment involved holding males off territory for about 24 hours, which allowed other males to take over the vacant territories, effectively delaying territory reestablishment by 12 d. As a result, experimental males fledged significantly fewer offspring than either early arriving-early breeding or early arriving-late breeding control males; and an indistinguishable number of fledglings compared to late arriving-late breeding control males. Numbers of offspring fledged as consequence of all these treatments corresponded with the expected linear seasonal decrease in young fledged, suggesting the importance of unknown extrinsic environmental conditions at the time of egg-laying that influence female body condition. However, some evidence also supported importance of individual female intrinsic traits such as age, genetics, and/or quality of territory held (see also Rushing et al. 2016 for similar conclusions about effects of overwintering territory quality on yearling redstart reproductive success).

Nest Survival Rate

Daily nest survival (DNS) rate at Hubbard Brook, New Hampshire was 0.966 (0.960–0.972 95% CI; n = 466 nests over 13 years), which corresponds to a 0.50 probability of fledging during a 20-d incubation/nestling period, but varied significantly among years (Sherry et al. 2015). This annual variation in DNS corresponded with a nest predator (scansorial mammal abundance) index. Furthermore, experimental protection of active nests from scansorial mammals across multiple years increased nesting success from an average of 50% to 77%. In the same study, 34% of annual variation in population change was accounted for by nesting success (itself almost exclusively determined by nest predation), but weather also influenced nesting success (increased by warmer temperatures in May and greater rainfall in June), probably indirectly by influencing predator and/or nesting bird behaviors (Sherry et al. 2015).

Redstart nests, both in fragmented and contiguous forest in Ontario, had DNS = 0.943 (i.e., a 23% success probability over entire nesting cycle; Falk et al. 2011). In agriculturally fragmented landscapes of the Driftless Area Ecoregion of Wisconsin, Minnesota, and Iowa,redstart nests had DNS = 0.952 ± 0.008 SE (29% ± 6 SE over the full nesting cycle; Knutson et al. 2004). DNS rate increased during the nestling stage in Alberta (Hannon et al. 2009), but declined with nest age (incubation & nestling) at Hubbard Brook (Sherry et al. 2015). Nesting success was density-dependent in an Ontario study (McKellar et al. 2014), but not at Hubbard Brook, possibly due to different nest predator communities (Sherry et al. 2015). The risk of nest predation also increased when nests were parasitized by Brown-headed Cowbird, likely because either more frequent parental provisioning or louder nestling begging calls may increase nest detectability to predators (Hannon et al. 2009). Thus, redstart nesting success is overwhelmingly impacted by nest predators, but is modified by effects of weather, brood parasites, population density, and habitat context. Weather alone (e.g., excessively rainy, windy conditions) can also cause nesting failure; e.g., via starvation of nestlings (Sherry et al. 2015).

Age-related differences in reproduction are well studied. An early, 1-yr study at Hubbard Brook, New Hampshire found that older males compared with yearlings paired more successfully, but did not differ in clutch size, more eggs hatching/clutch, a greater percent of eggs hatching, more fledglings/successful nest, more fledglings per territorial male, or a greater percent of males successfully fledging at least 1 chick (Procter-Gray and Holmes 1981). However, these results are based on small sample size within a single summer season.

Based on a longer, 15-yr study at Hubbard Brook (TWS and RTH, unpublished data), older males generally did better reproductively than yearling males in most, but not all respects. The probability of pairing (defined operationally in this study by discovery of a female with a nest on male’s territory) was 0.75 for older males (n = 125) compared to 0.43 for yearling males (n = 96). Probability that nest fledged ≥ 1 young (given that the clutch was completed) was 0.51 for older males (n = 343 nests) and 0.54 for yearlings (n = 82), not a significant difference (based on Mayfield method, for a 20-d incubation and nestling period; TWS). Fledging success averaged 0.48 for both male groups combined from 1981 to 1989, but varied significantly among years (range 0.20–0.73, n = 279 nests; Sherry and Holmes 1992b; see also Sherry et al. 2015). Females paired with older males laid larger clutch sizes than those mated with younger males (see Breeding: Eggs). Older males also had a higher probability of obtaining polygynous mating opportunities than did younger males (see Behavior: Sexual Behavior).

In the New Brunswick study as well, older males (n = 195) performed significantly better than yearling males (n = 116) in terms of pairing success (85.1 and 48.3%, respectively), nesting success (59.5% and 40.5%) and fledging success (29.2% and 12.1%; Lozano et al. 1996). Moreover, considering only those males whose nests were found and followed, 49.1% of older males (n = 115), but only 29.8% of yearlings (n = 47), fledged young. Finally, 41.9% of females mated to older males (n = 117), but only 14.8% of those mated to yearling males (n = 48) attempted to renest following failure of the first nest (≥ 2 attempts/season). In latter study, mating success was defined by presence of female on male’s territory, nest building success by start of nest on his territory, and fledging by production of ≥ 1 fledglings. In an Ontario study, age was the best predictor of pairing success (Germain et al. 2012); these authors also found that early arrival in spring was associated with increased pairing success, and for older males, multiple characteristics of song behavior contributed to pairing success (see also Sounds and Vocal Behavior: Vocalizations: Social Context and Presumed Function of Vocalizations).

Habitat Edge Effects On Nest Success

Findings indicate that edge effects (defined as increased nest predation rates at forest edges) may be context-dependent. In Alberta, where most predators are forest dwellers, no edge effect on nest predation was apparent (Hannon et al. 2009). The same finding occurred for redstarts nesting in fragmented forests of Ontario, but not for those in contiguous forests (Falk et al. 2011). Contrasting results in these various studies may result not so much from edge per se, but rather from the types and spatial relationships of landscape elements separating fragments; e.g., human habitations that increase nest predation more than commercial forestry.

Number Of Broods Reared Per Season

Redstarts are single-brooded where the species is best studied, which has been almost exclusively in northern parts of range. Season-long studies with marked individuals in southern parts of breeding range (e.g., Louisiana) have not been conducted. Redstarts do attempt to re-nest following loss of earlier nests. In Driftless Area Ecoregion of Wisconsin, Minnesota, and Iowa, redstarts attempted an average of 2.53 ± 1.41 nests per season, of which 0.75 ± 0.44 were successful (Powell and Knutson 2006).

Life Span and Survivorship

Maximum longevity record in American Redstart is 11.5 yr , in a female banded as a first winter female in Jamaica in 1996 and resighted there in 2007; https://nationalzoo.si.edu/scbi/migratorybirds/research/data/). Maximum longevity based on banding office records is a summer-recorded longevity of 10 yr, in a male; Klimkiewicz et al. 1983). A male banded at Hubbard Brook, in 1985 in Definitive Plumage (≥ 2 yr-old) was last resighted in summer 1992, when ≥ 9 yr-old (TWS and RTH). A male banded in Jamaica in 1993 in his second year was last sighted at same location in 2001, thus 10 yr longevity (https://nationalzoo.si.edu/scbi/migratorybirds/research/data).

