Molt of the Gray Vireo ’

The Condor 102:610-618 0 The Cooper Ornithological Society 2000

MOLT OF THE GRAY VIREO ’

GARY VOELKER Barrick Museum of Natural History, Box 454012, Universityof Nevada Las Vegas, Las Vegas,NV 891.54, e-mail: [email protected]

Abstract. Using museum specimens, I document the molt cycles and molting grounds of the Gray Vireo (Vireo vicinior). During prebasic molts, adult female Gray Vireos replace their primaries in 57 days, whereas adult males take 70 days; all body plumage is replaced during this molt. Prebasic molts occur almost exclusively on the breeding grounds;just 3 of 41 specimensreplacing primaries were collected away from breeding grounds.No molting specimens were collected from wintering areas. Prealternatemolt occurs on the wintering grounds,and appearslimited to the replacementof innermost secondariesand a limited molt of body plumage. By performing prebasicmolt on breeding grounds,the Gray Vireo differs from several other western breeding passerinespecies that use desert regions of the southwestern United States and northwesternMexico to exploit late summer food resources.The areasof these southwesterndesert regions used by other speciesform a small portion of the breeding grounds, and encompassthe entire wintering grounds,of Gray Vireos. I hypothe- size that this contrast in molting regions is not due to differences in the general timing of prebasic molts among these species;rather, the contrastmay be due to constraintsimposed on Gray Vireos by a dietary shift to fruit during winter and the need to defend winter territories. Key words: Gray Vireo, molt, plumage, Vireo vicinior.

INTRODUCTION era1 species, these desert regions do not form Molt is an integral part of the avian annual cy- part of either the breeding or wintering range. It cle, and recent studies have documented inter- has been argued that this stopover is performed esting life-history tradeoffs in the scheduling of in order to take advantage of food flushes pro- annual molts (Thompson 1991, Voelker and duced by late summer rains (Rohwer and Man- Rohwer 1998). In passerines, these tradeoffs ning 1990, Young 1991, Voelker and Rohwer generally involve the scheduling of fall molts 1998). relative to fall migration. For instance, many This molt-migration stopover raises an inter- eastern North American breeding species re- esting question with regard to food resourcepro- place flight feathers while still on their breeding ductivity on the breeding grounds versus non- grounds, perhapsbecause winter ranges are typ- breeding grounds: do birds that breed in the xe- ically much smaller than are breeding ranges, ric western/southwestern U.S., but that winter which could increase competition for the re- farther south, also take advantage of this late- sources necessary to support the high energetic summer increase in resources?If so, we should expect to find that molting birds are predomi- costs of molt (Rohwer 1971, Voelker and Roh- nately found in the molt-migration stopover re- wer 1998). gion, in specieswith disjunct breeding and win- In contrast to eastern birds, adults of many tering ranges that do not include this area. Or, western breeding species of North American we should find that molting birds are found pre- passerines begin fall migration prior to com- dominately in one portion of either the breeding mencing fall molts. This would allow these spe- or wintering range, specifically the molt-migra- cies to avoid the late-summer droughts typical tion area, for specieswhose breeding or winter- for much of western North America (Baldwin ing range includes this area. 1973), and an associatedlack of food resources. In this paper, I begin to addressthis question A taxonomically diverse set of these western by describing the molts and documenting the breeding species appear to stop during fall mi- molting grounds of the Gray Vireo (Vireo vici- gration to commence molt in the Sonoran and nior). This species breeds throughout arid re- Chihuahuan desert regions of the southwestern gions of the western/southwesternU.S., includ- United States and northwestern Mexico; in sev- ing part of the molt-migration area described for other species, and winters primarily along the ’ ’ Received9 July 1999. Accepted 28 February 2000. Gulf of California.

