RESEARCH ARTICLE
Timing of Feather Molt Related to Date of Spring Migration in Male White‐Throated Sparrows, Zonotrichia albicollis DANIEL AARON CRISTOL*, KAREN MICHELLE JOHNSON, KENDELL DALY JENKINS, AND DANA MICHELLE HAWLEY Institute for Integrative Bird Behavior Studies, Department of Biology, The College of William and Mary, Williamsburg, Virginia
ABSTRACT
In migratory birds, the ability to depart wintering grounds at the appropriate time is an important determinant of fitness. Understanding the regulation of this timing will be essential for predicting whether timing of bird migration keeps up with global climate change. We examined whether the timing of the late‐winter molt, in which white‐throated sparrows (Zonotrichia albicollis) replace head and body feathers in advance of mating, may constrain the timing of northward migratory departure. In an observational study, we found a significant correlation between timing of molt and the date on which free‐living male white‐throated sparrows disappeared from our study site during migration. The following year, we tested whether experimentally manipulating molt date by advancing photoperiod during temporary captivity would subsequently advance disappearance date once the birds were released. Sparrows that were experimentally induced to molt early disappeared from the wintering site before controls. However, the captive control birds also molted and disappeared from the site earlier than free‐living controls, suggesting that the diet during captivity had played a role. In the third winter we completed the study by advancing or delaying molt using only dietary manipulation. Together, these results show that the ability to molt early in spring is related to early disappearance from the wintering site. Early molt likely has carry‐over effects on reproduction and the requirements of molt may prevent populations from adjusting migration timing in response to global climate change. J. Exp. Zool. 321A:586–594, 2014. © 2014 Wiley Periodicals, Inc.
J. Exp. Zool. 321A:586–594, 2014
How to cite this article: Cristol DA, Johnson KM, Jenkins KD, Hawley DM. 2014. Timing of feather molt related to date of spring migration in male white‐throated sparrows, Zonotrichia albicollis. J. Exp. Zool. 321A:586–594.
Early arrival on the breeding range can lead to higher reproductive success in migratory birds, sometimes with important consequences for conservation of avian biodiversity (Rockwell, 2013). Yet there remains considerable variation among individuals in dates of spring migration and arrival at breeding sites (Stanley et al., 2012). Some of this variation stems from differential migration, in which different segments of a population migrate different distances in autumn to separate wintering areas (Cristol et al., '99). However, even within members of one age‐sex class of a population wintering in the same geographic area, spring departure dates can vary considerably (e.g., Dugger, '97). It is this variation, some of which is heritable, that may allow migratory bird populations to evolve in synch with the earlier spring onset of
Grant sponsor: National Science Foundation; grant number: IBN9876108. Present address of Kendell Daly Jenkins is Geographic Information Systems, Henrico County, Henrico, Virginia. Present address of Dana Michelle Hawley is Department of Biological Sciences, Virginia Tech, Blacksburg, Virginia. Correspondence to: D. A. Cristol, Department of Biology, College of William and Mary, Williamsburg, VA 23187‐8795. E-mail: [email protected] Received 2 February 2014; Revised 10 September 2014; Accepted 11 September 2014 DOI: 10.1002/jez.1899 Published online 6 October 2014 in Wiley Online Library (wileyonlinelibrary.com).
© 2014 WILEY PERIODICALS, INC.
MOLT TIMING CONSTRAINS BIRD MIGRATION optimal breeding conditions expected as the result of global climate change (Pulido, 2007). Events in one season can have profound effects on an individual animal's success in the following season, and these carry‐over effects may explain a large amount of the variation in individual fitness (Bêty et al., 2004; Harrison et al., 2011). In migratory birds, differences in wintering habitat quality can be related to differences in arrival date at nesting sites. Male songbirds wintering in higher‐quality habitat may arrive earlier, and thus attain greater reproductive success (Norris et al., 2004). It is imperative that we gain a better understanding of the factors regulating timing of bird migration because of its importance in reproductive success and the potential for adapting to global climate change. One factor intimately associated with timing of migration is timing of molt, because molt and migration are both energetically expensive activities difficult to carry on simultaneously (Hemborg and Lundberg, '98). While much is known of proximate mechanisms influencing departure for migration, for example day‐length triggered endocrine fluxes (Wingfield et al., '90), temperature (Gordo, 2007), and energetic reserves (Schaub et al., 2008), less is known about the relationship between timing of molt and migration. Most of what we know comes from the better‐studied autumn migration. In captive‐reared songbirds, photoperiodic manipulation of post‐juvenile molt termination was tightly correlated with onset of autumn migratory activity, but the degree of overlap was resilient to environmental perturbation and molt and migration timing appeared to be genetically correlated (Pulido and Coppack, 2004). This suggests that timing of autumn migration may be constrained in the face of global climate change. Spring migration timing is less constrained by the events of the preceding breeding season than autumn migration, and thus could be more labile and responsive to climate change. However, a comparative study of extent and occurrence of late‐winter molt (referred to hereafter as “prealternate” molt because it precedes the “alternate” plumage present during breeding) found that it was related to both diet and migration distance, suggesting the possibility of interdependence between diet, molt, and migration (Tökölyi et al., 2008). This linkage has been further explored by Danner et al. (2014); who found that molt can be advanced in a migratory songbird by improving diet in late winter. We hypothesized that there is a link between date of late‐winter molt (referred to hereafter as “prealternate” molt because it precedes the “alternate” plumage present during breeding) and date of spring migratory departure. If prealternate molt involved wing and tail feathers, it would be obvious that migration could not proceed until completion of molt. However, most migratory birds replace their entire plumage only once a year, after breeding but before autumn migration, whereas the prealternate molt usually involves a limited number of feathers, such as those involved in mate choice and territorial defence. Here we describe
587 the first experimental test of whether timing of prealternate molt is linked to date of disappearance from the wintering grounds.
MATERIALS AND METHODS Study Species and Site The well‐studied (Falls and Kopachena, 2010) white‐throated sparrow (Zonotrichia albicollis) was chosen for this experiment because it exhibits a high degree of variance in timing of spring migration; we observed that wintering males disappeared from our study site in Virginia, USA, over a period of 35 days in spring 1998 (D. Cristol, unpublished data), presumably beginning their northward migrations. White‐throated sparrows undergo an extensive prealternate molt in the spring involving the elaborately patterned black, white, tan, and yellow plumage of the head and neck, as well as some feather tracts on the sides, flanks, and back (Kuenzel and Helms, '74). We carried out a sequence of three consecutive studies over three winter/springs: January 15–May 20, 1998, January 6– May 13, 1999, and January 18–May 10, 2000 (for analysis all dates were converted to Julian dates; day 0 ¼ January 1), along the edge of a woodlot in Williamsburg, Virginia (N 37.2615; W 76.7069). Vegetation was primarily deciduous with a young oak‐hickory canopy and thick underbrush. The 4‐ha study site was bordered by a college campus to the north, a former sanitary landfill in the early stages of succession to the south, a wooded cemetery to the east, and mature forest to the west. Eight trapping stations were established approximately 50 m apart in a line parallel to, and 20 m from, the edge of the cemetery in 1998; four of these were continued in 1999 and three in 2000. Birds were captured using mist nets and ground traps baited approximately every third day with grain. All birds were color‐banded upon capture (white, red, black, light blue, or light green; Perler Bead Company plastic play beads). We determined whether individual white‐ throated sparrows were still present on the study site by observing them from outdoor blinds located at the trapping sites, 2–6 hr per day, 3–5 days per week from February 10 to May 20, 1998, and 2–14 hr daily from March 18 to May 13, 1999 and March 29 to May 10, 2000. The disappearance of a sparrow from the study site in early spring could have many explanations, including death, selection of an alternate wintering site, movement to a staging area, or initiation of migration. Based on the annual date of arrival of unbanded birds, assumed to be migrants from the south, and disappearance of multiple banded birds daily during periods of favorable weather, we set a date for the “beginning of migration” each spring (see below). We made the operational assumption that all birds disappearing from the study site after that date had departed for migration. The research presented here was described in animal use protocol 9850 approved January 4, 1999 by the IACUC of the J. Exp. Zool.