Average annual survival of individuals banded at Hubbard Brook at > 1 yr of age estimated to be 47.7%, on basis of assumption of constant survival with age and on regression of log of number of birds trapped versus number of years birds were resighted (n = 46, 74 males banded as yearling and older individuals, respectively); Sherry and Holmes 1991 and unpublished data). On basis of average probability of birds returning in subsequent year to site of initial marking (site fidelity), survival must be at least 51% (n = 111; Holmes and Sherry 1992); fidelity to winter sites (Jamaica) was greater than to breeding sites (Hubbard Brook). Site fidelity was also greater in males than in females. Jolly-Seber estimate of annual survival, based on Hubbard Brook site data (n = 232 birds banded), was 0.67, but this estimate had a high error (Nichols et al. 1981). Annual average survival thus appears to lie in range of 50–70%, once birds reach sexual maturity as yearlings, although all previous estimates were minimal estimates (“apparent survival”), because some birds disappeared due to emigration, not just mortality. A multistate capture-recapture analysis combining annual survival, reproduction, and movement estimates from redstarts captured/resighted at Hubbard Brook (TWS, J.-D. Lebreton, and RTH, unpublished data), yielded higher annual apparent survival estimates for individuals that reproduced most successfully (0.57 and 0.66 for males and females, respectively), estimates likely associated with greater breeding site fidelity in subsequent years. These higher estimates are probably the least biased for the species because of the addition of information about reproductive success into the models, the lower estimates being impacted more by unknown probability of individuals dispersing permanently.

Another multi-state capture-recapture study of annual survival, in Ontario (McKellar et al. 2015), estimated lower rates of adult survival overall than studies cited above, attributed to high levels of breeding dispersal (movements by breeding age individuals from one breeding season to next). Apparent survival probability was greater in older males (0.39 [95% CI 0.26–0.53]) than yearling males (0.15 [95% CI 0.09–0.23]), the difference attributed largely to greater breeding dispersal by the latter. Survival of older females (0.39 [95% CI 0.26–0.53]) did not differ from that of yearling females (0.35 [95% CI 0.20–0.55]). This study found a curvilinear relationship of annual adult survival, as measured on Ontario breeding grounds, with weather on the overwintering grounds: higher annual survival during Caribbean overwintering periods with an intermediate Southern Oscillation Index (SOI) compared to either El Niño or La Niña winters. This study also estimated annual dispersal rate between two breeding study areas with slightly different habitat to account for the relatively low apparent survival estimates: dispersal rate ranged from 0.03 (95% CI 0.01–0.14) in older females to 0.18 (95% CI 0.09–0.23) in yearling males; with breeding dispersal distances of up to 1.3 km.

In some studies females appear to survive at lower rate than males, on basis of differences in site fidelity (Holmes and Sherry 1992), and on preliminary Jolly-Seber models, taking age and sex into account (T. E. Martin, RTH, and TWS, unpublished data). However, in boreal forest in southern Canada, there was no difference in survival by sex, with the best models yielding survival for yearling male, older male, and females = 0.55 + 0.09 (Bayne and Hobson 2002a; see also previous paragraphs). No estimates of survival available for period from fledging until sexual maturity. Postfledging survival, however, is probably low, on basis of very low returns of banded fledglings in subsequent summers (0.6% = 1 of 161 banded, 7-d-old nestlings at Hubbard Brook; Sherry and Holmes 1991). In New Brunswick, 5% (n = 172) of banded nestlings returned to the study site the subsequent year, an equal number of males and females (Lemon et al. 1996).

Several estimates of annual survival come from modeling capture-recapture data from the winter grounds. In the Guanica dry forest (Puerto Rico), redstarts survived better in years with more rainfall; but numbers of individuals varied greatly among years, and capture rates were so low many years as to preclude reliable estimates of annual apparent survival (Dugger et al. 2004).

In Jamaica, birds in black mangrove habitat experienced greater annual survival and longevity, and males persisted better than those in adjacent coastal thorn scrub (Marra and Holmes 2001), although this habitat difference did not occur during the 1995–1996 overwintering period in which food and body condition were unusually high in both habitats (Angelier et al. 2009). Birds occupying thorn scrub tended to lose body mass over the winter, and although thorn scrub females had higher return rates, it appears their survival was low relative to that of males, suggesting overwintering habitat conditions may limit breeding female numbers (Marra and Holmes 2001). In another Jamaican study (Johnson et al. 2006b), similar differences in annual adult survival were found among habitats, predicted by better retention of body mass during the overwintering period in some habitats (black mangroves, shaded coffee plantations).

All these winter survival studies taken together suggest that winter ecological conditions may impact population dynamics primarily via influencing annual adult survival in response to winter food abundance (see also Wilson et al. 2011b). See Causes of Mortality: Environmental Conditions in Winter, and Population Regulation: Population Limiting Factors.

Disease and Body Parasites

Diseases

Enlarge
Figure 6. Migratory connectivity of American Redstart populations between summer and winter.

The distribution of the most likely breeding region (NW, Northwest; MW, Midwest; NE, Northeast; CE, Central-east; SE, Southeast) for individuals at each wintering region (M, Mexico; C, Central America; W, Western Greater Antilles; E, Eastern Greater Antilles; L, Lesser Antilles/South America). Black dots indicate sampling locations and bars indicate the proportion of individuals assigned to each breeding region (rounded to the nearest 5%). Figure from Martin et al. 2007, adapted from Norris et al. 2006a.

A spirochaete, possibly the Lyme Disease pathogen (Borrelia burgdorferi sensu lato), identified in a skin biopsy of 1 redstart mist-netted in Fall 1997 on St. Catherine’s Island, coastal Georgia (Durden et al. 2001). A cluster of records of this pathogen in several species of birds at the same time and place suggested the possibility of “periodic amplification or recrudescence of spirochetes in reservoir avian hosts” (Durden et al. 2001: 231).

Redstart populations were tested for impact of West Nile Virus on annual adult survival using a sample of 3,285 individuals captured at 41 bird banding stations across North America, but no significant effect was found (George et al. 2015).

Body Parasites

Four ectoparasites identified in study by H. S. Peters (cited in Bent 1953b): 3 species of lice (Menacanthus sp., Myrsidea incerta, and Philopterus subflavescens) and a tick (Haemaphysalis leporis-palustris). A new species of Cheyletoid mite (Harpyrhynchoides vulgaris: Harpirhynchidae), parasitic on skin of a variety of birds, described in part based on six female redstarts collected in Manitoba (Bochkov and Galloway 2004).

