We studied geographical and sexual variation in the extent of the first prebasic wing molt of Dark-eyed Juncos (Junco hyemalis) using data from migrant "Northern" juncos (J. h. hyemalis; n = 679) and resident juncos from "Pennsylvania" and "Southern" Appalachian populations (J. h. carolinensis; n = 150). In addition, we used skull pneumatization data to assess possible effects of hatching date on molt extent and migration timing of Northern juncos caught at Powdermill Nature Reserve in southwestern Pennsylvania. At a maximum, immature juncos replaced the lesser, median, greater secondary and carpal coverts, proximal and middle feathers of the alula, and proximal three secondaries (i.e., tertials) during their first prebasic wing molt. There were no differences in average molt extent between the Pennsylvania and Southern samples, but juncos in both these samples molted significantly more feathers than Northern juncos. Modal pattern of wing molt in the Northern sample (lesser, median, greater secondary and carpal coverts) was approximately the minimum observed for the Pennsylvania and Southern samples. Within all three samples males accomplished slightly more molt than females, although the difference was not statistically significant for the Southern sample. Juncos in the Northern sample caught during the second half of the spring and fall migration seasons had molted significantly fewer feathers than juncos caught during the first half of those seasons. Because there was little or no relationship between relative hatching date (estimated from skull data) and molt extent or migration timing, we attribute the observed intraseasonal variation to geographical variation within the Northern sample. Within the latitudinal range represented by this sample, juncos from more northerly areas likely molt less and migrate through Powdermill later in both the spring and fall compared to juncos from farther south.

()arnegie Museum of Natural History Powdermill Nature Reserve HC 64, Box 453 Rector, Pennsylvania 15677-9605 USA Department of Biology Indiana University of Pennsylvania Indiana, Pennsylvania 15705 USA VARIACIONES EN LA EXTENSION DE LA PRIMERA MUDA PREBfi. SICA DEL ALA EN JUNCO HYEMAHS Sinopsis.--Estudiamos variaciones geogrfificas y sexuales en la extensi6n de la primera muda prebfisica del ala en Junto hyemalis usando datos de los migrantes "nortefios" (J. h. hyemalis; n = 679) y de las poblaciones residentes de "Pennsylvania" y del sur de los Apalaches (/. h. Carolinensis; n = 150). Ademfis, usamos datos de pneumatizaci6n craneal para riotar posibles efectos de la fecha de eclosi6n en la extensi6n de la muda y del tiempo de la migraci6n de la poblaci6n migrante tornados en la Reserva Natural de Powdermill en el suroeste de Penn- sylvania. A lo mfiximo, los inroadufos reemplazaron las secundarias menores, medias y su- periores y las cubiertas carpales, plumas proximales y medias de la alula, y las tres secunda- rias proximales (i.e., terciarias) durante su primera muda prebfisica del ala. No hubo dife- rencias en la extensi6n promedio de la muda entre las muestras de Pennsylvania y del sur, pero hubo en ambas poblaciones aves que mudaron significativamente mas plumas que las aves nortefias. E1 patr6n modal de la muda en la muestra nortefia (secundarias menor, me- diana y mayor y cubiertas carpales) se aproxim6 al mlnimo obsexxado en las muestras de Pennsylvania y del sur. En las tres muestras los machos mudaron ligeramente mils que las hembras, aunque la diferencia no fu6 estadisticamente significativa para la muestra del sur. Aves de la poblaci6n nortefia capturados durante la segunda mitad de las estaciones migra- torias de primavera y de otofio hab[an mudado significativamente menos plumas que aves capturadas durante la primera mitad de estas estaciones. Debido a la poca o ninguna rela- ci6n entre la fecha de eclosi6n (estimada de datos craneales) y la extensi6n de la muda o el tiempo de la migraci6n, atribumos las variaciones intraestacionales a la variaci6n grogr/t- fica dentro de la muestra nortefia. En la extensi6n latitudinal representada pot esta mues- tra, aves de las poblaciones m/rs nortefias probablemente mudan menos y migran a trav6s de Powdermill m/rs tarde durante ambas migraciones primaveral y otofial en comparaci6n con aves de areas m/rs al sur. The first prebasic molt of most passerines is incomplete (Dwight 1900, Ginn and Melville 1983, Jenni and Winkler 1994, Mulvihill 1993, Pyle et al. 1987, Stresemann and Stresemann 1966, Svensson 1992). Within spe- cies the extent of this molt frequently varies with sex, hatching date, and/or geographical origin. Although this variation has been well docu- mented for many European passerines (Jenni and Winkler 1994), detailed quantitative studies of individual variation in the extent of this molt cur- rently are limited to a handful of North American species (Pyle 1995). This is unfortunate