Ecology, 78(8), 1997, pp. 2524-2531 ? 1997 by the Ecological Society of America GEOGRAPHIC CLINES OF AGE RATIOS OF BLACK-THROATED BLUE WARBLERS (DENDROICA CAERULESCENS) GARY R. GRAVES1 Department of Vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Waj&mgfoM, DC. 20J60 (/&4 Abstract. I investigated the age structure of breeding populations of Black-throated Blue Warblers {Dendroica caerulescens) in eastern North America. The proportion of year- ling males increases as relative abundance decreases toward the margins of the breeding range. Geographic dines in age structure are hypothesized to correlate with gradients in habitat quality radiating from the Appalachian Mountains, the probable axis of source populations. Data from the southern Appalachians suggest that despotism of older males may impose a ceiling on the proportion of yearlings in saturated high-quality habitats, regardless of local productivity. Populations that exhibit low variance in yearling recruit- ment and high population densities may fall toward the "source" end of the "source-sink" continuum, and vice versa. It is hypothesized that the regional distribution of source and sink populations of the Black-throated Blue Warbler might closely mirror the prevalence of successful multiple brooding. Key words: age ratio; Appalachian Mountains; Black-throated Blue Warbler; Dendroica caeru- lescens; despotism; dispersal; geographic dines; multiple brooding; source-sink dynamics. INTRODUCTION Long-term studies of avian populations have yielded significant insights as to why species thrive in some locations but not in others, or disperse from one habitat patch to the next (e.g., Kluijver 1951). Extrapolation of findings from local to continental scales has been less successful (see Wiens 1981, 1989). The advent of continent-wide breeding bird censuses (e.g., Robbins et al. 1986) has encouraged large-scale mapping (Price et al. 1995) and analyses (e.g., Bystrak 1981, Henge- veld and Haeck 1982, Brown 1984, Maurer and Brown 1989, Maurer 1994, Brown et al. 1995, Villard and Maurer 1996) that have verified and amplified biogeo- graphic hypotheses in circulation since early in this century (e.g., Grinnell 1917). These studies have re- vealed that significant density gradients occur within the breeding ranges of avian species. Continental spe- cies are relatively abundant only in a few locations within their geographic ranges, and peaks of abundance frequently occur near the centers of geographic ranges, implying that ecological conditions are less favorable near range margins where population densities ap- proach zero. Abundance gradients ultimately reflect habitat se- lection on a local and regional scale. Many data suggest that habitat selection in birds is influenced by intra- speciflc density (e.g., Mayr 1926, Kluijver 1951, Manuscript received 19 July 1996; revised 9 December 1996; accepted 9 January 1997; final version received 3 Feb- ruary 1997. 1 Present address: Division of Birds, National Museum of Natural History, Smithsonian Institution, Washington, DC. 20560. Brown 1969, Orians 1969, Fretwell and Lucas 1970, Krebs 1971, Newton 1992). For avian species with des- potic habitat selection, yearlings are often forced to settle in less favorable habitat (see Sutherland and Par- ker 1985, Parker and Sutherland 1986). Intrinsic natal dispersal, in combination with density-dependent ter- ritorially, may cause young or inexperienced individ- uals to accumulate in low-quality "sink" habitats (see Lidicker 1975, Van Home 1983). Clines in age struc- ture may thus occur along spatial gradients that reflect habitat quality. Although a number of studies have demonstrated despotic distributions within local pop- ulations (e.g., Goss-Custard et al. 1982), age-structured habitat selection at large spatial scales has not been well established, primarily because the pertinent de- mographic data (e.g., unbiased population samples) are nearly impossible to obtain. Consequently, we still do not know whether despotic and ideal free processes of habitat selection observed in local populations (Suth- erland and Parker 1985, Parker and Sutherland 1986) can result in geographic clines in age structure on a continental scale. Geographic clines in age structure, should they occur, would have important applied and theoretical consequences for the conservation and man- agement of populations. I offer evidence that geographic clines in age struc- ture are present among breeding populations of a small migratory passerine, the Black-throated Blue Warbler, Dendroica caerulescens, in eastern North America. Analysis of geographically dispersed population sam- ples is used to estimate the relationship between age structure, relative abundance, and the geographic po- sition of populations within the breeding range. 