Direct extraction and polymerase chain reaction (PCR) amplification of microbial (bacterial) DNA from feathers of American Redstarts and sympatric soil samples from both breeding (Maryland) and overwintering (Jamaica) sites (Bisson et al. 2007) revealed:

(1) a moderately diverse feather fauna overall (30 operational taxonomic units) dominated by Pseudomonas spp.,

(2) biogeographic and habitat-specific differences of feather bacterial communities (e.g., restarts occupying wet mangrove and dry scrub habitat in Jamaica had feather bacterial communities as different as between Jamaica and Maryland),

(3) considerable similarity of redstart feather bacterial communities in Maryland vs. Jamaica, probably due to the movements of the birds, compared to greater soil bacterial community differences, and

(4) one chloroplast (algal) molecular sequence at the Maryland site. Possible impacts of these bacterial communities on the birds were not discussed.

Two studies report endoparasites from American Redstarts. First, based on birds sampled during fall (September–October) in Jamaica, two hatch-year birds (5% of individuals sampled) were positive for Haemoproteus fringillae, and one after-hatch-year bird (3% of individuals sampled) was positive for Trypanosoma avium (Garvin et al. 2004b). None of these individuals was infected the following spring in Jamaica, suggesting the lack of local (Jamaican) transmission, but rather infection in North America, prior to fall migration, and probably by way of biting Diptera (Hippoboscidae or Ceratopogonidae).

The second endoparasite study found a small, but significant regional difference in redstart hematozoan parasites across the breeding range (Durrant et al. 2008). The southeastern region (two sites sampled in southern Louisiana) had the most distinctive blood parasite assemblages, including unique haplotypes of Haemoproteus spp., but also Plasmodium spp.; and some east-west (across North America) parasite assemblage structuring also detected. The southeastern birds also had the highest haematozoan prevalence (54% of birds sampled) compared to other breeding (22–32% prevalence) and overwintering (2–4% prevalence) regions. This pattern of haematozoan community structuring parallels patterns of migratory connectivity insofar as southeastern breeding redstarts tend to winter in Trinidad and Tobago, a distinctive part of the overwintering range (Norris et al. 2006b; Figure 6), but breeding range structuring probably arises also from local assemblages of dipteran vectors (mosquitos).

Causes of Mortality

Causes not comprehensively quantified to date, but include exposure, predation, and accidental collision deaths at TV towers and buildings (see Conservation and Management: Effects of Human Activity). Exhaustion during migration, probably exacerbated by storms, must be important cause of death, on basis of reports of exhausted flocks seeking resting spots offshore (Griscom and Sprunt 1957, Stevenson and Anderson 1994b).

Exposure

Proportion of nestlings dying from exposure or starvation during cold, rainy periods varies greatly, from 0% most years to 45% of nests in rare years, at Hubbard Brook Experimental Forest, New Hampshire (Sherry and Holmes 1992b).

Predation

Adult females are occasionally killed on or near the nest by predators (see Behavior: Predation). Nests at Hubbard Brook experienced an average 0.5 probability of fledging young over combined 20-d period during which either eggs or nestlings are typically exposed to predators (see Measures of Breeding Activity: Nest Survival Rate), and most nest losses were attributable to predators (Sherry and Holmes 1992b), especially red squirrels (see Behavior: Predation; Sherry et al. 2015). Nest success rate decreased with time after the start of incubation (Sherry et al. 2015), possibly due to increased parental activity in response to nestling growth that could attract attention of diurnal predators like squirrels. Predation risk to males in summer and to all individuals during other times of year has not been quantified. For details on predator identities, for adults as well as for eggs and nestlings, see Behavior: Predation.

Environmental Conditions In Winter

Redstart annual survival appears most strongly linked to winter ecological conditions. For example, Johnson et al. (Johnson et al. 2006b) show that 93% of variation in annual adult survival, winter-to-winter in Jamaica, is explained by variation in body mass maintenance among those winter habitats, suggesting the importance of body condition during winter.

Similarly, Angelier et al. 2013 implicated overwintering habitat as important to annual survival using length of telomeres (repetitive, non-coding DNA sequences at ends of chromosomes), an index to aging and longevity: Specifically, redstart telomere length, although not different on average between habitats, decreased more for banded individual birds from one winter to next in thorn scrub habitat in Jamaica compared to adjacent black mangroves, suggesting greater stress for birds in the former. In this same study telomeres shortened significantly in individuals from one winter to the next, and individuals with longer telomeres tended to survive better annually, habitats pooled.

Wilson et al. 2011b also implicated overwintering conditions as a driver of annual adult survival by linking changes in breeding season abundance in eastern North America (based on Breeding Bird Survey) to satellite-image-based estimates of rainfall (especially in Cuba). Overwintering conditions thus probably affect survival indirectly, by affecting body condition at the time of northward migration in spring, and thus vulnerability to predators, storms, fatigue, stress, and other direct causes of death, which may occur after northward departure on migration. Sillett and Holmes 2002 present calculations indicating disproportionate annual mortality taking place during migration, not winter, in the phylogenetically related Black-throated Blue Warbler (Setophaga caerulescens), but this mortality during migration could be a result of poor conditions during the preceding winter. A similar cause-and-effect relationship may be occurring in redstarts.

Marra and Holberton 1998 compared the physiological stress-response involving corticosteroid hormones in redstarts occupying male-dominated black mangrove habitat in Jamaica compared to those in female-dominated thorn scrub. No difference was found in baseline corticosterone concentration at time of capture in fall, but in spring individuals of both sexes in thorn scrub had elevated baseline concentrations of corticosterone, and were unable to mount an acute corticosterone secretion response to stress. In spring, individuals with higher baseline corticosterone concentration had lower body mass, suggesting that stresses of inhabiting thorn scrub habitat contributed to catabolic body mass loss, disadvantaging such individuals by delaying their spring migration and possibly increasing risk of mortality.

Angelier et al. 2009 expanded study of these stress–related mechanisms in annual survival of redstarts in the same black mangrove vs. thorn scrub habitats, and found condition-dependence: Individuals in the potentially more stressful scrub habitat were more likely to survive to the following winter (and return to Jamaica) the stronger their adrenocortical stress response. This relationship was not found in the black mangrove habitat, potentially because the latter individuals are less impacted during the winter (and other times of year) by the same stressors as thorn scrub-inhabiting individuals. Adrenocortical stress response was consistent among individuals, suggesting it is phenotypically constant; and baseline corticosterone levels do not predict annual survival under any circumstances. Angelier et al. (Angelier et al. 2010) discuss methods of obtaining reliable baseline corticosterone estimates necessary to estimate the adrenocortical stress response.