because knowledge of intraspecific variation in the pattern and extent of the first prebasic molt facilitates recognition of adult and immature age classes (Jenni and Winkler 1994, Mulvihill 1993, Pyle 1995), enabling researchers to account for an important variable in many studies of migration, population dynamics, reproductive and foraging ecology, and morphology (Mulvihill 1993). The purpose of our study was to describe individual variation in the first prebasic molt of feathers in the upper alar (wing) tract of the Dark- eyed Junco (Junco hyemalis). Data were analyzed for possible differences in the extent of this molt related to sex, geographic origin, and hatching date ofjuncos in our sample. MATERIALS AND METHODS We recorded the pattern and extent of the first prebasic molt of the major feathers (Fig. 1) of the upper alar tract of 829 immature Dark-eyed Juncos, mostly live birds (n = 785) trapped for banding, but also study skins in the collections of Carnegie Museum of Natural History (CMNH; n = 17) and U.S. National Museum (USNM; n = 27). The following com- prised our geographic subsamples: (1) 679 individuals of the migrant northern subspecies, J. h. hyemalis, caught during the course of year- round banding operations from 1986-1988 at the field research station of CMNH (Powdermill Nature Reserve [PNR]) in the mountains of south- western Pennsylvania (for details of the PNR banding program, see Leb- erman and Wood 1983); (2) 35 live birds banded in 1989 and 27 study skins of southern J. h. carolinensis from the mountains of Virginia, Tennessee, and North Carolina; and (3) 71 live birds banded from 1983-1989 and 17 study skins from a breeding population in the mountains of southwest- ern Pennsylvania. Juncos from this population also are ascribable to caro- linensis (Mulvihill and Chandler 1991, Mulvihill 1992), which differs from c C FIGURE 1. Major feathers and feather groups of the upper alar tract of Dark-eyed Juncos: LC (lesser coverts); MC (median coverts); AC (alula covert); A1 (middle alula feather); A2 (distal alula feather); CC (carpal covert); GPC (greater primary coverts); GSC (greater secondary coverts); Pl-P9 (primaries one through nine); S1-S9 (secondaries one through nine; S7-S9 are the "tertials"). nominate hyemalis in size, plumage color, and bill coloration (Miller 1941). These three groups were identified as Northern, Southern, and Pennsylvania samples, respectively. Both the Southern and Pennsylvania samples came from populations that are largely sedentary (Hostetter 1961, Mulvihill and Chandler 1991, Rabenold and Rabenold 1985). One problem in conducting a study such as this is the difficulty of dis- tinguishing between retained juvenal and molted first basic feathers. These two feather generations are similar in appearance in many small emberizids, includingjuncos (Mulvihill 1993). However, under good light- ing conditions differences between these feather generations in juncos are discernable, making possible the accurate identification of immature birds (hatching year [HY] or second year [SY] birds in banding terminol- ogy) beyond the completion of skull pneumatization, until the latter stages of the second (i.e., definitive) prebasic molt. First basic feathers were recognized (in approximate order of importance) by color, luster, wear, and size (see Mulvihill 1993). Retained juvenal feathers were always browner (with narrow tan edging), less shiny, and noticeably more worn than comparable molted feathers, which were usually lustrous dark gray, with broad light gray edges. In addition, juvenal feathers were often smaller than their first basic counterparts. They were particularly easy to distinguish when a given feather group (e.g., greater secondary coverts, alula, or tertials) was composed of both juvenal and recently grown first basic feathers. Rarely, adult juncos may retain some feathers following their usually complete prebasic molt (Yunick 1976). Four adult females caught during our study showed asymmetrical retention of one or more alular feathers, one or two greater primary coverts, and up to two second- aries. Such birds cannot be confused with immatures, however, because the retained, worn, brownish feathers are isolated in an otherwise entirely fresh plumage. Juncos in our study were sexed by the presence in breeding (SY) indi- viduals of a brood patch (female) or cloacal protuberance (male), or by a combination of wing length and plumage characters (summarized in Ketterson and Nolan 1976). Birds that could not be sexed by these crite- ria (n = 65; <10% of the total sample) were not included in the study. Molt pattern and extent was determined for both wings of juncos in the sample by recording the major feathers of the upper alar tract that had molted. Individuals with asymmetrical molt were included in the sample if, after the exclusion of tkathers that clearly were replaced ad- ventitiously (i.e., feathers never observed to have been molted symmetri- cally by anyjunco in the sample), there was a difference of no more than two feathers between wings. The molt of the right wing of these birds (n = 240) was arbitrarily chosen for all analyses; for the majority of these (70%) the extent of molt differed by only one feather between wings. Northern, Pennsylvania, and Southern samples were included in analy- ses of sexual and geographic variation. Statistical comparisons were made using Mann-Whitney U-tests; differences were considered to be significant at the P--< 0.05 table-wide c-level (sequential Bonfrroni adjustment; Rice 1989). Seasonal analyses were conducted only for the Northern sample, be- cause comparable data were not available for the other samples. For these analyses, the migration seasons were defined as follows: spring, 5 March-5 May; fall, 10 October-15 December. The end of the spring and the be- ginning of the fall season represent the latest and earliest dams of cap- ture, respectively, of immature juncos during the years of this study, prior to and following the species' summer-long absence at PNR. The onset of the spring migration and the end of the fall flight approximate the dates when the first and last non-winteringjuncos are encountered at PNR (i.e., juncos that were neither banded nor recaptured during the intervening winmr season). Therefore, with few exceptions juncos in the Northern sample were passing migrants, not local wintering birds. For intraseasonal analyses, spring and fall samples were divided chronologically into "Ear- ly" and "Late" half-seasons: these were calculated to be about 20 d for males and 25 d for females in the spring; fall half-seasons were about 30 d for both sexes. We assessed the effects of differences in hatching date on seasonal variation in molt extent by analyzing the molt of fall migrant Northern juncos in relation to both capture date and skull pneumatization stage, which was scored on an scale of 1-6 (>80% unpneumatized to fully pneumatized). Because extent of skull pneumatization is largely a function of a bird's age (Nero 1951, Winkler 1979), it provided information on relative hatching dates for juncos in our sample (cf. Jenni and Winkler 1994). Data from recaptures suggested that the interval between successive skull scores was about 35 d in juncos. Because the fall half-seasons were <35 d, we assumed thatjuncos at skull stage n q- 1 were older than juncos at skull stage n within the Early and Late fall samples. Similarly, we assumed that birds caught at skull stage n in the Early half of the season were, on average, about the same age as birds caught at skull stage n q- 1 in the Late half of the season. Seasonal differences in extent of molt were assessed using G-tests of indepen- dence (Wilkinson 1990); a-level was the same as for Mann-Whitney U-tests. There were no between-year differences in molt extent of male and female juncos in spring and fall; therefore data from 1986-1988 were combined for analyses. RESULTS Extent of the molt.--At a minimum, the first prebasic molt in Dark-eyed Juncos replaced the lesser and median coverts (and not necessarily all of these) of the upper alar tract. This pattern was recorded for three female Northern juncos prior to the current study (no junco failed to molt at least two greater secondary coverts during this study; Fig. 2); in these in- dividuals some retained body plumage was noted as well, including up- per tail coverts and feathers in the scapular and/or capital tracts (none was actively molting). The maximum extent of molt observed during this study included the following 16 feathers in addition to all of the lesser and median coverts: all ten greater secondary coverts, the carpal covert, the alula covert, the middle feather of the alula, and all three tertials (Fig. 2). A molt this extensive was recorded for just three juncos during the study (one male each from the Northern, Pennsylvania, and Southern samples) and for one Northern female caught after the study. A North- ern male caught after the study had additionally molted the large, distal feather of the alula on both wings. There was no evidence of normal (i.e., symmetrical) first prebasic molt of the primaries, greater primary coverts, secondaries (other than the tertials), or, with the single exception noted above, the distal feather of the alula. The sequence of the molt.