2524 December 1997 CLINES IN WARBLER AGE RATIOS 2525 - - - - V" --' /'" ---1 s /__ ^\ A LiaL FIG. 1. Distribution of collecting sites (see Table 1) within the core breeding range (stippled) of the Black-throated Blue Warbler (Dendroica caerulescens) in eastern North America. Hatched lines represent state and provincial boundaries. METHODS The focal species Black-throated Blue Warblers (Fringillidae, Tribe Parulini; taxonomy of Sibley and Monroe 1990) breed in cool deciduous and mixed deciduous-coniferous for- ests in eastern North America, and winter in the Ca- ribbean basin (Holmes 1994, Price et al. 1995; Fig. 1). Males returning from their wintering grounds establish territories in mixed deciduous-coniferous or deciduous forests as early as the last week of April at the southern terminus of their breeding range in Georgia, and by the last week of May near the northern margins of the range in Michigan, south-central Ontario, and the Maritime Provinces of Canada (Fig. 1). Older males precede yearling males and females in spring migration by a few days to a week or more (Hubbard 1965, Holmes 1994). Detailed information on the breeding ecology of the species can be found in Holmes et al. (1992, 1996), Rodenhouse and Holmes (1992), and Holmes (1994). The distribution and abundance of breeding Black- throated Blue Warblers (especially south of 43? N) are strongly influenced by elevation. In the southern Ap- palachian Mountains where it is one of the most abun- dant breeding birds (Wilcove 1988), most males estab- lish territories above 750 m (above sea level). The ele- vational width of the breeding zone varies from 300 to 500 m on spurs and outliers of the southern Appala- chians, to >1000 m in the Great Smoky and Black mountains (35-36? N) of Tennessee and North Caro- lina. Population samples Population samples of breeding males were collected over the course of eight field seasons (1986-1994). The research protocol required discrete population samples, meaning that sampling was limited to areas within the breeding range where Black-throated Blue Warblers were fairly common to abundant, and where a quota of territorial males could be collected quickly in a cir- cumscribed area without significantly diminishing local populations. Dozens of locations near the northern and western margins of the warbler's breeding range were censused and rejected as potential collecting sites (e.g., Minnesota, Ontario, Nova Scotia) because population density was too low. Population samples were obtained from the smallest area possible in large tracts of forest on public lands, usually along trails or old logging roads. Fragmented forest plots were avoided, as were sites near human habitation or agriculture. In the Ap- palachians, sampling was conducted well above the species' lower elevational limit. Sampling was oppor- tunistic, in that males were collected as they were en- countered. No attempt was made to age specimens in the field. Observation of males was necessarily brief, and I could not rule out the possibility that some were nonterritorial or that yearlings were more vocal than older males. However, because the collecting protocol was consistent at all of the collecting sites (i.e., all 2526 GARY R. GRAVES Ecology, Vol. 78, No. 8 TABLE 1. Population samples of male Black-throated Blue Warblers {Dendroica caerulescens) collected at breeding localities. Locality Older Relative State or province Date coordinates^ Yearling male abundance^ 1) Georgia 10-12 June 1989 34.75 84.03 6 14 3.64 2) South Carolina 5-8 June 1989 35.00 83.10 6 12 2.57 3) North Carolina 17-18 June 1989 35.38 82.67 8 17 3.85 4) Tennessee 13-15 June 1989 35.45 84.08 6 11 5.67 5) North Carolina 25-27 June 1989 35.71 82.18 4 16 5.00 6) Tennessee 19-20 June 1989 36.13 82.33 5 15 2.08 7) Tennessee 22-23 June 1989 36.20 82.13 3 12 5.00 8) Virginia 9-10 June 1988 36.72 81.58 3 16 2.53 9) Virginia 11-12 June 1988 36.88 82.58 4 10 5.80 10) Virginia 12-13 June 1988 37.46 80.52 6 14 2.50 11) West Virginia 6-7 July 1986 21 June 1988 38.31 80.50 3 18 2.53 12) West Virginia 19 June 1988 38.68 79.52 8 10 4.00 13) West Virginia 24-25 June 1988 39.13 79.30 6 14 4.30 12-13 June 1993 9 16 4.30 14) Pennsylvania 9-11 June 1994 41.20 76.23 12 20 2.28 15) New York 18-20 June 1992 41.92 74.50 11 14 1.28 16) New York 23-25 June 1992 43.70 74.72 14 11 2.45 17) Vermont 21-22 June 1992 43.98 73.00 16 9 2.78 18) New Hampshire 27-28 June 1991 44.00 71.32 8 17 3.96 19) Ontario 22-24 June 1990 45.37 79.58 14 21 0.88 20) Maine 23-25 June 1991 45.48 69.52 15 7 0.96 21) Ontario 25-29 June 1990 45.67 77.50 20 20 1.68 22) New Brunswick 14-17 June 1991 45.83 67.00 11 13 0.74 23) Michigan 12-16 June 1990 46.17 88.92 9 11 1.04 t Coordinates, presented as north latitude . . . west longitude, have been converted to decimals for computational purposes. % Total number of specimens collected in a locality divided by the greatest distance between collecting sites within a locality (males per kilometer). observation and collecting was performed by the au- thor), geographical bias owing to differential detect- ability of yearlings and older males was minimized. The low elevation of the central Appalachians (sel- dom >800 m) greatly restricts the breeding distribution of Black-throated Blue Warblers from western Mary- land north to the