16101 GRAY VIREO MOLT 611

METHODS 1988), and I use it here. A more accuratemethod apparently exists for estimating molt duration SPECIMENS (Underhill and Zucchini 1988), but is not readily I examined 178 specimens from 28 museums available. Furthermore, it is unclear whether a (see Acknowledgments). I requested from each basic assumption of this method, that observa- museum any specimen collected between 1 July tions constitute a random sample from the rele- and 30 April, up to 10 adult specimens from vant population, could be appropriately applied May and up to 10 from June, and any juvenile to specimen-baseddata, or for that matter to any specimen from May and June. I followed Pyle data set where a “relevant population” can not (1997) by referring to birds as either Hatching be unambiguously defined. Year (HY, birds in their first calendar year) or I followed the molt terminology of Humphrey After Hatching Year (AHY, birds in at least their and Parkes (1959). Following Langston and second calendar year). Sex was establishedfrom Rohwer (1996), I define a molt series as a set of museum labels. flight feathers that are replaced by a single set of rules. Yuri and Rohwer (1997) should be con- SCORING MOLT sulted for a detailed explanation of the rules of I generally followed Rohwer (1986) to deter- feather replacement and the identification of mine whether molt was adventitious. I did, how- molt series. Briefly, to identify the feathers of a ever, score asymmetrical primary feather molt if series and the rules by which they are replaced, one to several contiguous primaries on one each growing (focal) feather is placed in one of wing, but not the other, had been dropped, so three categories based on its stage of replace- long as primary one was included in this set of ment and the stage of replacement of adjacent feathers. I did this because primary feather molt feathers. Nodal feathers are those feathers re- is often not synchronized between wings, and placed first in a molt series and these are always thus it is possible for one wing to start replacing more fully grown than adjacent feathers. Ter- feathers slightly ahead of the other wing. minal feathers are those feathers replaced last in In examining specimensfor molt, I used a 3X a molt series and these are always less fully magnifying lamp lighted with a 22-W fluores- grown than adjacent feathers. Other growing cent bulb, and a small forceps to lift feathers and feathers provide directional information, show- quantify molt. For body molt, I estimated the ing whether replacement proceedsproximally to percentage of feathers growing in each of five distally, or distally to proximally. In the simplest regions: chin and throat, breast, belly, head, and possible scenario for primary replacement, Pl back (definitions in Rohwer 1986). I used Roh- would be nodal (dropped/replaced first), subse- wer’s (1986) scale of 0, 10, 30, 50, 70, and 90% quent primaries would be dropped in order with of feathers in development. An overall body P2 longer than P3, P3 longer than P4, and so on molt score was calculated for each specimen by until PlO was reached. This sequencewould in- averaging the scores of the five body regions. dicate proximal to distal replacement of prima- For scoring flight feather molt, I generally fol- ries, and, because PlO was dropped last, would lowed Rohwer’s (1986) practice of estimating indicate that PlO was the terminal feather in the the fraction of full length (by 0.1 intervals) at- series. More complex scenarios would involve tained by each growing feather, but used N to multiple series in a single feather tract, and designate new feathers, which allowed a score would be evidenced by multiple nodal feathers, of zero to be used for missing feathers (Voelker multiple directional waves, and multiple termi- and Rohwer 1998). Gray Vireos have a total of nal feathers (Langston and Rohwer 1995). 20 primaries, 18 secondaries,and 12 rectrices. I scored both sides of each specimen for flight BREEDING AND WINTERING RANGES feather condition. Gray Vireos breed from southern California, All regression models for estimating molt du- southern Nevada, southern Utah, western and ration are flawed. However, Pimm’s (1976) re- southeastern Colorado, and northwestern and gression method to estimate average duration of central New Mexico south to northwestern Baja molt in individual birds is less flawed because California, central and southeastern Arizona, the method overcomes the problem of heteros- southern New Mexico, western Texas, and cedasticity in molt data (Underhill and Zucchini northwestern to central Coahuila (Phillips 1991, 612 GARYVOELKER

AOU 1998). The winter range is central and southern Baja California, southwesternArizona, Sonora, the Big Bend region of western Texas (AOU 1998), and perhaps southwesternCoahui- la (Phillips 1991). In Sonora, the winter range is strongly associatedwith the distribution of elephant trees (Bursera microphylla; Bates 1992a) in coastal and lowland desert scrub areas (Rus- sell and Monson 1998).

MOLT AND MIGRATION To determine where Gray Vireos molt and to determine when they molt in relation to fall migration, I plotted the collecting localities of molting specimens on a map of southwestern North America.

RESULTS FIRST PREBASIC MOLT Based on six HY birds that were undergoing first prebasic molt, this molt appearsto be limited to an intense, complete body molt and the replacement of at least some flight feather coverts. In juvenile plumage, wing bars on the greater secondary coverts are brownish and relatively in- distinct; several specimens undergoing first prebasic molt had some new greater coverts edged white as in AHY birds. Two specimens that had completed this molt also had mostly white- edged greater coverts, along with a few older brownish-edged coverts. Thus, it appears that the first prebasic replacement of greater coverts may not be complete (Pyle 1997). None of these specimens were replacing flight feathers.