Competition With Other Species

Experiences interspecific aggression in summer from socially dominant Least Flycatcher (Empidonax minimus); yearling males (and their mates) attacked proportionately more than older birds (Sherry and Holmes 1988, Fletcher 2007, Fletcher 2008: see Behavior: Social and Interspecific Behavior: Non-predatory Interspecific Interactions).

During the overwintering period, competes with socially dominant Adelaide’s Warbler (Setophaga adelaidae) in Puerto Rico (Toms 2011, Toms 2013). Food limitation, a prerequisite for competition, indicated in drier of two years in this Puerto Rican study, in which both species lost body mass during the non-breeding dry season. Similar foraging behaviors (foraging locations, types of attack maneuver) indicate likely high overlap in diet; no resource partitioning documented between the two species. Adelaide’s Warbler did not shift foraging behavior after American Redstart departed in spring migration, indicating no measurable competitive effect of latter on former. No interspecific territoriality found, but both species responded strongly to simulated interspecific territorial intrusions, the larger bodied Adelaide’s the more aggressively. American Redstart thus appears to coexist via foraging flexibility, using “vagrant fugitive strategy” in which they forage relatively silently compared to Adelaide’s and in parts of Adelaide’s territories temporarily vacated by latter. Similar interspecific competition occurs with socially dominant resident Yellow Warbler (Setophaga petechia) in Jamaica (P. P. Marra and L. Powell, unpublished data); these interspecific dominance relationships could cause stress in American Redstart and thus worsen body condition, but are probably not a direct cause of death.

Redstarts probably compete interspecifically and diffusely with a variety of other insectivores, as indicated by three observations: First, high interspecific food overlap with four other wintering migrant bird species in Jamaican shaded coffee plantations, where all five species feed opportunistically on the same, small, patchily distributed insect food resources available, coupled with abundant evidence for strong intraspecific competition for food in winter (see Demography: Population Limitation) suggests diffuse exploitation competition among coexisting migrant species (Sherry et al. 2016). Second, the possibility of diffuse interspecific competition involving many migrant species competing for food in winter with many resident species is implicit in “Breeding Currency Hypothesis”, tested with seasonal food sampling across variety of Jamaican habitats with different proportions of overwintering migrant bird species (Bennett 1980, Johnson et al. 2005c). Third, large (continental scale) spatial distribution patterns of redstart in breeding season compared with the other 42 common North American Parulidae suggest interspecific competition, based on disproportionate absences of local (point survey stops along Breeding Bird Survey routes) coexistence between phylogenetically most closely related, and thus likely most intensively competing species (Lovette and Hochachka 2006). This latter study also finds evidence for phylogenetic niche conservatism, based on unusually high coexistence at survey points of sympatric (found on same BBS route) species, possibly because of inherited habitat preferences.

American Redstart may also coexist with other species via different genetically based morphological and behavioral traits that allow more aerial acrobatic and flycatching foraging compared to phylogenetically closest relatives (Sherry et al. 2016). In summer, redstarts also forage differently from phylogenetic relatives: For example, despite overall similarities in leg and wing morphology, different species of Setophaga attack prey differently compared to redstarts, and to other species when confronted with different foliage structure in experimental situations, suggesting intrinsic morphological differences that affect foraging behavior (Whelan 2001).

American Redstart segregates by habitat in winter from some other Neotropical migrant species overwintering in Jamaica, e.g., Common Yellowthroat (Geothlypis trichas), which is more tolerant of vegetation reduction associated with human disturbance, and Palm Warbler (Setophaga palmarum), which occurs almost exclusively in habitats with reduced vegetation (Confer and Holmes 1995), suggestive of interspecific competition.

Range

Natal Philopatry

Not known, but low rate of recovery of banded nestlings suggests low degree of fidelity to natal site: 1 of 161 (0.6%) of nestlings banded at Hubbard Brook, New Hampshire were sighted in a subsequent year (Sherry and Holmes 1991); 5% in New Brunswick study (Lemon et al. 1996); 26 of 632 (4.1%) of nestlings and fledglings banded in Ontario (McKellar et al. 2014).

Initial Dispersal From Natal Site; Fidelity To Breeding Site

Site fidelity of banded adult birds in summer is strong (to within 250 m of where bird nested in previous year; Holmes and Sherry 1992, Lemon et al. 1994, Lemon et al. 1996). Of older males at Hubbard Brook site, 16% returned to site 1 yr after banding (n = 51); of yearling males, 39% (n = 83); of females, 19% (n = 48). Yearling males that obtained mate were significantly more likely to return to site where banded than were individuals that never attracted mate, consistent with greater dispersal after first spring in latter individuals (TWS). Unmated yearling males sometimes move 1–2 km in spring, singing and displaying a territory at each location, suggesting that extended exploration consistent with natal (prior to first nesting) dispersal continues well into first summer (TWS, M.-A. Villard, personal communication). However, older males may also disperse up to 6 km (M.-A. Villard and S. Hannon, personal communication), indicating that dispersal occurs at multiple ages.

In Maryland, based on rectrix feather stable-isotope ratios, long-distance dispersal is relatively rare; i.e., most individuals (85.6%, n = 320) regardless of age or sex return to same site, statistically determined, as either where they were fledged or spent time during prior breeding season (Rushing et al. 2015a). When immigration did occur, involving both first-time breeders and older individuals, proportionately more individuals came from northern than southern sites, consistent with most of the breeding range situated to the north of Maryland (Norris et al. 2006b).

Variation in redstart dispersal behavior is linked to ecological conditions both in winter and in breeding areas. Based on feather stable-isotope ratios, first-winter individuals occupying preferred Jamaican black mangrove habitat and departing relatively early on spring migration tend to breed to south of their natal region in North America, whereas individuals tending to depart later from adjacent thorn scrub habitat breed relatively north of their natal region (Studds et al. 2008). This result was corroborated in a Maryland breeding study (Rushing et al. 2015a) that showed that drought during the overwintering period could have the same effect on first-year immigrant breeders (but not older individuals) as relatively poor winter habitats within a winter season: These first-year immigrants to Maryland site tended to shift northward compared to their natal breeding areas, i.e., arrive disproportionately from southern natal sites, but only in a year corresponding with drought conditions (and late spring departures) in their Caribbean overwintering areas. Furthermore, individuals of different ages tended to depart Caribbean overwintering areas relatively late in poor years there, and continue farther northward in spring to find suitable sites with plant phenology better synchronized to these birds’ reproductive conditions (and possibly more caterpillar food resources); i.e., timing of phenological (leaf-out) conditions in Maryland appear to have influenced decisions by dispersing individuals as to where to settle in spring (Rushing et al. 2015a). These findings suggest that dispersal between breeding sites, even in older individuals that have previously bred at least once, occurs regularly, and thus is more flexible in response to various ecological conditions (both in winter and spring) than once believed. Although such long-distance dispersal events are relatively rare (14.4% of individuals studied by Rushing et al. 2015a), the implications of flexible dispersal behavior are two-fold: (1) redstarts appear able to respond to earlier spring plant phenology in breeding areas by dispersing northwards, which would help adjustment to a warming climate; and (2) rare dispersal events increase movements that potentially enhances gene flow geographically and thus affect population dynamics.