South Branch of the Susquehanna River in north-central Pennsylvania (see Brauning 1992). Most males that occur in this region settle be- tween 500 m and the edge of the plateau (600?800 m) in a narrow elevational zone along the Allegheny front. The elevational zone of breeding populations is thus substantially narrower, and some 150 to 250 m lower in this region than in the southern Appalachians. Populations north of Pennsylvania are more widely and continuously distributed, and warblers may breed near sea level in Maine, Nova Scotia, and New Brunswick (e.g., Erskine 1992). Age-related plumage characteristics Two age classes of male Black-throated Blue War- blers can be distinguished by plumage characters in breeding specimens (see Dwight 1900, Holmes 1994, Graves, in press): (1) yearlings (first alternate plumage, SY in banding terminology) in their first potential breeding season, and (2) older individuals (definitive alternate plumage, ASY in banding terminology) in their second or later breeding season. The definitive alternate plumage of older males is completed on the wintering ground after the first potential breeding sea- son. Age-related plumage differences are subtle, and unlikely to result in collecting bias. Neither females nor fledglings were collected. To ensure that neither migrating nor wandering postbreeding birds were col- lected, I sampled and censused populations during the peak of breeding season, 5 June-7 July (obtaining all but one sample before 30 June). The brief breeding season and the large size of the breeding range (2000 km from east to west and north to south) precluded a geographically stratified sampling design each year. Instead, I collected population sam- ples within a limited region each field season, begin- ning in the central Appalachians in the first year, and in subsequent years collecting to the south, northwest, northeast, north-central, etc., until all regions of the breeding range had been sampled. Replicate samples at Locality 13 (Table 1) were obtained 5 yr apart. Relative abundance Geographic variation in relative abundance among sampled populations cannot be reliably extrapolated from Breeding Bird Survey (BBS) census data because the Appalachians are insufficiently sampled by BBS routes. Several of my population samples were col- lected in areas that exhibit near-zero BBS density es- timates (see Holmes 1994, Price et al. 1995). I esti- mated the relative abundance of local populations as the total number of specimens collected in a locality CLINES IN WARBLER AGE RATIOS 50? 2527 FIG. 2. Relative abundance (A) and percentage of yearling males (B) in breeding population samples of Black-throated Blue Warblers in eastern North America (see Table 1, Fig 1). Diameter of locality circles is proportional to the relative abundance (A) and the percentage of yearlings (B). divided by the greatest distance between collecting sites within the sampling locality (males per kilometer) (Table 1). When relative abundance was high, the sam- pling quota was quickly obtained over a relatively short distance, and vice versa. Data analysis The relationship between yearling percentage, rela- tive abundance, and latitude (for population samples) was examined with Spearman rank correlation coeffi- cients (Rs, two-tailed). The effects of collection year on age structure (arcsine transformed percentage of yearlings) were examined with one-way analysis of c q o o. o 4 # i 0.6 # 0.5 ??? # # - 04 0.3 # # # # # # m m # # 0.2 n 1 # I I i # i 0 12 3 4 5 6 Relative Abundance FIG. 3. Bivariate relationship of relative abundance and proportion of yearling males in population samples of Black- throated Blue Warblers. variance (ANOVA). To clarify the spatial and temporal correlates of age structure, I used a regression model that incorporated the percentage of yearling males as the dependent variable, and "year" (categorical) and latitude as independent variables. Chi-square tests of independence were performed on 2 X 2 contingency tables to test for differences between population sam- ples in the proportion of yearlings and older males. All statistical tests were performed on SYSTAT (Wilkinson 1989). RESULTS Relative abundance exhibited a sevenfold difference across the breeding range (Table 1), and was negatively correlated with latitude (Rs = -0.62, P < 0.005) (Fig. 2). The highest relative abundance of Black-throated Blue Warblers was recorded in the Appalachian Moun- tains, especially south of 40? N. The percentage of yearling males in population samples was positively correlated with latitude (Rs = 0.70, P < 0.001; Fig. 2) and negatively correlated with relative abundance (Rs = -0.51, P < 0.02; Fig. 3). The proportion of yearling male Black-throated Blue Warblers in population sam- ples appears to increase clinally from the Appalachian axis (from Georgia to New Hampshire) outward to the northern and northwestern margins of the species' breeding range. Moreover, the percentage of yearlings was inversely proportional to local relative abundance. The estimated proportions of yearling males in pop- ulations exhibited considerable variation among years (annual averages: 1988, 27%; 1989, 28%; 1990, 45%; 1991, 48%; 1992, 55%; 1993, 36%; 1994, 38%; see Table 1). Year effects were significant (ANOVA, F617, P = 0.009), but much of the variation can be explained by latitude and constraints of the sampling protocol. 