--Although all birds in this study had completed or nearly completed the first prebasic molt, an approximate sequence of feather replacement during this molt can be inferred from the following molt correlations. Of those birds that molted --<5 greater secondary co- verts (n = 15), only 20% (n = 3) replaced the carpal covert.; of those birds that molted >5 greater secondary coverts (n = 814), 92% (n = 749) replaced the carpal covert. Of those birds that molted one or more ter- tials (n = 320), all replaced the carpal covert, and 98% (n = 313) re- placed all ten greater secondary coverts; of those birds that molted just one tertial (n = 211), 91% (n = 191) replaced the proximal tertial (sec- ondary 9); of those birds that molted two tertials (n = 85), 92% (n = 78) Minimum wing molt LC MC GSC 9 8 6 5 4 3 2 1 7 ] CC AL 1 2 3 4 5 6 7 8 9 GPC PP Maximum wing molt LC MC GSC GPC SS PP 8 6 5 4 3 2 1 1 2 3 4 5 6 7 $ 9 7 FIGURE 2. Minimum and maximum molt (shaded areas) observed during the course of this study. LC = lesser coverts; MC = median coverts; CC = carpal covert; AL = alula; GSC = greater secondary coverts; GPC = greater primary coverts; SS = secondaries; PP = primaries. replaced the inner two (secondaries 8 and 9). Of those birds that molted at least one alular feather (n = 150), 99% (n = 149) replaced all ten greater secondary coverts and the carpal covert; of those birds that molted only one alular feather (n = 144), 21% (n = 30) replaced no tertials, 37% (n = 54) replaced one tertial, and 42% (n = 60) replaced two or more tertials. Of those birds that molted one tertial (n = 211), 28% (n = 60) replaced at least one alular feather; of those birds that molted two or more tertials (n = 109), 59% (n = 64) replaced at least one alular feather. Lastly, of those birds that molted two alular feathers (n = 6), 16.7% (n = 1) replaced no tertials, 33.3% (n = 2) replaced one tertial, and 50% (n = 3) replaced all three tertials. Based on these correlative patterns, we propose the following as a typical sequence of feather replacement during the first prebasic wing molt in Dark-eyed Juncos: lesser and median coverts, proximal greater secondary coverts (10 through 5), carpal covert, distal greater secondary coverts(4 through 1), proximal tertial (secondary 9), alula covert, middle tertial (secondary 8), distal tertial (secondary 7), and middle alula. Geographic and sexual variation.--The number of wing feathers re- placed by juncos during the first prebasic molt was variable within and among geographic regions (Fig. 3). In both sexes of Northern juncos, about 50% of the birds molted 11 upper alar feathers additional to the lesser and median coverts (ten greater secondary coverts and the carpal covert) (Fig. 3). The modal pattern for Northern juncos, however, was similar to the minimum for Pennsylvania and Southern juncos of either sex. The maximum extent of molt was the same across samples, while the minimum molt, as well as the average number of molted secondary co- verts, alula feathers, and tertials differed markedly for the Northern sample compared to the Pennsylvania and Southern samples (Table 1). Both sexes of the Northern sample molted significantly fewer total feath- ers than their counterparts in the Pennsylvania and Southern samples; there were no differences between Pennsylvania and Southern juncos (Table 1). Within each geographic sample, males accomplished more molt than females, although not significantly so the Southern sample (U= 360.0, P = 0.113). Differences in the number of molted tertials con- tributed most to both the sexual and geographic variation in our sample ofjuncos (Table 1, Fig. 4). Within-season variation.--Both male and female Northern juncos caught in the Late half of the spring and fall seasons molted significantly fewer feathers than their Early season counterparts (Table 2). By separat- ing Early and Late fall samples according to skull stage we were able to assess the possible effects of hatching date, both on migration timing and molt extent in juncos. Comparison of the number of individuals at skull stage n in the Early fall samples and skull stage n + 1 in the Late samples provided little or no evidence that later hatched juncos in fact migrate later in the fall. The number of birds with lower and higher skull scores did not show large increases and decreases, respectively, between the Early and Late samples, as would be expected if later hatched birds migrated through PNR later in the season (Table 3). Furthermore, within each half- season there was no consistent trend for more extensive molt across skull stages, as would be expected if earlier hatched juncos had a more exten- sive molt than late-hatched juncos. For juncos of similar age, however, there was a consistent pattern of less extensive molt byjuncos caught later in the season (compare average molt extent in Early skull stage n with Late skull stage n + 1; Table 3), in agreement with results obtained when differences in age were not taken into account (Table 2). There were no significant differences in molt extent between the fall and spring samples (Table 2). 