In contrast to yearling males, the breeding locations of older male redstarts do not change in response to different overwintering habitats occupied in Jamaica (based on stable-isotope ratios), probably due to their relatively greater commitment to breeding site (Studds et al. 2008).

Fidelity to Stopover Sites

Not known.

Overwinter Dispersal; Fidelity To Overwintering Site

Site fidelity of banded birds has been documented repeatedly in overwintering areas (McNeil 1982, Faaborg and Arendt 1984, Holmes et al. 1989, Sherry and Holmes 1996a, Wunderle and Latta 2000). Site fidelity in Jamaica appears to be stronger than that in summer in New Hampshire (Holmes and Sherry 1992). In Puerto Rico, redstarts have 65% overwinter site persistence rate in shade-coffee habitat, similar to natural forest, although this rate decreases with plantation size (Wunderle and Latta 2000). About one third of marked birds return the following year to these coffee plantations. In Jamaica (Holmes and Sherry 1992), redstarts generally return in subsequent years to overwinter within 75 m of sites occupied previously; percent of yearling males, older males, and females that return to Jamaican sites the year after being banded are 75% (n = 12), 49% (n = 57), and 46% (n = 42), respectively. However, in Venezuela individuals are less site-faithful (Lefebvre et al. 1992); and facultative movements of overwintering individuals in Panama is suggested by an inverse correlation of redstart abundance in nearby Pacific versus Caribbean coastal black mangrove sites, and by a late winter increase in the drier Pacific coast mangroves (possibly in response to increased abundance of small insect eggs in the late winter Pacific site) and other evidence (Lefebvre and Poulin 1996).

Population Connectivity

Population connectivity is the geographic structuring of migratory populations between breeding and non-breeding ranges: It is strong if a distinct breeding population segment also overwinters in a distinct portion of the overwintering range, and vice versa (Webster and Marra 2005). American Redstart is one of first migratory songbird species documented to show moderate to strong population connectivity (Boulet and Norris 2006, Norris et al. 2006b, Norris et al. 2006a; Figure 6). Although individual redstarts overwintering at a particular site come from several breeding regions (i.e., individuals mix to some extent in winter compared to summer, and vice versa), patterns of geographic structuring are nonetheless apparent, with the greatest degree of structuring from west to east (longitudinally), based on both stable-hydrogen isotope markers in feathers and seasonal tracks of banded individuals (Norris et al. 2006b). For example, individuals that overwinter in western and eastern Mexico tend to breed in the northwestern-most and Midwestern breeding sites, respectively; and individuals that overwinter in the eastern parts of the range tend to breed in the eastern part of the breeding range (Figure 6). Furthermore, populations in the eastern part of the overwintering range (Florida, Bahamas, and the Caribbean region) also follow chain migration (Norris et al. 2006b), with the northernmost overwintering populations tending to breed the furthest north and the southernmost populations further south (Figure 6). For example, Louisiana (southernmost) breeding individuals apparently tend to migrate farthest south (and east) to the Lesser Antilles, Trinidad, and Tobago. Hobson et al. 2014b confirm that eastern-breeding individuals overwintering in Venezuela correspond to more southerly breeding populations compared to individuals that overwinter in Cuba or Puerto Rico. This latter study also shows that individuals overwintering in Puerto Rico breed at consistent latitudes, based on stable-isotope markers in feathers collected over a five-year overwintering period.

Population connectivity is important to understand year-round population limitation and regulation, carry-over effects between seasons and stages of the annual cycle, conservation, life-histories including distance migrated (Norris et al. 2006a, Marra et al. 2006, Marra et al. 2010, Webster and Marra 2005), and possible genetic variation in nesting and breeding traits (see Breeding: Nest Site). Wilson et al. 2011b used redstart connectivity patterns to show that annual changes in abundance of eastern breeding redstart populations (based on Breeding Bird Survey data) are predicted by annual variability in rainfall in the Caribbean region, particularly Cuba, as indicated by the Normalized Difference Vegetation Index (NDVI) of vegetation greenness derived from satellite imagery. Western populations in their study did not show this linkage between winter ecological conditions and population change. Similarly, in another study taking advantage of redstart connectivity, an e. North American population showed somewhat stronger relationship between earlier arrival to the breeding areas corresponding with earlier departure from the overwintering grounds in wetter winters, compared with a western North American population around 3,000 km away (McKellar et al. 2013b). These authors argue that projected global climate change scenarios of increased drought in winter could contribute to long-term decline in this species, particularly the eastern-breeding populations, given the demonstrated detrimental carry-over effects between overwintering and breeding periods, such as delayed migration, in drought years.

Population Status

Numbers

Using data from the North American Breeding Bird Survey (BBS), the American Redstart population was estimated at 42,000,000 individuals for the United States and Canada, 2005–2014 (Rosenberg et al. 2016). Point-count abundance data collected by breeding bird atlas projects estimated populations at 3,000,000 individuals in Ontario, 2001–2005 (Cadman et al. 2007a), 730,000 singing males (95% CI: 700,000–760,000) in Pennsylvania, 2004–2009 (Wilson et al. 2012), and 97,000 singing males (95% CI: 85,000–110,000) in Ohio, 2007–2011 (Rodewald et al. 2016).

Generally speaking, highest abundance during the breeding season is from central British Columbia, southern Manitoba, southern Ontario, southern Quebec, Nova Scotia, and northern portions of Minnesota, Wisconsin, and Michigan, east within the U.S. through northern New York, northern Vermont, New Hampshire, and northern Maine (Figure 4). A variety of studies have published local density estimates for both summer and winter populations. Summer densities can be as great as 44 individuals/10 ha (Hamel et al. 1982; all numbers here have been converted to numbers of males and females per 10-ha basis for ease of comparison). Sabo 1980 reported 1.8 individuals/10 ha in virgin spruce forest in White Mountains, New Hampshire, and 8 individuals/10 ha in lower-elevation subalpine zone. At Hubbard Brook, numbers on one 10-ha study plot have ranged from a maximum of 44 individuals in the mid 1970s to zero since 2013 (Holmes et al. 1986, Holmes and Sherry 1988, Holmes and Sherry 2001, RTH). Other breeding densities: 15–21.9 individuals/10 ha in northern New York (Ficken 1962a); 7.1 in Maryland (Whitcomb et al. 1981); 13, 21, and 29 in Tennessee, Maryland, and Washington State, respectively (Bennett 1980); 5–15 in Quebec (cited by Cyr and Darveau 1996); 19.9 (95% CI: 15.1–26.1) pre-Hurricane Katrina to 27.5 (95% CI: 21.1–35.9) post-Katrina (difference not significant) in Bogue Chitto National Wildlife Refuge, Louisiana, in bottomland hardwoods forest in Pearl River floodplain (Brown et al. 2011a). These latter estimates are densities estimated from point counts, so represent mostly auditory (male territory) sightings, and therefore to be roughly comparable, should be doubled, to give total individuals as defined above. In this case the post-Katrina estimate (55 individuals) is the highest density recorded for breeding redstarts.