2528 GARY R. GRAVES Ecology, Vol. 78, No. 8 TABLE 2. Effects of latitude and year on the percentage of yearling males in population samples of Black-throated Blue Warblers (see Table 1). Source df ss MS F ratio P Latitude Yeart Error 1 6 16 0.035 0.080 0.186 0.035 0.013 0.012 3.016 1.151 0.102 0.379 f Entered as a categorical variable. Sampling localities spanned 11.4? of latitude, but the latitudinal span of population samples within years var- ied from 0.0? (single localities in 1993 and 1994) to 2.4? (1988). When year and latitude were entered into a multiple regression model, year effects were nonsig- nificant, whereas latitude explained a larger fraction of the continent*/ variation in age ratios (Table 2). The relationship between age structure and latitude is clearer when the population samples are partitioned into northern (>40? N) and southern (<40? N) groups. The percentage of yearling males in the northern group (collected 1990, 1991, 1992, and 1994) averaged 47.6%, vs. 28.3% for the southern group (collected in 1988, 1989, and 1993). This difference was significant (X2 = 21.6, df = 1, P < 0.001). The percentage of yearlings in population samples was uncorrelated with the date of collection (median day of year; Rs = 0.11, n = 24, P > 0.50). The percentages of yearling males from population samples taken 5 yr apart at the same site (Table 1, Locality 13) in West Virginia were not significantly different (X2 = 0.06, df = 1, P = 0.81). DISCUSSION Large-scale spatial variation in age structure of breeding bird populations has not been studied inten- sively (see Snow 1956, Greenberg 1980). Rohwer et al. (1980) examined the age structure of breeding American Redstarts (Setophaga ruticilla) represented in museum collections, and interpreted the significant difference in the proportion of yearlings collected east and west of the Great Plains as the result of a regional shift in molting chronology. Rohwer et al. (1983) later rejected this hypothesis, and Sherry and Holmes (1991) pointed out that a fixed age ratio was illusory unless replicated over large areas and long time periods. Nev- ertheless, geographic variation in age differences ob- served in American Redstarts warrants additional scru- tiny. The specimens examined by Snow (1956), Green- berg (1980), and Rohwer et al. (1980) were collected over a period of many decades, by many field collec- tors, and may have been systematically biased by fac- tors unknown to the investigators. In the present study, many of the potential biases exhibited in the accu- mulated series of specimens in museums have been minimized or eliminated because of the standardized sampling protocol. Hypotheses I considered four partially overlapping hypotheses that may account for the higher proportion of yearlings in peripheral populations: (1) males in peripheral regions of the breeding range suffer greater mortality, (2) peripheral populations exhibit higher productivity and local recruitment of yearlings, (3) observed dif- ferences are caused by sampling methods, and (4) pe- ripheral areas of the breeding range have a higher pro- portion of lower quality breeding sites that operate as population sinks for dispersing yearlings from source populations along the Appalachian axis. The first hypothesis does not seem likely as the prin- cipal cause of age and abundance dines in male Black- throated Blue Warblers. Assuming that habitat near the northern and western margins of the breeding range is relatively poor, males that settle there may suffer higher mortality than males in the Appalachians, although sur- vivorship data are lacking. However, the higher pro- portion of yearlings in peripheral populations would not be predicted under the differential mortality hy- pothesis, because mortality would presumably accrue to locally produced fledglings (hatching year, HY in banding terminology) as well as to breeding males. In fact, annual survivorship of adult passerines is signif- icantly higher than that of hatching-year birds. The second hypothesis is that the abundance of year- lings in local populations reflects the recruitment of young birds hatched at those locations the previous year (Nolan 1978, Sherry and Holmes 1991, 1992, Holmes et al. 1992). If yearling recruitment is directly linked to local productivity, the fledging rate in geo- graphically peripheral areas (e.g., Michigan, Ontario, Maine, New Brunswick) must be relatively high com- pared to populations from the Appalachian axis. How- ever, if productivity is elevated in peripheral areas, why is relative abundance so low? This hypothesis would require the unlikely scenario that peripheral popula- tions (but not core populations) were sampled during periods of rapid population