DEFINITIVE PREBASIC MOLT This molt involves the replacement of all feathers. Primaries are replaced as a single series, be- ginning at Pland proceeding distally to PI0 (Ta- ble 1). In 15 specimens, Pl was the sole nodal primary; 4 other specimens indicate that Pl and P2 may be dropped simultaneously. The regression of collection date on molt score suggests that primary molt lasts about 67 days when males and females are considered together (Ta- ble 2). However, when considered separately, males appear to start primary molt earlier, and take about 13 days longer to complete the molt than do females (Fig. 1, Table 2). Most passerines replace secondaries in two series: Sl to S6, and S7 to S9. However, Gray Vireos appear to have a replacement pattern that is accomplishedin either two different series, S 1 GRAY VIREO MOLT 613

TABLE 2. Averagestart and completiondates of primary molt for adult male and female Gray Vireos, based on regressions of date on primary molt scores.

n Start date Completion date Days in molt i-2 P

Males 29 3 July 11 September 70 0.93

to S5 and S6 to S9, or in three series, Sl to S5, where S7 and S8 where being replaced simul- S6, and S7 to S9. Without question, the outer taneously, S7 was longer than S8. This would series of secondaries is being replaced distally imply that S7 may have been nodal, and this to proximally (Sl to SS; Table 1). Sl is replaced resulted in one S8 being categorized as distal to when P5-P6 is the outermost growing primary, proximal replacement. Clearly, there does not and the Sl to S5 series is completed at the same appear to be a firm set of rules in the S7-S9 time as the primary series, or shortly thereafter. series, if indeed this set of feathers can be cor- S6 was longer than S5 in all cases where both rectly called a series. Despite the obvious incon- feathers were growing simultaneously. This sistency, I retain the term series for this group, means that S6 was dropped prior to S5, and im- as they are often in other species replaced to- plies that S6 is not part of the Sl-S5 series. gether in prealternate molts, where other sec- For the S7-S9 series in Gray Vireos, S8 is ondaries are most often not replaced. S7 to S9 most typically nodal, with either S7 or S9 being were being replaced when Pl-P5 were the out- dropped next in the series. In one of five cases ermost growing primaries. S6 may either be part of the S7-S9 series, or may perhaps more likely constitute a series by 17 act A itself. In only 1 of 12 cases was S6 growing T concurrently with S7 such that S6 could be 16 Sep viewed as the terminus of the inner molt series. In the other 11 cases, S6 was scored as a terminal feather but S7 was new and fully grown. 17 Aug In about half of these cases, S6 was quite short which suggeststhat it may have been dropped 19 Jul after S7 was fully grown. This implies that S6 may be an independent series, because in most 19 Jun molts, adjacent feathers are dropped before the 0 5 10 15 20 preceding feather in a series completes growth. Rectrices appear to be divided into two molt series: Rl-R5, and R6. Feather replacement pro- 17Oct T ceeds from RI to R5 in the inner series (Table 1). R6 was scored as terminal in nine cases 0 16 SW (shorter than R5) which would suggest that it was likely part of a single rectrix series, Rl-R6. However, R6 was scored as nodal in three cases (longer than R5); to be considered nodal, R6 must have been dropped prior to R5, and thus R6 constitutes a separate molt series. Replace- ment of the rectrices commences at about the same time as does the replacement of the inner 19 Jun / 0 5 10 15 20 secondary series, and is completed just prior to the completion of the inner secondary series. Molt score Definitive prebasic body molt begins soon af- FIGURE 1. Regression of date on primary molt ter primary molt begins and appears to be oc- score for the Gray Vireo. A = males, B = females. curring in all tracts by the time P4-P5 are being 614 GARY VOELKER

TABLE 3. Prealternate molt in Gray Vireos.