200 160 - 2o- 80- 40- 0 25 20- 0 0 15 12- :.:;[] FEMALE EN NSYLVAN IA = 88) SOUTHERN (a = 62) nt t, 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 NO. OF FEATHERS MOLTED FIGURE 3. Frequency distribution showing the extent of the first prebasic molt of upper wing feathers for male and female juncos from three geographic areas. Vol. 68, No. 2 Molt in Juntos [191 Northern Juncos (n = 679) LC MC GSC CC GPC $$ PP 8 6 5 4 3 2 1 1 2 3 7 4 5 6 7 8 9 0 10 30 50 70 90 I0O$ Pennsylvania & Southern Juneos (n = 150) LC MC CC GSC SS 9 8 7 GPC PP 6 5 4 3 2 1 I 2 3 4 5 6 7 8 9 FIC, URE 4. Geographic variation in the percentage ofjuncos replacing various feathers dur- ing the first prebasic molt. LC = lesser coverts; MC = median coverts; CC = carpal covert; AL = alula; GSC = greater secondary coverts; GPC = greater primary coverts; SS = secondaries; PP = primaries. TABIrE 2. Inter- and intraseasonal comparisons of the extent of the first prebasic molt (num- ber of feathers molted out of a possible 16) for female and male Northern juncos. Sta- tistical comparisons are two-tailed Mann-%hitney /::tests. Females Males Season Mean SD n Mean SD n Spring 10.8 1.7 157 11.5 1.4 117 Early b 11.0 1.6 90 11.8 1.4 54 Late b 10.5 1.9 67 11.3 1.3 63 FalP 10.4 1.9 197 11.1 1.6 200 Early c 11.0 1.3 84 11.6 1.4 94 Late c 10.0 2.1 113 10.7 1.8 106 females: U = 16502.0, P = 0.176; males: U = 12977.0, P = 0.086 females: U = 3643.0, P = 0.016; males: U = 2039.5, P = 0.047 females: U = 5706.5, P = 0.004; males: U = 6635.5, P = 0.000 DISCUSSION Dwight (1900) described the postjuvenal (first prebasic) molt of Dark- eyed Juncos as including "the body plumage and wing coverts, but not the rest of the wings or tail." Yunick (1981) observed that a small per- centage of immature juncos may replace 1-3 tertials. From Pyle et al. (1987) it can be inferred that from 1-5 distal greater secondary coverts may be retained in the first basic plumage. Based on the present study, the incomplete first prebasic molt of the Dark-eyed Junco typically replaces all of the body feathers, the lesser, median, greater secondary and carpal coverts; less often the alula covert and the proximal two ter- tials; rarely the middle alula feather and the distal tertial (Fig. 4). The pattern of the first prebasic molt in juncos is similar to that observed for several other North American and European emberizine finches (e.g., Melospiza lincolnii, Spizella passerina, S. arborea, Zonotrichia albicollis, Mul- vihill unpubl. data; Emberiza citrinella, E. cia, and E. schoeniclus, Jenni and Winklet 1994). The within and between wing-feather tract sequence of feather replacement during the first prebasic molt in Dark-eyed Juncos inferred from our molt correlations is similar to that observed in many European passetines (Jenni and Winklet 1994:32ff) and to several North American emberizids examined while in active molt (Mulvihill, unpubl. data). Typically, the juvenal feathers replaced during an incomplete first prebasic molt are those that primarily have an insulatory, rather than aero- dynamic, function; furthermore, they are the feathers most exposed to physical abrasion and most likely to become worn if they were retained until the definitive prebasic molt (Dwight 1900, Miller 1928). It may be significant that injuncos the juvenal distal feather of the alula is virtually always retained until the definitive prebasic molt, despite its exposed position on the wing. In fact, this feather is commonly included in the partial molt of only a very few passetines (e.g., two out of 58 spe- TABLE 3. Comparison of molt (mean number feathers replaced out of 16 possible) and sample sizes across skull stages between Early and Late halves of the fall migration sea- son for Northern juncos. The distribution of birds among skull stages in the Late half of the season was compared to three null distributions, each of which described an expected average incremental increase of one skull stage between the Early and Late periods (see Methods). Differences were assessed with likelihood ratio chi-square statis- tics. Skull Early Late Null distributions a stage Mean n Mean n A B C Females Probabilities Males Probabilities 1 10.8 15 10.5 6 0 3 5 2 11.1 28 9.4 22 15 15 14 3 10.8 30 10.0 36 28 25 24 4 11.4 10 10.3 38 30 27 23 5 11.0 1 10.3 11 10 12 13 6 -- 0 -- 0 1 2 5 0.120 0.447 0.055 1 10.8 18 8.5 10 0 3 6 2 11.8 34 10.6 20 18 18 17 3 11.5 31 11.1 40 34 30 27 4 11.8 11 10.9 28 31 29 26 5 -- 0 10.9 8 11 12 15 6 -- 0 -- 0 0 2 3 0.006 0.136 0.092 a Distributions based on null hypotheses that the number of birds in Late skull stage "n" equals: A) the number in Early skull stage n - 1; B) kg the number in Early skull stages n and n - 2 and % of the number in Early stage n - 1; or C) / each of the number of birds in Early skull stage n, n - 1, and n - 2. cies of European passerines studied by Jenni and Winkler 1994). Reten- tion by juveniles of the large, distal feather of the alula may be due in part to its putative aerodynamic function as a wing slot, providing addi- tional lift at slow flight speeds and facilitating take-offs (Kokshaysky 1973, Savile 1957). In juncos, the distal feather of the alula actually is slightly longer (by about 0.5 mm) in juveniles than adults (Mulvihill, unpubl. data), despite the fact that overall wing length is over 1 mm shorter on average in the immatures (Mulvihill and Chandler 1990). It is possible that a larger alula to wing length ratio gives an aerodynamic advantage to immature juncos (and probably immatures of other species as well; cf. Alatalo et al. 1984) sufficient to make replacement of the distal feather of the alula during the first prebasic molt disadvantageous. Additional lift provided by the longer juvenal alula may be especially important for im- mature juncos during the migration and winter seasons, because fat- related increases in body mass and wing loading are significantly greater in this age class than in adults (Chandler and Mulvihill 1992). Geographic and seasonal variation.--Southern populations of juncos (represented by the Pennsylvania and Southern samples in this study) molted significantly more feathers than birds in the Northern sample. Such geographical differences are common in passerines (e.g., Baillie and Swann 1980; Dwight 1900; Miller 1928, 1933; Pitelka 1945), and often are attributed to differences among populations in the onset of nesting (hence, hatching date differences) and or the date when molt necessarily ceases due to energetic demands associated with migration or survival in a cold environment (e.g., Hereward 1979, Mewaldt and King 1978, Miller 1933, Stangel 1985, Stresemann and Stresemann 1966). Within a population earlier-hatched birds typically have a longer po- tential molt period and consequently a more extensive molt than later hatched birds (Dhondt 1973, Ligon and White 1974, Michener and Michener 1940, Scott 1967, Wiseman 1977). The effect of hatching date on molt extent, however, may be reduced or even eliminated for popula- tions in which later-hatched birds both initiate molt sooner after hatch- ing and molt more intensely and rapidly than their earlier hatched coun- terparts (e.g., Dolnik and Blyumenthal 1967, Dolnik and Gavrilov 1980, Michener and Michener 1940, Newton 1966). Given the absence of sig- nificant differences in molt extent between juncos at different skull pneu- matization stages during each half of the fall migration (Table 4), there would appear to be factors promoting such molt synchronization in Northern juncos. This would also explain why molt extent for Northern juncos in this study was unimodal (Fig. 3), in spite of the fact that J. h. hyemalis occupies a wide latitudinal range and is double-brooded through- out that range (Bent 1968). Within the Northern sample, male and female juncos caught during the second half of both migration seasons had molted significantly fewer feathers. This pattern of intraseasonal variation has been found in fall but not in spring samples of several migrant European passetines (Jenni and Winklet 1994:41). Such differences in the fall frequently are attributed to less extensive molt and later migration of late-hatched birds, birds from more northerly populations, or some combination of the two (Jenni and Winklet 1994:42). Because we observed no significant effects of hatching date on molt extent or migration timing, however, we tentatively attribute the pattern of reduced molt in Northern juncos caught during the sec- ond half of both the spring and fall migration seasons at PNR to less ex- tensive molt and later migration by juncos from more northerly popula- tions within the range represented by our Northern sample. Because the nominate race of the Dark-eyedJunco exhibits little morphological varia- tion across most of its range (Miller 1941), we could not test the hypoth- esis of a relationship between molt extent, migration timing, and geographical origin using an independent morphological trait (e.g., wing length; Jenni and Winklet 1994:42). Differences between the sexes in extent of molt.-A significantly more ex- tensive first prebasic molt has been observed in males of several Euro- pean passetines (Jenni and Winklet 1994:41), but this variation has been documented for only a few North American species (e.g., Aphelocoma spp.; Pitelka 1945). The magnitude of the sex differences in molt extent in jun- cos (Table 1) is similar to that reported in thirteen species of European passetines (Jenni and Winklet 1994:41). Possible factors contributing to a slightly more extensive molt by male juncos include differences between the sexes in migratory behavior, social dominance interactions, and win- tering environment. The sex differences in molt observed for Northern juncos in this study are consistent with an hypothesis that migratory behavior affects the ex- tent of the first prebasic molt by reducing the time available for molting, because females migrate about ten days earlier than males on average (Chandler and Mulvihil 1990, Mulvihill and Chandler 1991). However, because more extensive molt by males has been observed in several