growth, yielding an excess of yearlings. However, survivorship and productivity data for the sampled populations are unavailable. A third possibility is that the observed dines in age structure are an artifact of regional variation in the relative abundance and elevational distribution of the Black-throated Blue Warbler, and the effect of this vari- ation on sampling protocol. As mentioned earlier, sam- pling of populations in the Appalachians (as elsewhere) was conducted where relative abundance of breeding populations was high enough to permit sampling. I avoided collecting in peripheral habitats where abun- dance was low, especially near the elevational bound- aries of local breeding populations, where yearlings may be more common. If older males tend to aggregate near the center of the warbler's elevational distribution, yearlings will be underrepresented in population sam- ples restricted to the warbler's optimal elevational December 1997 CLINES IN WARBLER AGE RATIOS 2529 zone. However, significant correlations with latitude (percentage of yearlings, Rs = 0.53, P < 0.03) persist when analyses are limited to Appalachian localities (Table 1, Localities 1-18). In other words, although it is possible that sampling bias accounts for some of the regional differences in age ratios, there is little evi- dence that the nearly 20% difference in yearling per- centages between southern and northern populations can be explained by regional collecting bias (see Meth- ods). A fourth hypothesis is that the higher proportion of yearlings in peripheral areas (presumably in poorer quality habitat) is due to emigration from highly pro- ductive areas along the Appalachian axis. This hy- pothesis is related to "source-sink" theory (e.g., Wiens and Rotenberry 1981, Van Home 1983, Pulliam 1988, Howe et al. 1991, Pulliam and Danielson 1991, Rob- inson 1992, Donovan et al. 1995, Brawn and Robinson 1996). "Source" populations are usually defined as those that exhibit no net change in size over several generations, and are net exporters of individuals to oth- er populations. Conversely, "sink" populations are net importers of individuals. The genesis of habitat segregation on a local scale (e.g., elevational transects) begins in spring when older, socially dominant males arrive early and preempt the most desirable habitat (see Brown 1969, Fretwell and Lucas 1970, Holmes 1994, Holmes et al. 1996). Year- ling males establish territories where they can, in the interstices of high-quality habitat and in less optimal patches. I hypothesize that the age-related habitat seg- regation observed in local habitats (e.g., Holmes et al. 1996; G. R. Graves, unpublished data) corresponds to a source-sink spectrum that may be extrapolated to the entire breeding range of the species. Under this inter- pretation, source populations are distributed primarily along the higher Appalachians, whereas the northern and western frontiers of the breeding range function as population sinks in some or most years. Although direct evidence of long-distance dispersal in Black-throated Blue Warblers has not been obtained, data from other migratory passerines suggest that natal dispersal is leptokurtic, and that a small fraction of yearlings may settle hundreds of kilometers from their natal site (for references see Pitelka 1974, Greenwood and Harvey 1982, Moore and Dolbeer 1989, Payne 1991). Holmes (1994) reported that only 1 of 800+ nestling Black-throated Blue Warblers banded on the 70-ha Hubbard Brook study site was resighted in sub- sequent years (a female that nested 2.5 km from her natal site). The spatial distribution of suitable breeding habitat and the dispersion probability field (Hengeveld 1989) of natal dispersal (from natal area to the site of first reproduction) varies substantially in different parts of the Black-throated Blue Warbler's breeding range. Birds returning to their natal regions or territories in the Appalachians (<40? N) are constrained by elevation to a corridor of habitat many times longer in one di- mension than the other (see Graves 1988). For example, the 700-m contour interval outlines the potential array of settlement sites at the southern terminus of the spe- cies' breeding range. If successful long-distance dis- persal occurs, vectors must be predominately aligned northeastward. Potential natal dispersal of Black- throated Blue Warblers from northern breeding sites (43?30'-47? N) is more rotationally symmetric, mean- ing that birds may disperse in any direction and find breeding habitat. The radii of symmetric dispersion fields in much of New England, for example, may be as large as several hundred kilometers. The relatively uniform percentage of yearlings in population samples from the southern Appalachians suggests a compensatory relationship between the pro- portion of yearlings and older males in high-quality, saturated habitats. In essence, there may be a ceiling on the number of yearlings recruited into saturated, high-quality habitats, regardless of local productivity. As long as the pool of males is