% in Mean body % replacing % with fresh % replacing % with fresh Date n molt molt score secondaries secondaries rectrices rectrices

ltol5March 2 100 36 50 0 0 0 16 to 31 March 1 0 0 0 0 0 0 1 to 15 April 12 17 18 17 8 0 8 16 to 30 April 12 25 6 25 25 0 0 1 to 15 May 17 12 2 6 18 6 0

replaced. Body molt is completed at about the age in all tracts (Table 3). Five banded speci- same time as the primary series is completed. mens of 35 trapped in March and April from the wintering grounds each showed a light to mod- DEFINITIVE PREALTERNATE MOLT erate amount of ventral molt (S. Russell, unpubl. Of 45 specimenscollected between 1 March and data). These data combine to suggestthat preal- 15 May, only 20% (n = 9) were molting body ternate molt may be generally limited to the re- feathers, flight feathers, or both (Table 3). Body placement of S7 to S9 and at least a limited re- molt was not very intense on any specimen; no placement of body feathers. specimen had more than a 40% total body molt score, and just two specimens were simulta- MOLT IN RELATION TO MIGRATION neously replacing feathers in all tracts. Just one Definitive prebasic molt appears generally to be specimen was replacing rectrices, and another both initiated and completed on the breeding had some fresh rectrices (Table 3). Seven spec- grounds. Of 41 AHY specimensundergoing this imens were replacing S7 to S9. Seven other molt, 93% (n = 38) were replacing flight feath- specimens, collected in April and May, had ers on the breeding grounds, and 24% (n = 10) fresh-appearing S7 to S9, and fresh body plum- of these were replacing PlO, which suggeststhat most birds are completing the prebasic molt on the breeding grounds. Of the three molting specimens collected away from the breeding grounds, none had yet reached probable wintering areas (Fig. 2). Two of these were taken in Sonora in late September (15, 30), and had nearly completed the molt, whereas the third was taken in Durango on 13 August and had just dropped Pl-P2; these three collection dates were later than the normal ini- tiation and completion dates (Table 2). This suggeststhat these birds finished breeding later than usual and were forced to initiate migration prior to initiating molt.