Density in summer is related to successional stage and structure of vegetation (Webb et al. 1977, Welsh and Healy 1993, Cyr and Darveau 1996, Hunt 1996), and to deciduousness of vegetation in New Hampshire (TWS), but underlying ecological determinants of breeding density (e.g., food abundance, predation) are not well understood.

Overwinter estimates of redstart densities reach can be as great as 150 individuals/10 ha. Estimates (per 10 ha) from a variety of Jamaican study sites and years: 14–44 (Holmes et al. 1989), 17–56 (Sherry and Holmes 1996a), 40–60 in mangrove and thorn scrub (Marra and Holmes 2001), 63 in black mangroves (range 49–85) and 40 in adjacent thorn scrub (range 25–61; Marra et al. 2015), and up to 150 individuals/10 ha, at least temporarily, in mangrove and scrub, densities boosted by surprisingly abundant transient individuals in habitats previously considered to contain largely territorial individuals (Peele et al. 2015). These high densities, and the nature and abundances of transient individuals need further study (see Priorities for Future Research). Some of these estimates come from same study areas at different times, and some are based on different methods. Other overwinter estimates are 3.8 and 7.5 (Grand Cayman Island), 10 (Mexico), and 17.5 (Puerto Rico; Bennett 1980); 0–6.5 in Yucatán Peninsula (Waide et al. 1980); 20–52 in Belize (S. Baird, unpublished data); and 34 (shade coffee) and 9 (primary forest) in Venezuela (Bakermans et al. 2009). Overwinter densities are correlated with arthropod abundance in foliage and possibly interspecific competitor abundance (Sherry and Holmes 1996a, Johnson and Sherry 2001).

Terborgh 1989 and others have argued that Neotropical–Nearctic migrants overwinter in geographic areas that are so much smaller than summer ones that densities must be greater in overwintering areas than in breeding areas. Although our review above indicates considerable variability (and maybe patchiness) within a season, the fact that some redstart winter densities far exceed any summer ones suggest that Terborgh’s generalization may be reasonable for the American Redstart. However, redstart density is also variable across the breeding range, making generalizations and comparisons based on geographic range simplistic and risky.

Trends

An examination of long-term data from the Breeding Bird Survey highlights the complex and variable nature of population trends over time and geographic regions. From 1966–2013, American Redstart breeding populations in the U.S. and Canada declined by -0.34% annually, and little if at all since around 1990 (Sauer et al. 2014b); this trend was not significantly different from zero based on the Interval Method, nor the Regression Method. Nonetheless, over a 45-yr period (1970–2014), BBS data indicated that the breeding population has declined by an estimated 12% (Rosenberg et al. 2016), illustrating that a statistically insignificant long-term decline can amount to a large change in a population. More recently (2003–2013), the survey-wide annual trend was 0.56%, but this was not statistically significant (Interval Method). From 1966–2013, the American Redstart population showed no long-term trend in Canada (0.01%), but declined by -1.43% per year in the U.S. (Sauer et al. 2014b).

Between 1966 and 2013 (Sauer et al. 2014b; Figure 5), the most statistically significant annual declines occurred in northern New England, including Maine (-4.24%), New Hampshire (-3.07%), New York (-2.06%), Vermont (-1.79%), and nearby the Maritime Provinces of Canada including Nova Scotia (-2.34%) and New Brunswick (-1.33%). This northeastern decline, which was most dramatic after about 1980, appears to have resulted at least in part from regional maturation of hardwood forests, which reduced suitability of habitat for this species (Holmes et al. 1986, Holmes and Sherry 1988, Hunt 1996, Holmes and Sherry 2001, Holmes 2007). Other areas of significant decrease include Alabama (-1.75%), Arkansas (-2.6%), Tennessee (-4.27%), and western states including Montana (-4.33%), Idaho (-3.22%), and the broader Northern Rockies Ecoregion (-0.95%). Some of these declines may have resulted in part from human-caused loss or degradation of riparian habitats preferred by the species in the western U.S. (Sallabanks 1993e), but recent droughts linked to climate change may also have contributed, particularly in the western U.S. James et al. 1996 also found continental-scale declines in American Redstart populations between early 1970s and late 1980s, but the decline was significant using only 1 of 2 analytical methods.

By contrast, annual population trends have increased from 1966–2013 in Wisconsin (1.66%), Ohio (1.75%), New Jersey (3.02%), and Manitoba (1.8%) (Sauer et al. 2014b; Figure 5). Although not quite significant, an increase in Quebec (0.95%) may have resulted from forest regeneration after logging (Cyr and Darveau 1996). Declines are more widespread than increases, and have been dramatic in parts of range where abundance is greatest (New England, Maritime Provinces). However, declines to date are not extensive enough to warrant management concern, because the species is locally abundant, widely distributed, tolerant of range of habitat conditions, and increasing in some locations (Butcher 1992, Sauer et al. 2014b).

Population Regulation

Population Limiting Factors

Ecological factors that limit or regulate populations in the American Redstart, as for almost all Neotropical-Nearctic migrant birds, are poorly understood (Sherry and Holmes 1995), but knowledge is expanding rapidly due to increasing recent research focus. Theoretically, populations of migratory birds are potentially limited simultaneously in summer and winter (Sherry and Holmes 1995, Sherry and Holmes 1996a, Runge and Marra 2005), as well as being affected during migration (Faaborg et al. 2010b). Summer factors limiting redstarts appear to be primarily linked to fecundity, while independent winter ecological circumstances appear to have their greatest effect on population dynamics by affecting annual adult survival (Sherry et al. 2015). Winter effects also carry-over to breeding season, and impact redstart fecundity as described below. Summer season effects may also carry-over into winter, as documented in Louisiana Waterthrush (Parkesia motacilla; Latta et al. 2016), but this has not been confirmed in redstart. In both winter and summer habitat quality, defined as the conditions favoring either fecundity or survival, vary considerably. In summer a variety of additive ecological population limiting factors are documented:

(1) Redstarts prefer nesting in forests of moderate age (see Habitat: Breeding Range), which may explain regional decline in northern New England, where forest stands appear to be maturing past preferred age (Holmes et al. 1986, Hunt 1996, Holmes and Sherry 2001);

(2) Increase in Brown-headed Cowbird populations can cause locally high rates of brood parasitism (see Breeding: Brood Parasitism) that decrease fecundity. The inference that brood parasites have contributed to population declines is strengthened by observation that recruitment of yearlings into local populations, and subsequent changes in population size, are directly related to fledging success the previous summer (Sherry and Holmes 1992b, Sherry et al. 2015). Note, however, that declines in New Hampshire and other New England localities often occur in habitats without cowbirds (Holmes and Sherry 2001).