sufficiently large in adjacent suboptimal habitats, annual fluctuations in productivity in optimal habitats may well be masked by immigration (see Wilcove and Terborgh 1984)?i.e., the proportion of yearlings in saturated habitats may decline after poor reproductive years, but may never rise above the ceiling dictated by mortality of older males, even after a season of superior productivity. This situation may arise frequently along steep elevational gradients, where yearlings that settle in peripheral, low- quality territories in one year emigrate to adjacent high- quality habitat (as older males) in a subsequent year. In some instances, most regional reproduction could occur in areas that exhibit low yearling recruitment. Rather than reflecting fluctuations in local productivity, as suggested by Holmes et al. (1992), variance in an- nual recruitment of yearlings may serve to distinguish source from sink populations (see Winker et al. 1995). Those that exhibit low variance in yearling recruitment and high population densities might be considered to fall toward the source end of the source-sink contin- uum, and vice versa. In conclusion, settlement patterns of yearlings and older males in local habitats are likely to result from social dominance and the preemption of high-quality habitat by older males, although this remains to be tested. The question of geographic dines in age struc- ture is more difficult to address with the available cor- relative data. Subtle changes in sampling procedure associated with elevation may result in minor variation in observed age ratios, but as I have suggested, the bulk of the regional variance may be caused by the dispersal of yearlings from source populations (high relative abundance) along the Appalachian axis to pe- ripheral sink populations (low relative abundance). In any event, the demography of Appalachian and pe- ripheral populations seems to be significantly different. 2530 GARY R. GRAVES Ecology, Vol. 78, No. 8 Multiple brooding and source-sink dynamics Holmes et al. (1992) concluded that multiple brood- ing in the Black-throated Blue Warbler accounted for most of the annual variation of young fledged per fe- male. An average of 48% of females at Hubbard Brook initiated second clutches after successfully fledging a first brood. Successful multiple brooding coincided with warm, dry springs, low levels of predation, and an abundance of caterpillars (Holmes et al. 1992, Ro- denhouse and Holmes 1992). Egg-laying was delayed by cool, wet weather. Late frosts affected leafing phe- nology and the appearance of phytophagous insects eat- en by warblers. The percentage of clutches in which one or more young died of starvation was negatively correlated with caterpillar abundance and biomass. Growth rates and body masses of nestlings were neg- atively correlated with precipitation, and positively correlated with daily minimum temperature. The Hubbard Brook data suggest that the reproduc- tive potential of Black-throated Blue Warblers is linked to the length of the growing season (the period without killing frost). Based on data presented by Holmes et al. (1992), Holmes (1994), and Nolan (1978), I esti- mated the nesting cycle of the Black-throated Blue Warbler (initiation of nest building to independence of young) to be a minimum of 53 d. Key in this calculation was Nolan's (1978) conclusion that fledgling Prairie Warblers (Dendroica discolor) <33 d old (from hatch- ing day) could not survive without parental assistance. This suggests that successful double brooding in the Black-throated Blue Warbler may not be possible in regions with growing seasons of less than ?95-100 d, allowing a week of overlap between first and second broods. Temperature gradients within the breeding range of the Black-throated Blue Warbler are correlated with latitude, elevation, and proximity to bodies of water, especially the Atlantic Ocean and Great Lakes. For example, the mean frost-free period at Hanover, New Hampshire, on the Connecticut River is 135d(17 May- 30 September), 118 d (24 May-19 September) at Woodstock on the Pemigewasset River some 60 km northeast of Hanover, and only 39 d (8 July-17 August) on Mount Washington, the highest peak in the White Mountains (Koss et al. 1988). In regions near the 100-d growing season isotherm, any nesting failure that oc- curs >2 wk after the first eggs are laid automatically halves the potential annual reproductive output of a female. These data suggest that on a continental scale, a rapid transition in per capita productivity may occur near the 100-d isotherm, and that the location of source and sink populations of the Black-throated Blue War- bler might correlate with the regional distribution of successful multiple-brooding. ACKNOWLEDGMENTS I am grateful to Bill Boecklen, Dan Brauning, Jeff Brawn, Nick Gotelli, Russell Greenberg, Richard Holmes, Tom Mar- tin, Brian Maurer, John Rappole, Van Remsen, Scott Rob- inson, Tom Sherry, David Wilcove, Kevin Winker, and an anonymous reviewer for comments, criticism, and advice. Phil Angle, Carla