588 College of William and Mary. Birds were captured and color banded under a USGS Bird Banding Lab permit, and held in accordance with all applicable Virginia and USA permits. Study 1: Unmanipulated Timing of Molt and Migration (1998) To determine when each bird molted we quantified molt every time an individual was captured. Beginning on March 1, molt of the head feathers was scored as follows. The head was divided into eight areas (left and right lores combined, left and right superciliary stripes combined, left and right lateral stripes combined, median stripe, left and right auriculars combined, nape, chin, left and right eye margins combined). Each area was scored on a scale of 0–3 where 0 ¼ no incoming sheathed (hereafter “pin”) feathers, 1 ¼1–10 pin feathers, 2 ¼ 11–50 pin feathers, 3 ¼ more than 50 pin feathers. A score was calculated for each bird on each recapture by summing the scores for all areas. Birds typically progressed from 0 through a maximum recorded score of 24 (more than 50 pin feathers in each of the eight regions), and then declined as pin feathers erupted from their sheaths and were no longer counted as such. Molt can vary in terms of date of onset and/or rate of progression and is notoriously difficult to quantify because it has a long duration, progresses differently among individuals, and varies across different parts of the body. Because the number and date of recaptures varied, we could not track the progress of each bird's molt at a fine scale. Instead, we used two indices that could be calculated even for birds with very different recapture schedules. One index was the estimated date of molt onset. We back‐calculated to estimate date of onset using a typical trajectory based on data from a group of birds for which we had excellent data on onset. To produce this plot of typical onset we used data from captive birds held under natural photoperiod and examined weekly during the onset of molt in 1999. Ear and chin data were used because these were the first areas to molt in every captive bird examined in this set. Using only data from ear and chin molts we developed a plot of the changes in molt scores over 7‐day periods. We could then back‐calculate the date of molt onset for any free‐living bird that was captured with ear or chin molt before its date of peak molt. Our other index was date of peak molt, that is, the estimated date on which the bird was molting the most feathers on its eight head regions combined. First, we used the nine free‐living birds for which we had the most recaptures to plot a typical molt progression. We then divided the molt period into 5‐day intervals and plotted the mean molt scores against time since onset of molt (see above) for our nine model birds. Using this plot we could then estimate the date of peak molt score for any bird captured two or more times during the increasing phase of molt by placing it on the typical trajectory. The two indices of molt were both intended only to distinguish early birds from late birds, and it should be noted that they were calculated based on entirely different data using independent sets of birds. J. Exp. Zool.
CRISTOL ET AL. Studies 2 and 3: Manipulations of Molt After examining the naturally occurring relationship between molt and disappearance from the wintering site in free‐living birds during 1998, we experimentally manipulated molt in temporarily captive birds the following winter/spring and examined whether birds with advanced molt also had advanced spring disappearance dates. In 1999 we advanced molt by advancing photoperiod for wild‐caught sparrows temporarily housed in an aviary and given a high‐quality diet. Because we observed that temporarily captive control birds, maintained on high‐quality diet but natural photoperiod, also advanced molt somewhat compared to free‐living birds, the following year we used strictly diet manipulation to speed or slow molt. We used the same molt indices described above for free‐living birds, but because birds were experimentally manipulated we calculated treatment‐specific molt trajectories from the most complete birds in each of our treatment groups and estimated scores for the remaining birds using these trajectories. It should be noted that sample sizes for onset and peak of molt, as well as date of disappearance, are smaller than total sample sizes because, even for captive birds, we were sometimes unable to calculate these dates if we did not have sufficient recaptures after release. Calculating molt onset required a reliable estimate of peak molt date (except for some frequently‐caught free‐living birds in which it could be documented at a finer scale). If subjects had not molted enough before release, and we did not recapture them enough times to estimate peak, then we also could not estimate onset. Study 2: Photoperiodic Manipulation of Molt (1999) In 1999, we tested whether molt date constrained departure date by providing some white‐throated sparrows with an advanced photoperiod and a high‐quality diet, which we expected to cause them to molt early. All adult birds caught before April 1 with wings 70 mm were assumed to be males and were assigned to one of three treatment groups: advanced, captive control, or free‐ living. We established two advanced and two control flocks (n ¼ 21 birds per flock) in four indoor (but unheated), light‐proof rooms (4.5 4.3 2.7 m). After one week of acclimatization under simulated natural day length, we began increasing the day length in the advanced rooms by 5 min/day to induce early molt. The photoperiod in control rooms continued to track natural changes in day length. Our third treatment was 42 free‐living birds released at their original sites of capture after color‐banding and measuring. It should be noted that 50% of our birds had wings 74 mm and were thus unquestionably males, but by including in our experimental study birds with wings as small as 70 mm, some (5 unless shown in parentheses.