(3) Increased populations of nest predators in the eastern U.S. resulting from increased forest edge habitat associated with fragmentation of the once-continuous forests in region are implicated as a major cause of local population declines (Sherry and Holmes 1992b). More direct evidence for the importance of nest predators comes from experimental and correlational studies of nesting success linked to subsequent population changes. In an experiment conducted at Hubbard Brook (Sherry et al. 2015), sheet metal baffles placed beneath active nests and designed to prevent access to nest by scansorial animals had a strong effect increasing nesting success and implicating sciurid mammals, particularly red squirrels, as the most important predators at least in this New Hampshire study site.

(4) Food availability in breeding season is also implicated as causing population change (Jones et al. 2003b): Redstart abundance tended to increase in breeding seasons following those with relatively high abundance of caterpillars, an important food for nestlings and fledglings, and to vary synchronously with caterpillar abundance over a wide region in the White Mountains of New Hampshire.

Evidence is increasing that the overwintering period is important in limiting redstart abundance, probably via habitat quality as determined primarily by food supply through the period, particularly into the dry season prior to northward migration (reviewed by Sherry and Holmes 1995, Sherry and Holmes 1996a, Sherry et al. 2005, Wilson et al. 2011b; Table 2). Habitat quantity could also be important insofar as the redstart overwintering range is compressed into an area 63% of breeding range, because of how land area becomes constricted longitudinally in mainland Central America and on Caribbean islands, compared to longitudinally broader North American breeding range (Mills 2006). Some overwintering habitats appear to be saturated, on basis of several lines of evidence for redstarts from Jamaica and other high-density overwintering sites (Marra et al. 1993, Marra et al. 2015, Sherry and Holmes 1996a, Cooper et al. 2015). Many migratory birds, including American Redstart, appear to be food-limited during the overwintering period (Sherry et al. 2005, Toms 2011).

Overwintering habitat quality likely limits individual fitness. Body condition of redstarts improves as the season progresses in Venezuelan shade coffee, but not in nearby primary forest (Bakermans et al. 2009). Birds that upgrade to high-quality mangrove habitat from scrub tend to maintain their overwinter body mass, depart on average 6 d earlier for spring migration, and are more likely to return the following winter (Studds and Marra 2005). Similarly, in Jamaica, redstarts survive best from one overwintering period to the next in those habitats (especially black mangroves, shaded coffee) where they can maintain overwinter body mass (~ 91% of variance in this relationship so explained) consistent with food limitation (Johnson et al. 2006b). Wet limestone forest in Jamaica is also high quality overwintering habitat based on relatively high annual adult survival, early spring migration departure times, age composition biased in favor of older males, and other lines of evidence (Peele 2015). See also Habitat: Overwintering Range.

Overwinter rainfall is probably a major determinant of habitat quality both temporally and spatially, via its effect on food availability, and consequently on overwinter body condition, migration timing, annual adult survival, and population dynamics. In Jamaica, El Niño-Southern Oscillation cycles correspond with annual adult survival of Black-throated Blue Warblers, abundant in similar habitats to the redstart, with survival poorer in drier El Niño years (Sillett et al. 2000). Most of the variation in redstart annual survival, determined among habitats in Jamaica, is attributable to over-winter maintenance of body mass (Johnson et al. 2006b), which is correlated with overwinter rainfall (Studds and Marra 2007, Studds and Marra 2011) and probably overwinter food supplies (Sherry et al. 2005). Eastern, but not western U.S. breeding redstart population fluctuations (determined using 26 years of Breeding Bird Survey data, 1982–2007) are predicted by an index of rainfall in the corresponding wintering regions for eastern populations, namely Cuban satellite-imaged Normalized Deviance Vegetation Index (NDVI) = dry season vegetation greenness (Wilson et al. 2011b). In a followup study, Wilson et al. 2013b experimentally irrigated thorn scrub habitat during dry season in Jamaica, and significantly altered plants, specifically reducing leaf fall and thus maintaining vegetation cover and thus overall greenness. This experiment did not have detectable impacts on arthropod abundance, or on redstart body condition, probably due to sampling and treatment issues (e.g., limited water for sustained irrigation). However, a marginally significant difference occurred in turnover of redstart individuals, with older males replacing females or yearling males on territories that had been irrigated, suggesting the importance of vegetation cover to winter territory quality and occupancy.

Body condition, foraging behavior, and habitat use are likely interrelated, based on study of female redstarts occupying coastal thorn scrub and black mangrove habitat in Jamaica (Powell et al. 2015). Females were studied because of their likely greater vulnerability to drought conditions during winter in the scrub compared to mangrove habitat. Females switched to relatively aerial foraging tactics between wetter early winter vs. later drought-impacted conditions in scrub, but not mangrove habitat. In both habitats, individuals in better body condition fed less frequently than lighter individuals. In spring (during drought), individuals in better body condition in scrub (but not mangroves) tended to forage with more aerial tactics and with more wing-powered movements. The proportion of aerial foraging tactics in these individuals was also greater in individuals making fewer attacks per unit time. Powell et al. 2015 argue for potential positive feedback involving birds in better body condition allowing exploitation efficiency (energetic tactics and more rewarding prey), allowing the luxury of not having to work as energetically to maintain body condition, and simultaneously allowing the use of better habitats, and vice versa. Changing foraging opportunities during the dry season due to drought deciduousness in scrub, but not mangroves, may have contributed as much as changing insect abundances and types to influence these relationships.

Carry-over Effects (Seasonal Interactions)

Carry-over effects, also known as seasonal interactions, are the individual or population residual effects in one season of ecological events or circumstances in a previous season. Carry-over effects have been modeled using redstart biology and data to parameterize a population dynamics model and test its consequences (Runge and Marra 2005). This model incorporates density-dependence in summer and winter, carry-over effects from winter to summer, and social dominance of males over females regarding access to winter habitat (all documented in redstarts) to show how habitat in either winter or summer can affect equilibrium size of the population as well as sex ratio; and sex ratio can also affect population dynamics. See Norris and Marra 2007 and Harrison et al. 2011 for reviews, including discussions of American Redstart.