non- migratory passetines (e.g., Dhondt 1973, Newton 1966, Pitelka 1945), as well as in the largely sedentary Pennsylvania juncos in this study, other factors also may contribute to this variation. Jenni and Winklet (1994:41) proposed that for species in which the wing plumage of males is more brightly colored or conspicuously patterned, a sex difference in molt extent might be related to the impor- tance of status signaling for males in social contests during winter and the succeeding nesting season. Although the wing plumage of male jun- cos is neither bright nor strongly patterned, retained brownish juvenal wing coverts and tertials are more conspicuous in the males than females (Mulvihill, pets. obs.). Because brown plumage may be associated with status signaling in juncos (Ketterson 1979a, 1979b), at least within their cohort, males that have replaced more juvenal plumage may be more likely to enjoy benefits associated with social dominance including increased access to food resources, decreased exposure to predation, and greater success in acquiring and defending a territory in the following breeding season (Rabenold and Rabenold 1985). Differences in the levels of nutrient and energy stress also have been suggested as a possible cause for sex differences in the extent of molt (Jenni and Winklet 1994:41). Male juncos are socially dominant to females in winter (Ketterson 1979b), but whether or not male dominance in any way limits the access of female juncos to resources necessary for achieving maximum rates of feather synthesis during the course of their first prebasic molt is not known. Because male juncos differentially winter in areas characterized by more severe climatic conditions than females (more northerly latitudes in hyemalis; higher elevations in carolinensis) (Ketterson and Nolan 1983, Ra- behold and Rabenold 1985), a more extensive molt by this sex may be related to better insulation provided by the additional nonjuvenal wing feathers. In juncos, a difference between the sexes in the inclusion of ter- tials in the first prebasic molt accounts for most of the intersexual varia- tion in all three samples. When a bird is at rest, with its wings folded over its back, the tertials, sometimes termed "shield feathers" (Svensson 1992), may have an important insulative function as a barrier against heat loss through the lateral apteria. Finally, at least in the case of Northern juncos, sex differences in molt may simply reflect the within-sample geographical variation discussed above. Because Northern juncos exhibit differential migration, with many more immature females than males wintering to the south of PNR (Ketter- son and Nolan 1985), the proportion of females passing through PNR from more northerly areas within the extensive geographic region repre- sented by our "Northern" sample (i.e., from populations possibly charac- terized by less extensive molt) probably is greater than for males (Ketter- son and Nolan 1982). ACKNOWLEDGMENTS We gratefully acknowledge field assistance given to us at Powdermill by R. C. Leberman and numerous volunteers and at Mountain Lake Biological Station by C. ZiegenfBs. For the loan of specimens used in this study we thank those responsible at the U.S. Na6onal Mu- seum. For their help and logistical support we also thank K. C. Parkes, D. S. Wood, J. Lough- lin, R. Panza, and M. Niedermeier. For allowing us access to field sites for the purpose of banding Pennsylvania breedingjuncos we thank R. Ache, D. Schiller and D. Willians (Penn- sylvania Bureau of Forestry), B. Bromley, L. Golden, and IL Williams (Pennsylvania Bureau of State Parks), University of Pittsburgh, R. McMullen (Valley School of Ligonier), Mr. and Mrs. H. Stiff (Rolling Rock Farms), Mr. and Mrs. G. Rennels, T. Kobil, Mr. and Mrs. W. McCray, and R. C. Abbott. LITERATURE CITED ALATALO, R. V., L. GUSTAFSSON, AND A. LUNDBERG. 1984. %qy do young passetines have shorter wings than older birds? Ibis 126:410-415. B?dLL1E, S. R., AND R. L. SWANN. 1980. The extent of post-juvenile mottit in the Blackbird. Ringing and Migr. 3:21-26. BENT, A. C. 1968. Life histories of North American cardinals, grosbeaks, buntings, towhees, finches, sparrows, and allies, Part 2. U.S. National Mus. Bull. no. 237, Smithsonian In- stitution Press, Washington, D.C. 1248 pp. CHANDLER, C. R., AND R. S. MULVIHILL. 1990. Wing-shape variation and differential timing of migration in Dark-eyed Juncos. Condor 92:54-61. --, AND --. 1992. Effects of age, sex, and fht level on wing loading in Dark-eyed Juncos. Auk 109:235-241. DHONDT, A.A. 1973. Postjuvenile and postnuptial moult in a Belgian population of Great Tits, Parus major, with some data on captive birds. Gerfaut 63:187-209. DOLNIK, V. R., AND T. I. BLYUMENTHAL. 1967. Autumn premigratory and migratory periods in the Chaffinch and some other temperate zone passerines. Condor 69:435-468. --, AND V. M. GAVRILOV. 