DISCUSSION

MOLT PATTERNS Gray Vireos replace their primaries according to the rules followed by most passerines.Under the assumption that multiple nodal feathers indicate multiple series, three specimenssuggest that rec- FIGURE 2. Geographic distribution of molting Gray trices are being replaced in two series, Rl-R5 Vireo specimens (open circles). The cross-hatched area and R6; this pattern has been documented in denotes the approximate breeding range. The stippled several other species (Jenni and Winkler 1994, area denotes both the approximate wintering range of Gray Vireos, and the distribution of elephant tree (Bur- Yuri and Rohwer 1997, Voelker and Rohwer sera microphylla). 1998) and may be a general pattern among pas- GRAY VIREO MOLT 615 serines as more detailed molt studies are con- Vireos to take advantage of food flushes asso- ducted. Data from several specimensthat had S6 ciated with late summer monsoons in south- longer than S5, suggestthat S6 is either part of western United States and northwestern Mexi- the inner series of secondaries (i.e., S6-S9) or can deserts (Szarek 1979, Nielson 1986). Four constitutes a molt series by itself. If S6 is part other western North American speciesthat breed of the S7-S9 series, then this is not a novel re- in arid, or in dry, riparian habitats perform such sult among passerines; individuals of several a molt-migration (Bullock’s Oriole Zcterus bul- Old World species have been found to drop S6 lockii, Rohwer and Manning 1990; western prior to SS (Jenni and Winkler 1994), meaning Painted Bunting Passerina ciris pallidior, that in these individuals, S6 is not part of the Thompson 1991; Lazuli Bunting Passerina Sl-S5 series. Indeed, in several species, all in- amoena, Young 199 1; Western Warbling-Vireo, dividuals show a similar “convergent” pattern Voelker and Rohwer 1998). of secondary replacement proximally from Sl Prior to examining specimens, I had predicted and distally from S6, such that S6-S5 and per- that Gray Vireos would be taking advantage of haps S6-S4 are replaced as a series (Jenni and the late summer food flush by moving away Winkler 1994). I have assumed that there is no from northern and eastern portions of their individual variation in replacement patterns, and breeding range where resource productivity pre- that therefore each feather must belong to a spe- sumably drops in late summer. Thus, I had ex- cific molt series. But the available molt data pected to find that molting birds were either col- does not preclude the possibility that S6 can be lected only in the southwestern portion of the replaced in any series, depending on the indi- breeding range (i.e., the Sonoran and Chi- vidual bird. However, variation in flight feather huahuan deserts), or that molting specimenshad replacement within many species seems to in- been taken there as well as farther south into volve the number of feathers replaced (due to Sonora, to include the wintering grounds. This time constraints), rather than the pattern in does not appear to be the case, although collec- which feathers are replaced. tion localities of molting specimens in Figure 2 A lack of quantitative molt data from most would seem to suggest that a late summer dis- other vireo speciescurrently precludes a general tribution shift has occurred such that most Gray comparison of molt patterns across the family. Vireos are concentrated in the southwesternde- However, I find no evidence to suggestthat Gray sert portion of their range similar to Bullock’s Vireos are replacing body feathers during mi- Orioles (Rohwer and Manning 1990). Instead, gratory flights, as do western Warbling-Vireos this effect is the result of large numbers of (Vireo swainsonii; Voelker and Rohwer 1998); breeding-grounds specimens having been col- nor do Gray Vireos exhibit the replacement of lected in Arizona versus other states (i.e., less outer, but not inner, primaries exhibited by than 10 specimens each from Texas, Colorado, White-eyed Vireos (K griseus; George 1973, Utah, and Nevada). Lloyd-Evans 1983). Gray Vireos replace some, One possible reason for not migrating prior to but apparently not all, of the juvenile secondary molting is that abundant, or at least sufficient, coverts during first prebasic molt, thus differing resources may be available late in the summer from Eastern Warbling-Vireo (V. gilvus), which throughout the breeding range of Gray Vireos replaces all secondary coverts, as well as differ- such that molt-migration would not provide a ing from Western Warbling-Vireo and Hutton’s benefit. This argument has been used to explain Vireo (V. huttoni) which appear to retain their why western breeding Hermit (Dendroica occi- buffy secondarycoverts (Davis 1995, Pyle 1997, dentalis) and Townsend’s Warblers (D. townsen- Voelker and Rohwer 1998). di) do not perform a molt-migration. These two species breed in montane forest habitats which TIMING OF MOLT AND MIGRATION are surely more productive in late summer than Like most other north-temperate vireos, Gray are lowland habitats west of the Rocky Moun- Vireos undergo prebasic molt on the breeding tains which several molt-migration species oc- grounds. This species does not perform a molt- cupy during the breeding season (Jackson et al. migration, that is, migrating to a nonbreeding, 1992). nonwintering area to commence prebasic molt. Another possible reason may be that, given Performing molt-migration might allow Gray the very