early also disappeared earlier from the wintering site. This suggests the possibility that molt is a constraint on departure and must be at least partly completed before birds can migrate to breeding grounds. However, causality cannot be inferred from the existence of a relationship between molt and disappearance dates. The same individuals may have the ability to both molt and depart early, for example dominant birds or those in better nutritional condition. To explore whether variables that influenced molt date also influenced disappearance date, we did two experimental tests of the relationship. We first induced an early molt onset and peak in temporarily captive birds by advancing photoperiod in an indoor aviary. These advanced birds were then released back onto the study site several weeks before migration, along with controls held under identical conditions but on natural photoperiod. Advanced birds disappeared from the wintering site more than 2 weeks before the captive controls, which left no earlier than free‐living birds. We were surprised to find that the captive controls molted 3 weeks earlier than their free‐living counterparts. We concluded that their captive diet had likely played a role in advancing molt, and we used this finding to design our next experimental manipulation of molt. Because the initial photoperiod manipulation also caused experimental birds to increase their singing behavior early (unpublished data), we assume that a cascade of physiological changes was triggered by the advanced photoperiod, and that some factor(s) other than early molt, such as endocrine fluxes, could have contributed to the early migration. Because the endogenous control of prealternate molt is poorly understood, we instead repeated the experimental manipulation of molt date
in an outdoor aviary under natural photoperiod using another exogenous cue, alteration of diet, to shift molt date. One group of birds received an ad libitum diet high in calories and protein that advanced the onset of molt by approximately one week relative to free‐living birds, while another received an insufficient diet that delayed the peak of molt (although onset was not delayed significantly relative to free‐living birds). Birds in the advanced treatment group disappeared from the wintering site approximately 5 days earlier than free‐living or delayed birds, which did not differ from one another. Our results are interesting for two reasons. First, these results suggest that diet can affect molt with a cascading effect on migration timing, such that birds induced to molt a week early apparently departed almost one week early for spring migration. This is consistent with the recent demonstration that free‐living swamp sparrows provisioned with food underwent early molt (Danner et al., 2014). Together these findings suggest a tight link between food availability and timing of molt and migration. Second, the ease with which we were able to manipulate molt timing contrasts with some previous studies (e.g., Murphy, '96) and raises questions about the exogenous regulation of spring and autumn molt; for example prebasic molt in bluethroats (Luscinia svecica) was not altered by photoperiod or captivity (Lindström et al., '94) whereas molt and timing of autumn migration behavior both responded to experimental change in photoperiod in hand‐reared juvenile blackcaps (Sylvia atricapilla, Pulido and Coppack, 2004). The pre‐eminence of photoperiod as an exogenous cue for molt may vary between spring and fall, or the role of food availability may have been overlooked in both seasons. J. Exp. Zool.
592 Because we manipulated molt by changing additional variables (photoperiod or diet), we still cannot conclusively point to molt as the causative agent for estimated departure date. Other explanations must be considered. For example, the early molting individuals might have migrated early in 1998 not because they had molted first, but because they were dominant, and dominant white‐throated sparrows may have had attributes that made them both molt and migrate early. In the 1999 experiment the photoperiodically advanced birds might have left the wintering site early not because they were done molting, but because premature environmental signals had induced migratory readiness in addition to an early molt. Finally, birds that were fed rich diets in 2000 might have migrated early not because of molt, but because they were in better nutritional condition. Only an experiment in which birds were forcibly molted by feather removal could establish causality with certainty, but it is not clear whether such an experiment is possible. Molt involves the pushing out of old feathers by emerging feathers, a process that would not be mimicked through experimental plucking. Additionally, caution must be used when interpreting the relationship between molt and migration timing because our estimates of molt onset and peak were based on an imperfect series of measurements of molt progression. We used data from sets of birds for which we had numerous samples to produce “typical” curves for estimating onset and peak, but the finding that onset and peak responded independently to manipulations and did not both correlate with timing of disappearance from the study site suggests that individual birds progressed differently from onset to peak. Two unexpected results require further explanation. In the 1999 experimental manipulation of molt using photoperiod, the captive control group molted somewhat earlier than their free‐ living counterparts, although their molt was not nearly as rapid as that of the advanced birds. This was most likely the result of their superior diet relative to free‐living birds or the reduced energetic demands of indoor life (although normal outdoor temperatures were maintained), but may also have occurred because these captive controls heard a greater amount of singing by the experimental birds housed in the same building. Interestingly, when these captive controls were released, they did not migrate earlier than their free‐living counterparts, despite having begun molt early. Our tentative explanation for their failure to migrate early stems from the fact that these birds had not reached peak molt before they were released. Their progression through molt may have been slowed once they were released and faced the same nutritional constraints as the free‐living birds. The experimental photoperiodically advanced birds, on the other hand, were mostly past peak molt when they were released, so could have migrated without further delay. The other unexpected finding was that in the 2000 dietary manipulation of molt, which included a delayed as well as an advanced treatment, the birds molting late were able to migrate at J. Exp. Zool.
CRISTOL ET AL. the same time as their free‐living counterparts, despite the fact that their molts had progressed less far on the release date. Two explanations deserve consideration. First, it may be that molting early facilitates early migration, but experimentally delaying molt does not delay migration because at some point birds will migrate regardless of molt status. If this were the case, molt would not be an absolute constraint on departure date, but might still explain variation in timing early in spring. Alternately, the effect of diet may be swamped by the effect of accelerating increases in day length that occur later in spring. Another explanation would be that delayed birds were able to rapidly catch up to the molt status of their free‐living counterparts just before departure because they were no longer eating a poor diet, as occurred when malnourished white‐crowned sparrows (Zonotrichia leucophrys) were returned to normal diets (Murphy et al., '88). It may be that late in the season there is so much high‐quality food available that all birds can molt rapidly. A post hoc analysis of molt rate supports this explanation, because rate of molt increased dramatically after release in these delayed birds (D. Cristol, unpublished data). Again, perhaps only earlier in the spring, when protein‐rich food may be limiting, would small differences in molt schedules lead to earlier migration. In conclusion, we suggest that molt and departure for spring migration are related in a wintering songbird, just as they are during autumn migration (Stutchbury et al., 2011). These important life‐history events are connected directly, or through a common response to environmental cues. Further experiments will be necessary before we can conclude that molt is a primary constraint on timing of migration, but this is a parsimonious explanation of our results: molt and departure were correlated in free‐living birds and two different methods of advancing molt accelerated apparent departure. One possible mechanism for the relationship between molt and migration timing would be the following scenario: (1) socially dominant individuals obtain more of the rare high‐protein food resources on the winter range; (2) dominants are thus in better condition which allows them to molt earlier; (3) earlier‐molting, socially dominant birds migrate earlier to breeding grounds. There is ample support for a link between social status and access to resources in white‐throated sparrows (Schneider, '84; Piper and Wiley, '90). A study on our sparrow population indicated that degree of despotism in captive flocks affected rate of replacement feather growth (Jenkins et al., 2001). The present study establishes a link between timing of molt and migration for this species. Thus, there is some evidence for all aspects of a proposed mechanistic link between dominance status, molt, and migration. This is the same type of relationship proposed to explain why warblers from better tropical habitats arrived early on temperate wintering grounds. Based on stable isotope profiles, Marra et al. ('98) determined that early spring migrant male American redstarts (Setophaga ruticilla) were more likely to have wintered in higher quality mangrove habitat, whereas late male and female
MOLT TIMING CONSTRAINS BIRD MIGRATION redstarts came from lower‐quality scrub forest. Habitat segregation on the wintering grounds was influenced by dominance status and affected nutritional condition and spring migration date, so there is a case for the importance of wintering ground conditions in determining breeding success (Marra, 2000; Marra and Holmes, 2001). In another songbird, the blackcap, it has also been suggested that only the fittest males are capable of an early spring departure (Izhaki and Maitav, '98). Our results suggest that in species with a significant prealternate molt, the ability to complete feather replacement early may be one mechanism for the relationship between dominance, habitat quality/nutritional status, and migration date. Previous studies have shown that status or condition were related to speed or timing of molt and reproductive success (e.g., Mulder and Magrath, '94; Robertson et al., '98), but ours is the first study to suggest by experimental manipulation that molt and timing of migration are tightly linked, with implications for whether and how much white‐ throated sparrow populations will be able to adapt migration timing to expected earlier onset of spring.