Dove, Jim Dean, Chris Milensky, and Brian Schmidt skillfully prepared the specimens. Field work was supported by the Alexander Wetmore Fund and by the Re- search Opportunities Fund of the Smithsonian Institution. I thank the following agencies for granting scientific collecting permits or assistance: U.S. Fish and Wildlife Service, Ca- nadian Wildlife Service (Atlantic and Ontario Region), U.S. Forest Service, Georgia Department of Natural Resources, Maine Department of Inland Fisheries and Wildlife, Maryland Department of Natural Resources, Michigan Department of Natural Resources, New Hampshire Fish and Game Depart- ment, New York Department of Environmental Conservation, North Carolina Wildlife Resources Commission, Pennsyl- vania Game Commission, South Carolina Wildlife and Marine Resources Department, Tennessee Wildlife Resources Agen- cy, Vermont Fish and Wildlife Department, Virginia Com- mission of Game and Inland Fisheries, West Virginia De- partment of Natural Resources, and Wisconsin Department of Natural Resources. LITERATURE CITED Brauning, D. W. 1992. Atlas of breeding birds of Pennsyl- vania. University of Pittsburgh Press, Pittsburgh, Pennsyl- vania, USA. Brawn, J. D., and S. K. Robinson. 1996. Source-sink pop- ulation dynamics may complicate the interpretation of long- term census data. Ecology 77:3-12. Brown, J. H. 1984. On the relationship between abundance and distribution of species. American Naturalist 124:255- 279. Brown, J. H., D. W. Mehlman, and G. C. Stevens. 1995. Spatial variance in abundance. Ecology 76:2028-2043. Brown, J. L. 1969. Territorial behavior and population reg- ulation in birds (a review and re-evaluation). Wilson Bul- letin 81:293-329. Bystrak, D. 1981. The North American breeding bird survey. Studies in Avian Biology 6:34-41. Donovan, T. M., R. H. Lamberson, A. Kimber, F. R. Thomp- son III, and J. Faaborg. 1995. Modeling the effects of habitat fragmentation on source and sink demography of neotropical migrant birds. Conservation Biology 9:1396- 1407. Dwight, J., Jr. 1900. The sequences of plumages and moults of the passerine birds of New York. Annals of the New York Academy of Sciences 13:73-360. Erskine, A. J. 1992. Atlas of breeding birds of the Maritime Provinces. Nimbus, Halifax, Nova Scotia, Canada. Fretwell, S. D., and H. L. Lucas, Jr. 1970. On territorial behavior and other factors influencing habitat distribution in birds. I. Theoretical development. Acta Biotheoretica 19: 16-36. Goss-Custard, J. D., S. E. A. Le V dit Durrell, S. McGrorty, and C. J. Reading. 1982. Use of mussel Mytilus edulis beds by oystercatchers Haemotopus ostralegus according to age and population size. Journal of Animal Ecology 51: 543-554. Graves, G. R. 1988. Linearity of geographic range and its possible effect on the population structure of Andean birds. Auk 105:47-52. . In press. Age determination of free-living male Black-throated Blue Warblers during the breeding season. Journal of Field Ornithology. Greenberg, R. 1980. Demographic aspects of long-distance migration. Pages 493-504 in A. Keast and E. S. Morton, editors. Migrant birds in the Neotropics: ecology, behavior, distribution, and conservation. Smithsonian Institution Press, Washington, DC, USA. Greenwood, P. J., and P. H. Harvey. 1982. The natal and December 1997 CLINES IN WARBLER AGE RATIOS 2531 breeding dispersal of birds. Annual Review of Ecology and Systematics 13:1-21. Grinnell, J. 1917. Field tests of theories concerning distri- butional control. American Naturalist 51:115-128. Hengeveld, R. 1989. Dynamics of biological invasions. Chapman and Hall, London, UK. Hengeveld, R., and J. Haeck. 1982. The distribution of abun- dance. I. Measurements. Journal of Biogeography 9:303- 316. Holmes, R. T 1994. Black-throated Blue Warbler {Dendroica caerulescens). In A. Poole and E Gill, editors. The birds of North America, Number 87. Academy of Natural Sci- ences, Philadelphia, Pennsylvania, and American Orni- thologists' Union, Washington, DC, USA. Holmes, R. T, P. P. Marra, and T W Sherry. 1996. Habitat- specific demography of breeding Black-throated Blue War- blers {Dendroica caerulescens): implications for population dynamics. Journal of Animal Ecology 65:183-195. Holmes, R. T, T W Sherry, P. P. Marra, and K. E. Petit. 1992. Multiple breeding and productivity of a Neotropical migrant, the Black-throated Blue Warbler {Dendroica cae- rulescens), in an unfragmented forest. Auk 109:321-333. Howe, R. W, G. J. Davis, and V. Mosca. 1991. The de- mographic significance of "sink" populations. Biological Conservation 57:239-255. Hubbard, J. P. 1965. Migration of the Black-throated Blue Warbler in southern Michigan. Jack-Pine Warbler 43:162- 163. Kluijver, H. N. 1951. The population ecology of the Great Tit, Parus m. major L. Ardea 39:1-135. Koss, J. W, J. R. Owenby, P. M. Steurer, and D. S. Ezell. 