American Redstarts were the subject of one of the earliest studies documenting a carry-over effect in any migratory bird, showing that winter habitat influences arrival timing—and thus by inference reproductive opportunities—on the breeding grounds (Marra et al. 1998). Using stable carbon isotope signatures of winter habitat detected in muscle tissues of birds arriving to breed, this study showed that redstarts wintering in black mangroves leave this habitat in spring earlier and in better body condition, compared to thorn scrub birds, and correspondingly arrive earlier to breed in New Hampshire. Subsequent work has shown that stable isotopes in blood can serve equally well as a winter habitat marker to study carry-over effects (Norris et al. 2004a, Norris et al. 2005).

Timing of departure from winter habitat (and corresponding arrival times to breeding habitat) links winter habitat quality of redstarts to their reproductive success the following spring (Marra et al. 1998, Norris et al. 2004a, Studds and Marra 2005, Reudink et al. 2009a, Tonra et al. 2011, Tonra et al. 2013). Tonra et al. 2011 and Tonra et al. 2013 also linked migration timing, winter body condition, and preparation for breeding in older males to blood androgen levels determined and/or manipulated both in spring prior to departing winter sites and upon arrival to breed (see Distribution, Migration, and Habitat: Control and Physiology). Norris et al. 2004a used path analysis to link arrival date (corresponding with winter habitats differing in quality) of birds to breed in Ontario to fledging success: In older males, later arrivers fledged ~1 fewer young and did so ~1 week later as a consequence of wintering in drier scrub compared to wet limestone winter habitats; in females the corresponding cost was two young fledged, and almost a month delay in fledging date. The mechanism involved delayed egg laying (males) and fledging date (both sexes), linked negatively to number of young fledged.

Winter territory quality and early winter departure for spring migration is also linked to reproductive success via increased probability of mating and nesting success in yearling males (Rushing et al. 2016) and via more polygynous matings, extra-pair copulations, and reduced likelihood of cuckoldry (Reudink et al. 2009a). In a bottomland hardwoods habitat in Maryland, individuals with relatively poor quality winter habitats (based on stable carbon isotope ratios in toenails) experience 90% reduced reproductive success, measured as number of young fledged, compared to those in better habitats, but only in males (both yearling and older) and not females (Rushing et al. 2016). Earlier arriving males in northern Michigan settle on higher quality territories and hatch nestlings relatively early; earlier females start laying earlier, clutch size is greater, and their nestlings are heavier, increasing the likelihood of offspring recruiting to adult populations (Smith and Moore 2005a). Circumstances during northward migration in spring can also carry over into the breeding season, as suggested by some individuals (in some years) arriving to breed with excess body fat, and increased reproductive performance by females, in particular, that arrive to breed with higher body lipid deposits (i.e., better body condition; Smith and Moore 2003, Smith and Moore 2005b). These latter two studies provide a direct link between body condition during migration and subsequent reproductive success. Carry-over effects from winter habitats and overall winter habitat quality have also been linked to natal and breeding dispersal (Studds et al. 2008, Rushing et al. 2015a). Effects of migration per se on redstarts, and several other long-distance migrant warblers, confirmed by delayed arrival and poorer body condition in years with colder spring weather (average May temperature) in Manitoba stopover site (González-Prieto and Hobson 2013).

No information available for redstarts on possible carryover effects from summer breeding grounds to wintering areas. Such effects have been shown in another long-distance migratory parulid, the Louisiana Waterthrush (Parkesia motacilla; Latta et al. 2016).

Density Dependence and Population Regulation

Based on 11-year study in Ontario, redstart reproductive success decreased with density of territories present (but see Sherry et al. 2015), indicating potential for population regulation during the breeding season (McKellar et al. 2014). Both number of offspring fledged per nest and fledgling success rate decreased in years with more total territories, and within years on territories with more close neighbor territories (within 200 m). Nest predation was implicated as the mechanism for this density-dependence because (1) many nest predators in this study (e.g., crows, accipiters) have large home ranges allowing encounter with many nests and development of a stronger search image at higher nest densities, (2) the proportion of nesting attempts that succeeded per female declined with more close neighbor nests, and most nest failure was attributable to predators, and (3) greater residual (abnormally high) nestling mass was found in high-density territories within years, the opposite of expectation if food competition were the mechanism. McKellar et al. 2014 attribute high-density territories to aggregation in areas with more food, resulting in relatively high fledgling mass despite higher nest predation risk. Density of breeding adults did not affect annual adult survival, but habitat-specific density did influence the likelihood of individuals dispersing between two nearby study sites (McKellar et al. 2015).

Density-dependence of body condition and apparent annual survival also documented in winter, but mediated by habitat quality and sex-specific habitat use (Marra et al. 2015): In highest density, male-dominated (60% of individuals on average) black mangrove wintering habitat, body condition and annual survival decreased with density of territorial individuals among years, indicative of a crowding mechanism, probably mediated via food abundance (see also Cooper et al. 2015), but did not vary with annual rainfall. By contrast, in adjacent thorn scrub habitat, body condition increased while survival remained unchanged with increased individual density, probably via a density-independent, food-related mechanism: More rainfall led to more insect food in some years, allowing more individuals to crowd into this habitat and maintain better body condition (Studds and Marra 2007, Studds and Marra 2011, Studds and Marra 2012, Wilson et al. 2013b). Thus males appear to be limited by different processes in winter, due at least in part to their predominance in mangroves (wetter, more favorable habitats) compared with females, which occur mostly in thorn scrub (drier, less favorable habitat). The influence on body condition and annual survival of transient (non-territorial) individuals, abundant in both these habitats (Peele et al. 2015), not yet known, but transients appear to play role in population regulation at greater (among-habitat, and regional) spatial scales (Peele 2015).

Transient individuals may be linked to another population regulatory mechanism suggested to be operating in non-breeding season, namely a buffer effect (Marra et al. 2015), in which proportionately more individuals may be constrained (e.g., through territorial aggression) to occupy worse habitat when total density is greater. The result is that individuals in poorer habitat can move into vacancies in the better habitat, thus buffering the population. This may be operating in wintering redstarts, as evidenced by the observations that (1) proportionately more individuals occur in thorn scrub in years with overall greater density (Marra et al. 2015), (2) replacement rate of individuals following territory removals is greater in the better quality black mangroves than adjacent thorn scrub (Marra 2000), and (3) individuals replacing mangrove birds tend to be those moving in from scrub habitat (Studds and Marra 2005). Because multiple regulatory mechanisms occur within a season (winter) and between seasons, the relative importance of these different mechanisms on population dynamics is not known, and will probably require year-round and large scale population modeling and include data from a number of localities in both breeding and wintering areas.

Recommended Citation

Sherry, T. W., R. T. Holmes, P. Pyle, and M. A. Patten (2016). American Redstart (Setophaga ruticilla), version 3.0. In The Birds of North America (P. G. Rodewald, Editor). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bna.amered.03