ACKNOWLEDGMENTS We are greatly indebted to Jessica Armstrong, Brandi Jenks Brown, Dave Cerasale, Elise Donnelly, Chris Farabaugh, Dee Fritz, Chris Gleason, Josh LeClerc, Katie Martin, Jen McCall, Andrew McCormick, Katie Murphy, Hagai Nassau, Erica Reynolds, and Zakiya Thomas for spending many hours in the blind confirming the continued presence of each bird daily. Walter Piper and Thomas Grubb, Jr. made helpful suggestions on an earlier version of the manuscript. This study was supported by NSF grant IBN‐ 9876108 as well as the Virginia Society of Ornithology, the Virginia Academy of Science, the Williamsburg Bird Club, and the College of William & Mary.
LITERATURE CITED Bêty J, Giroux J‐F, Gauthier G. 2004. Individual variation in timing of migration: Causes and reproductive consequences in greater snow geese (Anser caerulescens atlanticus). Behav Ecol Sociobiol 57:1–8. Cristol DA, Baker MB, Carbone C. 1999. Differential migration revisited: Latitudinal segregation by age and sex class. Curr Ornith 15:33–88. Danner RM, Greenberg RS, Danner JE, Walters JR. 2014. Winter food limits timing of moult in a short‐distance migratory bird. Funct Ecol Available online ahead of print: DOI: 10.1111/1365–2435. 12322. Dugger BD. 1997. Factors influencing the onset of spring migration in Mallards. J Field Ornith 68:331–337. J.B. Falls, J.G. Kopachena.2010. White‐throated Sparrow (Zonotrichia albicollis), The Birds of North America Online. A. Poole. Ithaca: Cornell Lab of Ornithology Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/12DOI:10. 2173/bna.128 .
593 Gordo O. 2007. Why are bird migration dates shifting? A review of weather and climate effects on avian migration phenology. Clim Res 35:37–58. Harrison XA, Blount JD, Inger R, Norris DR, Bearhop S. 2011. Carry‐ over effects as drivers of fitness differences in animals. J Anim Ecol 80:4–18. Hemborg C, Lundberg A. 1998. Costs of overlapping reproduction and moult in passerine birds: An experiment with the pied flycatcher. Behav Ecol Sociobiol 43:19–23. Izhaki I, Maitav A. 1998. Blackcaps Sylvia atricapilla stopping over at the desert edge: Inter‐ and intra‐sexual differences in spring and autumn migration. Ibis 140:234–243. Jenkins KD, Hawley DM, Farabaugh CS, Cristol DA. 2001. Ptilochronology reveals differences in condition of captive white‐throated sparrows. Condor 103:579–586. Kuenzel WJ, Helms CW. 1974. An annual cycle study of tan‐striped and white‐striped white‐ throated sparrows. Auk 91:44–53. Lindström A, Daan S, Visser GH. 1994. The conflict between moult and migratory fat deposition: a photoperiodic experiment with bluethroats. Anim Behav 48:1173–1181. Marra PP. 2000. The role of behavioral dominance in structuring habitat occupancy of a migrant bird during the nonbreeding season. Behav Ecol 11:299–308. Marra PP, Hobson KA, Holmes RT. 1998. Linking winter and summer events in a migratory bird by using stable carbon isotopes. Science 282:1884–1886. Marra PP, Holmes RT. 2001. Consequences of dominance‐mediated habitat segregation in American redstarts during the nonbreeding season. Auk 118:92–104. Mulder RA, Magrath MJL. 1994. Timing of prenuptual molt as a sexually selected indicator of male quality in superb fairy‐wrens (Malurus cyaneus). Anim Behav 5:393–400. Murphy ME. 1996. Energetics and nutrition of molt. Carey C, Avian Energetics and Nutritional Ecology. New York: Chapman and Hall. pp 158–198. Murphy ME, King JR, Lu J. 1988. Malnutrition during the postnuptial molt of white‐ crowned sparrows: Feather growth and quality. Can J Zool 66:1403–1413. Norris DR, Marra PP, Kyser TK, Sherry TW, Ratcliffe LM. 2004. Tropical winter habitat limits reproductive success on the temperate breeding grounds in a migratory bird. Proc R Soc Lond B 271:59–64. Piper WH, Wiley RH. 1990. The relationship between social dominance, subcutaneous fat, and annual survival in wintering white‐throated sparrows (Zonotrichia albicollis). Behav Ecol Sociobiol 26:201–208. Piper WH, Wiley RH. 1991. Effects of laparotomies on wintering white‐throated sparrows and the usefulness of wing chord as a criterion for sexing. J Field Ornith 62:40–45. Pulido F. 2007. Phenotypic changes in spring arrival: Evolution, phenotypic plasticity, effects of weather and condition. Clim Res 35:5–23. J. Exp. Zool.
594 Pulido F, Coppack T. 2004. Correlation between timing of juvenile moult and onset of migration in the blackcap, Sylvia atricapilla. Anim Behav 68:167–173. Robertson GJ, Cooke F, Goudie RI, Boyd WS. 1998. Molt speed predicts pairing success in male harlequin ducks. Anim Behav 55:1677–1684. Rockwell SM. 2013. Carry-over effects from the non-breeding season influence spring arrival dates, reproductive success, and survival in an endangered migratory bird, the Kirtland's warbler (Setophaga kirtlandii). [Dissertation]. College Park, MD: University of Maryland http://hdl.handle.net/1903/14363. Schaub M, Jenni L, Bairlein F. 2008. Fuel stores, fuel accumulation, and the decision to depart from a migration stopover site. Behav Ecol 19:657–666. Schneider KJ. 1984. Dominance, predation, and optimal foraging in white‐throated sparrow flocks. Ecology 65:1820–1827.
J. Exp. Zool.
CRISTOL ET AL. Siegel S, Castellan NJ, Jr. 1988. Nonparametric Statistics for the Behavioural Sciences. 2nd edition. New York: McGraw‐Hill. Stanley C, MacPherson M, Fraser KC, McKinnon EA, Stutchbury BJM. 2012. Repeat tracking of individual songbirds reveals consistent migration timing but flexibility in route. PLoS ONE 7:-e40688.DOI: 10.1371/journal.pone.0040688. Stutchbury BJM, Gow EA, Done T. 2011. Effects of post‐ breeding moult and energetic condition on timing of songbird migration into the tropics. Proc R Soc Lond B 178:131–137. Tökölyi J, Bókony V, Barta Z. 2008. Seasonal colour change by moult or by the abrasion of feather tips: A comparative study. Biol J. Linn Soc 94:711–721. Wingfield JC, Schwabl H, Mattocks PW, Jr. 1990. Endocrine mechanisms of bird migration. Gwinner E, Bird Migration. Berlin, Germany: Springer‐Verlag. pp 232–256.