short distance between breeding and 616 GARY VOELKJZR wintering ranges, Gray Vireos simply have suf- 1950) males may be able to begin molting while ficient time available to perform prebasic molt females rear young; however, limited evidence after breeding, and prior to migration. By com- suggests that both sexes of Gray Vireos are parison, species which migrate farther may lack equally involved in parental care (Bent 1950). sufficient time to molt on breeding grounds, and Finally, it is apparent that Gray Vireos are not are thus constrained to molt elsewhere. Gener- performing any portion of the prebasic molt on alized time-constraints hypotheses are widely wintering grounds. No specimens from this applied to trans-Saharan migration systems, al- study were replacing flight feathers after arriving though there are complex tradeoffs, such as en- on wintering grounds, nor were 60 banded birds vironmental predictability, that vary consider- that were checked for molt while on wintering ably between speciesand which result in excep- grounds (Bates 1987; S. Russell, unpubl. data). tions to generalized patterns (Jenni and Winkler This lack of prebasic molt on the wintering 1994). Because ecological comparisonsbetween grounds may be due to constraints imposed by breeding and wintering grounds are often un- winter ecology and diet. During winter, both explored (Jenni and Winkler 1994) for individ- males and females defend territories dominated ual species, unique species- or group-specific by elephant trees (Bates 1992b) (Fig. 2). Gray patterns may be going unnoticed. Vireos undergo a dietary shift from being pri- The reasonswhy Gray Vireos are not exploit- marily insectivorous on the breeding grounds ing food resourcesin the southwest are not due (Chapin 1925, Barlow 1978) to having a heavy to the general timing of the prebasic molt as reliance on elephant tree fruits during winter. compared to molt-migrating species; Gray Vir- Winter territoriality and frugivory imply several eos, and all four species that exhibit molt-mi- things. First, if the defended resource is season- gration, undergo prebasic molt from July to Sep- al, it may not pay to migrate early. And, indeed, tember-October. That Gray Vireos and molt-mi- Gray Vireos do not arrive on wintering grounds gration species do not differ in the general tim- until late September, just after Bursera micro- ing of prebasic molt raises an interesting phylla bears fruit (Bates 1992a). Second, al- question with regard to the prebasic molt dura- though Bursera fruits have a high caloric con- tions of western breeding passerine species that tent, they may be unable to meet the demands undergo molt-migration as compared to western for increased protein, iron, and calcium required breeding species which do not. Namely, does for successfulmolt (Murphy and Ring 1992). If performing molt-migration to exploit late sum- so, undergoing molt on the breeding grounds mer resources simply facilitate molt by provid- would be a better strategy for Gray Vireos. ing resources,or, can performing molt-migration Third, if birds must defend winter territories, it also increase the rate at which molt can be ac- would be reasonable to do so while not simul- complished? taneously constrained by the need to replace By comparing the prebasic molt duration of feathers. Gray Vireos to those of molt-migration species, it appears that by stopping in southwestern de- Similar to Gray Vireos, White-eyed Vireos on serts, molt-migrating species may in fact be in- the Yucatan Peninsula have a heavy reliance on creasing the rate at which prebasic molt can be Bursera fruit during winter, and both sexes vig- accomplished. Whereas Gray Vireos, as a spe- orously defend foraging territories from conspe- cies, appear to take 67 days to complete prebasic cifics as well as congeners (Greenberg et al. molt, molt-migrating speciestake only about 55 1993, 1995). This species also appears to molt days to complete the molt (54-57 days; Thomp- on its breeding grounds (Pyle 1997) and there son 1991, Young 1991, Voelker and Rohwer appears to be a close connection between the 1998). Just one of these previous studies time it spends on the Yucatan Peninsula portion (Thompson 1991) had sufficient molting speci- of its wintering grounds and the fruiting of Bur- mens to document that males and females were sera trees (Greenberg et al. 1993, 1995). Fur- molting at the same rate. It is unclear why fe- thermore, White-eyed Vireos which do not win- male Gray Vireos appear able to complete pre- ter in areas with Bursera show a heavy reliance basic molt faster than males (Table 2), at a rate on other fruits (Barlow 1980). This presentsthe similar to that of molt-migrating species. One interesting possibility that White-eyed Vireos possibility is that during second broods (Bent may also be following a strategy of molting on GRAY VIREO MOLT 617 breeding grounds while waiting for a seasonal seum of Natural History (Salt Lake City); Cornell Uni- resource. versity (Ithaca, New York); University of Arizona (Tucson); California Academy of Sciences (Oakland): A number of other vireo species,such as Phil- Museum of Comparative Zoology (Harvard Universi- adelphia (V. philudelphicus) and Yellow-throat- ty); University of California Los Angeles Dickey Col- ed (V. Jlavifrons) have a similar heavy reliance lection. I also thank those curatorswithout appropriate on fruits during winter (Skutch 1980); it is not specimensthat took the time to reply to my request. clear whether these species, like Gray and White-eyed Vireos, defend this resource. Be- LITERATURE CITED cause modem molt studies have not been done AMERICANORNITHOLOGISTS UNION.’ 