1988. Freeze/frost data. Climatology of the U.S. Number 20, Supplement Number 1. National Climatic Data Center, National Oceanic and Atmospheric Administration, Ashe- ville, North Carolina, USA. Krebs, J. R. 1971. Territory and breeding density in the Great Tit Parus major. Ecology 52:2-22. Li dicker, W. Z., Jr. 1975. The role of dispersal in the de- mography of small mammals. Pages 103-128 in F. B. Gol- ley, K. Petrusewicz, and L. Ryszkowski, editors. Small mammals: their productivity and population dynamics. Cambridge University Press, New York, New York, USA. Maurer, B. A. 1994. Geographical population analysis: tools for the analysis of biodiversity. Blackwell, Oxford, UK. Maurer, B. A., and J. H. Brown. 1989. Distributional con- sequences of spatial variation in local demographic pro- cesses. Annales Zoologici Fennici 26:121-131. Mayr, E. 1926. Die Ausbreitung des Girlitz {Serinus canaria serinus L.). Journal fur Ornithologie 74:571-671. Moore, W. S., and R. A. Dolbeer. 1989. The use of banding recovery data to estimate dispersal rates and gene flow in avian species: case studies in the Red-winged Blackbird and Common Crackle. Condor 91:242-253. Newton, I. 1992. Experiments on the limitation of bird num- bers by territorial behaviour. Biological Review 67:129- 173. Nolan, V., Jr. 1978. The ecology and behavior of the Prairie Warbler Dendroica discolor. Ornithological Monographs Number 26. Orians, G. H. 1969. On the evolution of mating systems in birds and mammals. American Naturalist 103:589-603. Parker, G. A., and W J. Sutherland. 1986. Ideal free distri- butions when individuals differ in competitive ability: phe- notype-limited ideal free models. Animal Behaviour 34: 1222-42. Payne, R. B. 1991. Natal dispersal and population structure in a migratory songbird, the indigo bunting. Evolution 45: 49-62. Pitelka, F. 1974. An avifaunal review for the Barrow region and North Slope of arctic Alaska. Arctic and Alpine Re- search 6:161-184. Price, J., S. Droege, and A. Price. 1995. The summer atlas of North American birds. Academic Press, London, UK. Pulliam, H. R. 1988. Sources, sinks, and population regu- lation. American Naturalist 132:652-661. Pulliam, H. R., and B. J. Danielson. 1991. Sources, sinks, and habitat selection: a landscape perspective on population dynamics. American Naturalist 137:S50-S66. Robbins, C. S., D. Bystrak, and P. H. Geissler. 1986. The Breeding Bird Survey: its first fifteen years, 1965-1979. U.S. Fish and Wildlife Service Resource Publication 157. Robinson, S. K. 1992. Population dynamics of breeding neo- tropical migrants in a fragmented Illinois landscape. Pages 408-418 in J. M. Hagan III and D. W Johnston, editors. Ecology and conservation of Neotropical migrant land- birds. Smithsonian Institution Press, Washington, DC, USA. Rodenhouse, N. L., and R. T Holmes. 1992. Effects of ex- perimental and natural food reductions for breeding Black- throated Blue Warblers. Ecology 73:357-372. Rohwer, S., S. D. Fretwell, and D. M. Niles. 1980. Delayed maturation in passerine plumages and the deceptive ac- quisition of resources. American Naturalist 115:400-437. Rohwer, S., W. P Klein, Jr., and S. Heard. 1983. Delayed plumage maturation and the presumed prealternate molt in American Redstarts. Wilson Bulletin 95:199-208. Sherry, T W, and R. T Holmes. 1991. Population age struc- ture of long-distance migratory passerine birds: variation in time and space. Proceedings of the International Con- gress XX: 1542-1556. Sherry, T W, and R. T Holmes. 1992. Population fluctua- tions in a long-distance Neotropical migrant: demographic evidence for the importance of breeding season events in the American Redstart. Pages 431-442 in J. M. Hagan and D. W. Johnston, editors. Ecology and conservation of Neo- tropical migrant landbirds. Smithsonian Institution Press, Washington, DC, USA. Sibley, C. G., and B. L. Monroe, Jr. 1990. Distribution and taxonomy of birds of the world. Yale University Press, New Haven, Connecticut, USA. Sutherland, W. J., and G. A. Parker. 1985. Distribution of unequal competitors. Pages 255-274 in R. M. Sibly and R. H. Smith, editors. Behavioural ecology: ecological con- sequences of adaptive behaviour. Blackwell, Oxford, UK. Snow, D. 1956. Annual mortality of the Blue Tit in different parts of its range. British Birds 49:174-177. Van Home, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893- 901. Villard, M.A., and B. A. Maurer. 1996. Geostatistics as a tool for examining hypothesized declines in migratory songbirds. Ecology 77:59-68. Wiens, J. A. 1981. Scale problems in avian censusing. Stud- ies in Avian Biology 6:513-521. . 1989. The ecology of bird communities. Cambridge University Press, New York, New York, USA. Wiens, J. A., and J. T Rotenberry. 1981. Censusing and the evaluation of avian habitat occupancy. Studies in Avian Biology 6:522-532. Wilcove, D. S. 1988. Changes in the avifauna of the Great Smoky Mountains: 1947-1983. Wilson Bulletin 100:256- 271. Wilcove, D. S., and J. W. Terborgh. 1984. Patterns of pop- ulation decline in birds. American Birds 38:10-13. Wilkinson, L. 1989. SYSTAT: the system for statistics. SYS- TAT, Evanston, Illinois, USA. Winker, K., J. H. Rappole, and M. A. Ramos. 1995. The use of movement data as an assay of habitat quality. Oecologia 101:211-216.