Timing of Feather Molt Related to Date of Spring Migration in Male White‐Throated Sparrows, Zonotrichia albicollis DANIEL AARON CRISTOL*, KAREN MICHELLE JOHNSON, KENDELL DALY JENKINS, AND DANA MICHELLE HAWLEY Institute for Integrative Bird Behavior Studies, Department of Biology, The College of William and Mary, Williamsburg, Virginia
ABSTRACT
In migratory birds, the ability to depart wintering grounds at the appropriate time is an important determinant of fitness. Understanding the regulation of this timing will be essential for predicting whether timing of bird migration keeps up with global climate change. We examined whether the timing of the late‐winter molt, in which white‐throated sparrows (Zonotrichia albicollis) replace head and body feathers in advance of mating, may constrain the timing of northward migratory departure. In an observational study, we found a significant correlation between timing of molt and the date on which free‐living male white‐throated sparrows disappeared from our study site during migration. The following year, we tested whether experimentally manipulating molt date by advancing photoperiod during temporary captivity would subsequently advance disappearance date once the birds were released. Sparrows that were experimentally induced to molt early disappeared from the wintering site before controls. However, the captive control birds also molted and disappeared from the site earlier than free‐living controls, suggesting that the diet during captivity had played a role. In the third winter we completed the study by advancing or delaying molt using only dietary manipulation. Together, these results show that the ability to molt early in spring is related to early disappearance from the wintering site. Early molt likely has carry‐over effects on reproduction and the requirements of molt may prevent populations from adjusting migration timing in response to global climate change. J. Exp. Zool. 321A:586–594, 2014. © 2014 Wiley Periodicals, Inc.
J. Exp. Zool. 321A:586–594, 2014
How to cite this article: Cristol DA, Johnson KM, Jenkins KD, Hawley DM. 2014. Timing of feather molt related to date of spring migration in male white‐throated sparrows, Zonotrichia albicollis. J. Exp. Zool. 321A:586–594.
Early arrival on the breeding range can lead to higher reproductive success in migratory birds, sometimes with important consequences for conservation of avian biodiversity (Rockwell, 2013). Yet there remains considerable variation among individuals in dates of spring migration and arrival at breeding sites (Stanley et al., 2012). Some of this variation stems from differential migration, in which different segments of a population migrate different distances in autumn to separate wintering areas (Cristol et al., '99). However, even within members of one age‐sex class of a population wintering in the same geographic area, spring departure dates can vary considerably (e.g., Dugger, '97). It is this variation, some of which is heritable, that may allow migratory bird populations to evolve in synch with the earlier spring onset of
Grant sponsor: National Science Foundation; grant number: IBN9876108. Present address of Kendell Daly Jenkins is Geographic Information Systems, Henrico County, Henrico, Virginia. Present address of Dana Michelle Hawley is Department of Biological Sciences, Virginia Tech, Blacksburg, Virginia. Correspondence to: D. A. Cristol, Department of Biology, College of William and Mary, Williamsburg, VA 23187‐8795. E-mail: [email protected] Received 2 February 2014; Revised 10 September 2014; Accepted 11 September 2014 DOI: 10.1002/jez.1899 Published online 6 October 2014 in Wiley Online Library (wileyonlinelibrary.com).
© 2014 WILEY PERIODICALS, INC.
MOLT TIMING CONSTRAINS BIRD MIGRATION optimal breeding conditions expected as the result of global climate change (Pulido, 2007). Events in one season can have profound effects on an individual animal's success in the following season, and these carry‐over effects may explain a large amount of the variation in individual fitness (Bêty et al., 2004; Harrison et al., 2011). In migratory birds, differences in wintering habitat quality can be related to differences in arrival date at nesting sites. Male songbirds wintering in higher‐quality habitat may arrive earlier, and thus attain greater reproductive success (Norris et al., 2004). It is imperative that we gain a better understanding of the factors regulating timing of bird migration because of its importance in reproductive success and the potential for adapting to global climate change. One factor intimately associated with timing of migration is timing of molt, because molt and migration are both energetically expensive activities difficult to carry on simultaneously (Hemborg and Lundberg, '98). While much is known of proximate mechanisms influencing departure for migration, for example day‐length triggered endocrine fluxes (Wingfield et al., '90), temperature (Gordo, 2007), and energetic reserves (Schaub et al., 2008), less is known about the relationship between timing of molt and migration. Most of what we know comes from the better‐studied autumn migration. In captive‐reared songbirds, photoperiodic manipulation of post‐juvenile molt termination was tightly correlated with onset of autumn migratory activity, but the degree of overlap was resilient to environmental perturbation and molt and migration timing appeared to be genetically correlated (Pulido and Coppack, 2004). This suggests that timing of autumn migration may be constrained in the face of global climate change. Spring migration timing is less constrained by the events of the preceding breeding season than autumn migration, and thus could be more labile and responsive to climate change. However, a comparative study of extent and occurrence of late‐winter molt (referred to hereafter as “prealternate” molt because it precedes the “alternate” plumage present during breeding) found that it was related to both diet and migration distance, suggesting the possibility of interdependence between diet, molt, and migration (Tökölyi et al., 2008). This linkage has been further explored by Danner et al. (2014); who found that molt can be advanced in a migratory songbird by improving diet in late winter. We hypothesized that there is a link between date of late‐winter molt (referred to hereafter as “prealternate” molt because it precedes the “alternate” plumage present during breeding) and date of spring migratory departure. If prealternate molt involved wing and tail feathers, it would be obvious that migration could not proceed until completion of molt. However, most migratory birds replace their entire plumage only once a year, after breeding but before autumn migration, whereas the prealternate molt usually involves a limited number of feathers, such as those involved in mate choice and territorial defence. Here we describe
587 the first experimental test of whether timing of prealternate molt is linked to date of disappearance from the wintering grounds.
MATERIALS AND METHODS Study Species and Site The well‐studied (Falls and Kopachena, 2010) white‐throated sparrow (Zonotrichia albicollis) was chosen for this experiment because it exhibits a high degree of variance in timing of spring migration; we observed that wintering males disappeared from our study site in Virginia, USA, over a period of 35 days in spring 1998 (D. Cristol, unpublished data), presumably beginning their northward migrations. White‐throated sparrows undergo an extensive prealternate molt in the spring involving the elaborately patterned black, white, tan, and yellow plumage of the head and neck, as well as some feather tracts on the sides, flanks, and back (Kuenzel and Helms, '74). We carried out a sequence of three consecutive studies over three winter/springs: January 15–May 20, 1998, January 6– May 13, 1999, and January 18–May 10, 2000 (for analysis all dates were converted to Julian dates; day 0 ¼ January 1), along the edge of a woodlot in Williamsburg, Virginia (N 37.2615; W 76.7069). Vegetation was primarily deciduous with a young oak‐hickory canopy and thick underbrush. The 4‐ha study site was bordered by a college campus to the north, a former sanitary landfill in the early stages of succession to the south, a wooded cemetery to the east, and mature forest to the west. Eight trapping stations were established approximately 50 m apart in a line parallel to, and 20 m from, the edge of the cemetery in 1998; four of these were continued in 1999 and three in 2000. Birds were captured using mist nets and ground traps baited approximately every third day with grain. All birds were color‐banded upon capture (white, red, black, light blue, or light green; Perler Bead Company plastic play beads). We determined whether individual white‐ throated sparrows were still present on the study site by observing them from outdoor blinds located at the trapping sites, 2–6 hr per day, 3–5 days per week from February 10 to May 20, 1998, and 2–14 hr daily from March 18 to May 13, 1999 and March 29 to May 10, 2000. The disappearance of a sparrow from the study site in early spring could have many explanations, including death, selection of an alternate wintering site, movement to a staging area, or initiation of migration. Based on the annual date of arrival of unbanded birds, assumed to be migrants from the south, and disappearance of multiple banded birds daily during periods of favorable weather, we set a date for the “beginning of migration” each spring (see below). We made the operational assumption that all birds disappearing from the study site after that date had departed for migration. The research presented here was described in animal use protocol 9850 approved January 4, 1999 by the IACUC of the J. Exp. Zool.