1980. Photoperiodic control of molt in the Chaffinch. Auk 97:50-62. DWIGHT, J. 1900. The sequence of plumages and moults of the passerine birds of New York. Ann. New York Acad. Sci. 13:73-360. GINN, H. B., AND D. S. MELVILLE. 1983. Moult in birds. BTO Guide no. 19, British Trust for Ornithology, Tring, England. 112 pp. HEREWARD, A. C. 1979. The autumn moult of the Yellow Wagtail. Ringing and Migr. 2:113- 117. HOSTETTER, R. 1961. Life history of the carolinaJunco (Junto hyemalis carolinensis Brew- ster). Raven 32:97-170. JENNI, L., AND R. WINKLER. 1994. Moult and ageing of European passetines. Academic Press, San Diego, California. 224 pp. KETTEISON, E. D. 1979a. Status signaling in Dark-eyedJuncos. Auk 96:94-99. 1979b. Aggressive behavior in wintering Dark-eyed Juncos: determinants of domi- nance and their possible relation to geographic variation in sex ratio. Wilson Bull. 91:371-383. , AND V. NOLN, JR. 1976. Geographic variation and its climatic correlates in the sex ratio of eastern-wintering Dark-eyed Juncos (Junco hyemalis). Ecology 57:679- 693. , AND --.. 1982. The role of migration and winter mortality in the life history of a temperate-zone migrant, the Dark-eyed Junco, as determined from demographic analyses of winter populations. Auk 99:243-259. , AND 1983. The evolution of differential bird migration. Curr. Ornithol. 1:357-402. , AND 1985. Intraspecific variation in avian migration: evolutionary and regu- latory aspects. p. 553-579 in M. A. Rankin, ed. Migration: mechanisms and adaptive significance. Contrib. Marine Sci. 27 (Supplement). KOKSHAYSKY, N.V. 1973. Functional aspects of some details of bird wing configuration. Syst. Zool. 22:442-450. LEBERMAN, R. C., AND D. S. WOOD. 1983. Bird-banding at Powdermill: twenty years reviewed. P.N.R. Research Report No. 42, Carnegie Mus. Nat. Hist., Pittsburgh, Pennsylvania. 49 pp. LIGON, J. D., AND J. W. WHITE. 1974. Molt and its timing in the Pition Jay, Gymnorhinus cyanocephalus. Condor 76:274-287. MEWALDT, L. R., AND J. R. FANG. 1978. Latitudinal variation of postnuptial molt in Pacific coast White-crowned Sparrows. Auk 95:168-179. MICHENER, H., AND J. R. MICHENER. 1940. The molt of House Finches of the Pasadena region, California. Condor 42:140-153. MILLER, A. H. 1928. The molts of the Loggerhead Shrike, Lanius ludovicianus Linnaeus. Univ. Calif. Publ. Zool. 30:393-417. 1933. Postjuvenal molt and the appearance of sexual characters of plumage in Painopepla nitens. Univ. Calif. Publ. Zool. 38:425-444. 1941. Speciation in the avian genusJunco. Univ. Calif. Publ. Zool. 44:173-434. MULVIHILL, R. S. 1992. Dark-eyedJunco. Pp. 394-395, in D. Brauning, ed. Atlas of breed- ing birds in Pennsylvania. Univ. of Pittsburgh Press, Pittsburgh, Pennsylvania. 1993. Using wing molt to age passetines. N. Am. Bird Bander 18:1-10. , AND C. R. CHANDLER. 1990. The relationship between wing shape and differential migration in the Dark-eyed Junco. Auk 107:490-499. , AND --. 1991. A comparison of wing shape between migratory and sedentary Dark-eyed Juncos (Junco hyemalis). Condor 93:172-175. NERO, R.W. 1951. Pattern and rate of cranial 'ossification' in the House Sparrow. Wilson Bull. 63:84-88. NE'TON, I. 1966. The moult of the Bullfinch Pyrrhula pyrrhula. Ibis 108:41-67. PITELKA, F. A. 1945. Pterylography, molt, and age determination in American jays of the genus Aphelocoma. Condor 47:229-260. P.E, P. 1995. Update on banding guide revision. N. Amer. Bird Bander 20:28-29. , S. N. G. HOWELL, R. P. YUNICK, AND D. F. DESANTE. 1987. Identification guide to North American passetines. Slate Creek Press, Bolinas, Calitbrnia. 278 pp. RABENOLD, K. N., AND P. P. RABENOI,D. 1985. Variation in altitudinal migration, winter seg- regation, and site tenacity in two subspecies of Dark-eyed Juncos in the southern Appa- lachians. Auk 102:805-819. RICE, W. R. 1989. Analyzing tables of statistical tests. Evolution 43:223-225. S^VILE, D. B. O. 1957. Adaptive evolution in the avian wing. Evolution 11:212-224. ScoTt, D.M. 1967. Postjuvenal molt and determination of age of the cardinal. Bird-Banding 38:37-51. STANGEL, P. W. 1985. Incomplete first prebasic molt of Massachusetts House Finches. J. Field Ornithol. 56:1-8. STRESEM,42qN, E., AND V. STRESEMANN. 1966. Die Mauser der V6gel. J. Ornithol. 107, Sonder- heft. SVENSSON, L. 1992. Identification guide to European passetines. 4th ed. Published by the author, Stockhohn, Sweden. 368 pp. WILKINSON, L. 1990. SYSTAT: the system for statistics. SYSTAT, Inc., Evanston, Illinois. 676 pp. WINKLER, R. 1979. Zur Pneumatisation des Schdeldachs der V6gel. Orn. Beob. 76:49-118. WlSEMAN, A. J. 1977. Interrelation of variables in postjuvenal molt of Cardinals. Bird-Banding 48:206-223. YUNICK, R. P. 1976. Incomplete prebasic molt in a Dark-eyed Junco. Bird-Banding 47:276- 277. 1981. Age determination of wintering and spring Dark-eyed Juncos. N. Am. Bird Binder 6:97. Received 6 Mar. 1996; accepted 3 Jun. 1996.