588 College of William and Mary. Birds were captured and color banded under a USGS Bird Banding Lab permit, and held in accordance with all applicable Virginia and USA permits. Study 1: Unmanipulated Timing of Molt and Migration (1998) To determine when each bird molted we quantified molt every time an individual was captured. Beginning on March 1, molt of the head feathers was scored as follows. The head was divided into eight areas (left and right lores combined, left and right superciliary stripes combined, left and right lateral stripes combined, median stripe, left and right auriculars combined, nape, chin, left and right eye margins combined). Each area was scored on a scale of 0–3 where 0 ¼ no incoming sheathed (hereafter “pin”) feathers, 1 ¼1–10 pin feathers, 2 ¼ 11–50 pin feathers, 3 ¼ more than 50 pin feathers. A score was calculated for each bird on each recapture by summing the scores for all areas. Birds typically progressed from 0 through a maximum recorded score of 24 (more than 50 pin feathers in each of the eight regions), and then declined as pin feathers erupted from their sheaths and were no longer counted as such. Molt can vary in terms of date of onset and/or rate of progression and is notoriously difficult to quantify because it has a long duration, progresses differently among individuals, and varies across different parts of the body. Because the number and date of recaptures varied, we could not track the progress of each bird's molt at a fine scale. Instead, we used two indices that could be calculated even for birds with very different recapture schedules. One index was the estimated date of molt onset. We back‐calculated to estimate date of onset using a typical trajectory based on data from a group of birds for which we had excellent data on onset. To produce this plot of typical onset we used data from captive birds held under natural photoperiod and examined weekly during the onset of molt in 1999. Ear and chin data were used because these were the first areas to molt in every captive bird examined in this set. Using only data from ear and chin molts we developed a plot of the changes in molt scores over 7‐day periods. We could then back‐calculate the date of molt onset for any free‐living bird that was captured with ear or chin molt before its date of peak molt. Our other index was date of peak molt, that is, the estimated date on which the bird was molting the most feathers on its eight head regions combined. First, we used the nine free‐living birds for which we had the most recaptures to plot a typical molt progression. We then divided the molt period into 5‐day intervals and plotted the mean molt scores against time since onset of molt (see above) for our nine model birds. Using this plot we could then estimate the date of peak molt score for any bird captured two or more times during the increasing phase of molt by placing it on the typical trajectory. The two indices of molt were both intended only to distinguish early birds from late birds, and it should be noted that they were calculated based on entirely different data using independent sets of birds. J. Exp. Zool.
CRISTOL ET AL. Studies 2 and 3: Manipulations of Molt After examining the naturally occurring relationship between molt and disappearance from the wintering site in free‐living birds during 1998, we experimentally manipulated molt in temporarily captive birds the following winter/spring and examined whether birds with advanced molt also had advanced spring disappearance dates. In 1999 we advanced molt by advancing photoperiod for wild‐caught sparrows temporarily housed in an aviary and given a high‐quality diet. Because we observed that temporarily captive control birds, maintained on high‐quality diet but natural photoperiod, also advanced molt somewhat compared to free‐living birds, the following year we used strictly diet manipulation to speed or slow molt. We used the same molt indices described above for free‐living birds, but because birds were experimentally manipulated we calculated treatment‐specific molt trajectories from the most complete birds in each of our treatment groups and estimated scores for the remaining birds using these trajectories. It should be noted that sample sizes for onset and peak of molt, as well as date of disappearance, are smaller than total sample sizes because, even for captive birds, we were sometimes unable to calculate these dates if we did not have sufficient recaptures after release. Calculating molt onset required a reliable estimate of peak molt date (except for some frequently‐caught free‐living birds in which it could be documented at a finer scale). If subjects had not molted enough before release, and we did not recapture them enough times to estimate peak, then we also could not estimate onset. Study 2: Photoperiodic Manipulation of Molt (1999) In 1999, we tested whether molt date constrained departure date by providing some white‐throated sparrows with an advanced photoperiod and a high‐quality diet, which we expected to cause them to molt early. All adult birds caught before April 1 with wings 70 mm were assumed to be males and were assigned to one of three treatment groups: advanced, captive control, or free‐ living. We established two advanced and two control flocks (n ¼ 21 birds per flock) in four indoor (but unheated), light‐proof rooms (4.5 4.3 2.7 m). After one week of acclimatization under simulated natural day length, we began increasing the day length in the advanced rooms by 5 min/day to induce early molt. The photoperiod in control rooms continued to track natural changes in day length. Our third treatment was 42 free‐living birds released at their original sites of capture after color‐banding and measuring. It should be noted that 50% of our birds had wings 74 mm and were thus unquestionably males, but by including in our experimental study birds with wings as small as 70 mm, some (5 unless shown in parentheses.
early also disappeared earlier from the wintering site. This suggests the possibility that molt is a constraint on departure and must be at least partly completed before birds can migrate to breeding grounds. However, causality cannot be inferred from the existence of a relationship between molt and disappearance dates. The same individuals may have the ability to both molt and depart early, for example dominant birds or those in better nutritional condition. To explore whether variables that influenced molt date also influenced disappearance date, we did two experimental tests of the relationship. We first induced an early molt onset and peak in temporarily captive birds by advancing photoperiod in an indoor aviary. These advanced birds were then released back onto the study site several weeks before migration, along with controls held under identical conditions but on natural photoperiod. Advanced birds disappeared from the wintering site more than 2 weeks before the captive controls, which left no earlier than free‐living birds. We were surprised to find that the captive controls molted 3 weeks earlier than their free‐living counterparts. We concluded that their captive diet had likely played a role in advancing molt, and we used this finding to design our next experimental manipulation of molt. Because the initial photoperiod manipulation also caused experimental birds to increase their singing behavior early (unpublished data), we assume that a cascade of physiological changes was triggered by the advanced photoperiod, and that some factor(s) other than early molt, such as endocrine fluxes, could have contributed to the early migration. Because the endogenous control of prealternate molt is poorly understood, we instead repeated the experimental manipulation of molt date
in an outdoor aviary under natural photoperiod using another exogenous cue, alteration of diet, to shift molt date. One group of birds received an ad libitum diet high in calories and protein that advanced the onset of molt by approximately one week relative to free‐living birds, while another received an insufficient diet that delayed the peak of molt (although onset was not delayed significantly relative to free‐living birds). Birds in the advanced treatment group disappeared from the wintering site approximately 5 days earlier than free‐living or delayed birds, which did not differ from one another. Our results are interesting for two reasons. First, these results suggest that diet can affect molt with a cascading effect on migration timing, such that birds induced to molt a week early apparently departed almost one week early for spring migration. This is consistent with the recent demonstration that free‐living swamp sparrows provisioned with food underwent early molt (Danner et al., 2014). Together these findings suggest a tight link between food availability and timing of molt and migration. Second, the ease with which we were able to manipulate molt timing contrasts with some previous studies (e.g., Murphy, '96) and raises questions about the exogenous regulation of spring and autumn molt; for example prebasic molt in bluethroats (Luscinia svecica) was not altered by photoperiod or captivity (Lindström et al., '94) whereas molt and timing of autumn migration behavior both responded to experimental change in photoperiod in hand‐reared juvenile blackcaps (Sylvia atricapilla, Pulido and Coppack, 2004). The pre‐eminence of photoperiod as an exogenous cue for molt may vary between spring and fall, or the role of food availability may have been overlooked in both seasons. J. Exp. Zool.
592 Because we manipulated molt by changing additional variables (photoperiod or diet), we still cannot conclusively point to molt as the causative agent for estimated departure date. Other explanations must be considered. For example, the early molting individuals might have migrated early in 1998 not because they had molted first, but because they were dominant, and dominant white‐throated sparrows may have had attributes that made them both molt and migrate early. In the 1999 experiment the photoperiodically advanced birds might have left the wintering site early not because they were done molting, but because premature environmental signals had induced migratory readiness in addition to an early molt. Finally, birds that were fed rich diets in 2000 might have migrated early not because of molt, but because they were in better nutritional condition. Only an experiment in which birds were forcibly molted by feather removal could establish causality with certainty, but it is not clear whether such an experiment is possible. Molt involves the pushing out of old feathers by emerging feathers, a process that would not be mimicked through experimental plucking. Additionally, caution must be used when interpreting the relationship between molt and migration timing because our estimates of molt onset and peak were based on an imperfect series of measurements of molt progression. We used data from sets of birds for which we had numerous samples to produce “typical” curves for estimating onset and peak, but the finding that onset and peak responded independently to manipulations and did not both correlate with timing of disappearance from the study site suggests that individual birds progressed differently from onset to peak. Two unexpected results require further explanation. In the 1999 experimental manipulation of molt using photoperiod, the captive control group molted somewhat earlier than their free‐ living counterparts, although their molt was not nearly as rapid as that of the advanced birds. This was most likely the result of their superior diet relative to free‐living birds or the reduced energetic demands of indoor life (although normal outdoor temperatures were maintained), but may also have occurred because these captive controls heard a greater amount of singing by the experimental birds housed in the same building. Interestingly, when these captive controls were released, they did not migrate earlier than their free‐living counterparts, despite having begun molt early. Our tentative explanation for their failure to migrate early stems from the fact that these birds had not reached peak molt before they were released. Their progression through molt may have been slowed once they were released and faced the same nutritional constraints as the free‐living birds. The experimental photoperiodically advanced birds, on the other hand, were mostly past peak molt when they were released, so could have migrated without further delay. The other unexpected finding was that in the 2000 dietary manipulation of molt, which included a delayed as well as an advanced treatment, the birds molting late were able to migrate at J. Exp. Zool.
CRISTOL ET AL. the same time as their free‐living counterparts, despite the fact that their molts had progressed less far on the release date. Two explanations deserve consideration. First, it may be that molting early facilitates early migration, but experimentally delaying molt does not delay migration because at some point birds will migrate regardless of molt status. If this were the case, molt would not be an absolute constraint on departure date, but might still explain variation in timing early in spring. Alternately, the effect of diet may be swamped by the effect of accelerating increases in day length that occur later in spring. Another explanation would be that delayed birds were able to rapidly catch up to the molt status of their free‐living counterparts just before departure because they were no longer eating a poor diet, as occurred when malnourished white‐crowned sparrows (Zonotrichia leucophrys) were returned to normal diets (Murphy et al., '88). It may be that late in the season there is so much high‐quality food available that all birds can molt rapidly. A post hoc analysis of molt rate supports this explanation, because rate of molt increased dramatically after release in these delayed birds (D. Cristol, unpublished data). Again, perhaps only earlier in the spring, when protein‐rich food may be limiting, would small differences in molt schedules lead to earlier migration. In conclusion, we suggest that molt and departure for spring migration are related in a wintering songbird, just as they are during autumn migration (Stutchbury et al., 2011). These important life‐history events are connected directly, or through a common response to environmental cues. Further experiments will be necessary before we can conclude that molt is a primary constraint on timing of migration, but this is a parsimonious explanation of our results: molt and departure were correlated in free‐living birds and two different methods of advancing molt accelerated apparent departure. One possible mechanism for the relationship between molt and migration timing would be the following scenario: (1) socially dominant individuals obtain more of the rare high‐protein food resources on the winter range; (2) dominants are thus in better condition which allows them to molt earlier; (3) earlier‐molting, socially dominant birds migrate earlier to breeding grounds. There is ample support for a link between social status and access to resources in white‐throated sparrows (Schneider, '84; Piper and Wiley, '90). A study on our sparrow population indicated that degree of despotism in captive flocks affected rate of replacement feather growth (Jenkins et al., 2001). The present study establishes a link between timing of molt and migration for this species. Thus, there is some evidence for all aspects of a proposed mechanistic link between dominance status, molt, and migration. This is the same type of relationship proposed to explain why warblers from better tropical habitats arrived early on temperate wintering grounds. Based on stable isotope profiles, Marra et al. ('98) determined that early spring migrant male American redstarts (Setophaga ruticilla) were more likely to have wintered in higher quality mangrove habitat, whereas late male and female
MOLT TIMING CONSTRAINS BIRD MIGRATION redstarts came from lower‐quality scrub forest. Habitat segregation on the wintering grounds was influenced by dominance status and affected nutritional condition and spring migration date, so there is a case for the importance of wintering ground conditions in determining breeding success (Marra, 2000; Marra and Holmes, 2001). In another songbird, the blackcap, it has also been suggested that only the fittest males are capable of an early spring departure (Izhaki and Maitav, '98). Our results suggest that in species with a significant prealternate molt, the ability to complete feather replacement early may be one mechanism for the relationship between dominance, habitat quality/nutritional status, and migration date. Previous studies have shown that status or condition were related to speed or timing of molt and reproductive success (e.g., Mulder and Magrath, '94; Robertson et al., '98), but ours is the first study to suggest by experimental manipulation that molt and timing of migration are tightly linked, with implications for whether and how much white‐ throated sparrow populations will be able to adapt migration timing to expected earlier onset of spring.
ACKNOWLEDGMENTS We are greatly indebted to Jessica Armstrong, Brandi Jenks Brown, Dave Cerasale, Elise Donnelly, Chris Farabaugh, Dee Fritz, Chris Gleason, Josh LeClerc, Katie Martin, Jen McCall, Andrew McCormick, Katie Murphy, Hagai Nassau, Erica Reynolds, and Zakiya Thomas for spending many hours in the blind confirming the continued presence of each bird daily. Walter Piper and Thomas Grubb, Jr. made helpful suggestions on an earlier version of the manuscript. This study was supported by NSF grant IBN‐ 9876108 as well as the Virginia Society of Ornithology, the Virginia Academy of Science, the Williamsburg Bird Club, and the College of William & Mary.
LITERATURE CITED Bêty J, Giroux J‐F, Gauthier G. 2004. Individual variation in timing of migration: Causes and reproductive consequences in greater snow geese (Anser caerulescens atlanticus). Behav Ecol Sociobiol 57:1–8. Cristol DA, Baker MB, Carbone C. 1999. Differential migration revisited: Latitudinal segregation by age and sex class. Curr Ornith 15:33–88. Danner RM, Greenberg RS, Danner JE, Walters JR. 2014. Winter food limits timing of moult in a short‐distance migratory bird. Funct Ecol Available online ahead of print: DOI: 10.1111/1365–2435. 12322. Dugger BD. 1997. Factors influencing the onset of spring migration in Mallards. J Field Ornith 68:331–337. J.B. Falls, J.G. Kopachena.2010. White‐throated Sparrow (Zonotrichia albicollis), The Birds of North America Online. A. Poole. Ithaca: Cornell Lab of Ornithology Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/12DOI:10. 2173/bna.128 .
593 Gordo O. 2007. Why are bird migration dates shifting? A review of weather and climate effects on avian migration phenology. Clim Res 35:37–58. Harrison XA, Blount JD, Inger R, Norris DR, Bearhop S. 2011. Carry‐ over effects as drivers of fitness differences in animals. J Anim Ecol 80:4–18. Hemborg C, Lundberg A. 1998. Costs of overlapping reproduction and moult in passerine birds: An experiment with the pied flycatcher. Behav Ecol Sociobiol 43:19–23. Izhaki I, Maitav A. 1998. Blackcaps Sylvia atricapilla stopping over at the desert edge: Inter‐ and intra‐sexual differences in spring and autumn migration. Ibis 140:234–243. Jenkins KD, Hawley DM, Farabaugh CS, Cristol DA. 2001. Ptilochronology reveals differences in condition of captive white‐throated sparrows. Condor 103:579–586. Kuenzel WJ, Helms CW. 1974. An annual cycle study of tan‐striped and white‐striped white‐ throated sparrows. Auk 91:44–53. Lindström A, Daan S, Visser GH. 1994. The conflict between moult and migratory fat deposition: a photoperiodic experiment with bluethroats. Anim Behav 48:1173–1181. Marra PP. 2000. The role of behavioral dominance in structuring habitat occupancy of a migrant bird during the nonbreeding season. Behav Ecol 11:299–308. Marra PP, Hobson KA, Holmes RT. 1998. Linking winter and summer events in a migratory bird by using stable carbon isotopes. Science 282:1884–1886. Marra PP, Holmes RT. 2001. Consequences of dominance‐mediated habitat segregation in American redstarts during the nonbreeding season. Auk 118:92–104. Mulder RA, Magrath MJL. 1994. Timing of prenuptual molt as a sexually selected indicator of male quality in superb fairy‐wrens (Malurus cyaneus). Anim Behav 5:393–400. Murphy ME. 1996. Energetics and nutrition of molt. Carey C, Avian Energetics and Nutritional Ecology. New York: Chapman and Hall. pp 158–198. Murphy ME, King JR, Lu J. 1988. Malnutrition during the postnuptial molt of white‐ crowned sparrows: Feather growth and quality. Can J Zool 66:1403–1413. Norris DR, Marra PP, Kyser TK, Sherry TW, Ratcliffe LM. 2004. Tropical winter habitat limits reproductive success on the temperate breeding grounds in a migratory bird. Proc R Soc Lond B 271:59–64. Piper WH, Wiley RH. 1990. The relationship between social dominance, subcutaneous fat, and annual survival in wintering white‐throated sparrows (Zonotrichia albicollis). Behav Ecol Sociobiol 26:201–208. Piper WH, Wiley RH. 1991. Effects of laparotomies on wintering white‐throated sparrows and the usefulness of wing chord as a criterion for sexing. J Field Ornith 62:40–45. Pulido F. 2007. Phenotypic changes in spring arrival: Evolution, phenotypic plasticity, effects of weather and condition. Clim Res 35:5–23. J. Exp. Zool.
594 Pulido F, Coppack T. 2004. Correlation between timing of juvenile moult and onset of migration in the blackcap, Sylvia atricapilla. Anim Behav 68:167–173. Robertson GJ, Cooke F, Goudie RI, Boyd WS. 1998. Molt speed predicts pairing success in male harlequin ducks. Anim Behav 55:1677–1684. Rockwell SM. 2013. Carry-over effects from the non-breeding season influence spring arrival dates, reproductive success, and survival in an endangered migratory bird, the Kirtland's warbler (Setophaga kirtlandii). [Dissertation]. College Park, MD: University of Maryland http://hdl.handle.net/1903/14363. Schaub M, Jenni L, Bairlein F. 2008. Fuel stores, fuel accumulation, and the decision to depart from a migration stopover site. Behav Ecol 19:657–666. Schneider KJ. 1984. Dominance, predation, and optimal foraging in white‐throated sparrow flocks. Ecology 65:1820–1827.
J. Exp. Zool.
CRISTOL ET AL. Siegel S, Castellan NJ, Jr. 1988. Nonparametric Statistics for the Behavioural Sciences. 2nd edition. New York: McGraw‐Hill. Stanley C, MacPherson M, Fraser KC, McKinnon EA, Stutchbury BJM. 2012. Repeat tracking of individual songbirds reveals consistent migration timing but flexibility in route. PLoS ONE 7:-e40688.DOI: 10.1371/journal.pone.0040688. Stutchbury BJM, Gow EA, Done T. 2011. Effects of post‐ breeding moult and energetic condition on timing of songbird migration into the tropics. Proc R Soc Lond B 178:131–137. Tökölyi J, Bókony V, Barta Z. 2008. Seasonal colour change by moult or by the abrasion of feather tips: A comparative study. Biol J. Linn Soc 94:711–721. Wingfield JC, Schwabl H, Mattocks PW, Jr. 1990. Endocrine mechanisms of bird migration. Gwinner E, Bird Migration. Berlin, Germany: Springer‐Verlag. pp 232–256.