1998. Check-list on these other vireo species,it is unclear wheth- of North American birds. 7th ed. American Or- er a possible connection between the timing of nithologists’ Union, Washington,DC. molt and fruiting of tropical trees can be in- BALDWIN,.I. L. 1973. Climates of the United States. National Oceanic and Atmospheric Administra- ferred. An added obfuscating factor is that, in tion, Washington, DC. eastern North America, there are abundant late- BARLOW.J. C. 1978. The effect of habitat attrition on summer food resources(insects). This food flush vireo distribution and population density in the has been hypothesized as a major factor influ- northern Chihuahuan desert region-U.S. and encing breeding ground molts in eastern birds Mexico, p, 591-596. In R. H. Wauer and D. H. Riskind [EDS.], Transactionsof the symposiumon (Voelker and Rohwer 1998). biological resourcesof the ChihuahuanDesert Re- My study presentsyet another example of the gion-united States and Mexico. Natl. Park Ser- evolutionary lability of the fall molt and migra- vice Trans. Proc. Ser. 3:1-658. tion schedulesof North American passerinesin BARLOW,J. C. 1980. Patternsof ecological interactions among migrant and resident vireos in the winter- general, and of vireos in particular. Because a ing grounds, p. 79-107. In A. Keast and E. S. phylogeny of vireos is available (Murray et al. Morton [EDS.], Migrant birds in the Neotropics: 1994), as more molt studies of vireos are con- ecology, behavior, distribution, and conservation. ducted comparisons of the molt patterns among Smithson. Inst. Press, Washington, DC. vireo species in a phylogenetic context will be BATES,J. M. 1987. The winter ecology of the Gray Vireo, Vireo vicinior, in Sonora, Mexico. M.Sc. possible; such a comparison could help clarify thesis, Univ. Arizona, Tucson, AZ. the evolutionary history of these patterns. BATES,J. M. 1992a. Frugivory on Burseramicrophylla (Burseraceae)by wintering Gray Vireos (Vireo vi- ACKNOWLEDGMENTS cinior, Vireonidae) in the coastal deserts of So- I thank Stephen Russell and John Bates for unpub- nora, Mexico. Southwest.Nat. 37:252-258. lished data and insights into Gray Vireo winter ecol- BATES,J. M. 1992b. Winter territorial behavior of Gray ogy. Chris Thompson provided a typically pithy re- Vireos. Wilson Bull. 104:425-433. view of the penultimate draft. I thank J. Bates and an BENT, A. C. 1950. Life histories of North American anonymous reviewer for insightful comments. Dave wagtails, shrikes, vireos, and their allies. U.S. Smee kindly drew Figure 2. I thank the curators and Natl. Mus. Bull. 197. staff at the following museums for loans: San Diego CHAPIN,E. A. 1925. Food habits of vireos; a family of Natural History Museum; Museum of Vertebrate Zo- insectivorousbirds. USDA Bull. 1355:1-44. ology (University of California, Berkeley); Natural DAVIS, J. N. 1995. Hutton’s Vireo (Vireo huttoni). In History Museum of Los Angeles County; Western A. Poole, l? Stettenheim, and E Gill [EDS.], The Foundation of Vertebrate Zoology; Moore Laboratory birds of North America, No. 189. The Academy of Zoology (Occidental College, Los Angeles); Denver of Natural Sciences, Philadelphia, and the Amer- Museum of Natural History; PeabodyMuseum of Nat- ican Ornithologists’ Union, Washington,DC. ural History (Yale University); Delaware Museum of GEORGE,W. G. 1973. Molt of juvenile White-eyed Vir- Natural Historv: Field Museum of Natural Historv eos. Wilson Bull. 85:327-330. (Chicago); Un&rsity of Kansas Museum of Naturil GREENBERG,R., M. S. FOSTER,AND L. MARQUEZ-VAL- History; Museum of-Natural Science, Louisiana State DELAMAR.1995. The role of the White-eyed Vireo Universitv: Universitv of Michigan Museum of Zool- in the dispersal of Bursera fruit on the Yucatan 2 ogy; American Museim of Natiral History; Bell Mu- Peninsula.J. Trouical Ecol. 11:619-639. seum (University of Minnesota); Barrick Museum GREENBERG,R., D. K: NIVEN, S. HOPP,AND C. BOONE. (University of Nevada Las Vegas); Academy of Nat- 1993. Frugivory and coexistencein a resident and ural Sciences (Philadelphia); Carnegie Museum of a migratory vireo on the Yucatan Peninsula.Con- Natural History (Pittsburgh, Pennsylvania); U.S. Na- dor 95:990-999. tional Museum of Natural History (WashingtonD.C.); HUMPHREY,F? S., AND K. C. PARKES.1959. An ap- Canadian Museum of Nature (Ottawa); Texas Coop- proach to the study of molts and plumages. Auk erative Wildlife Collection (College Station); Museum 76:1-31. of SouthwesternBiology (Albuquerque,New Mexico); JACKSON,W. M., C. S. WOOD, AND S. ROHWER.1992. Brigham Young University (Provo, Utah); Utah Mu- Age-specific plumage charactersand annual molt 618 GARY VOELKER

schedules of Hermit Warblers and Townsend’s Chuck-will’s_widow, Caprimulgus carolinensis. Warblers. Condor 94:490-50 1. Auk 88:485-519. JENNI, L., AND R. WINKLER.1994. Moult and ageing ROHWER,S. 1986. A previously unknown plumage of of Europeanpasserines. Academic Press,London. first-year Indigo Buntings and theories of delayed LANGSTON,N. E., AND S. ROHWER.1995. Unusual pat- plumage maturation. Auk 103:281-292. terns of incomplete primary molt in Laysan gnd ROHWER,S., AND J. MANNING. 1990. Differences in Black-footed Albatrosses.Condor 97:1-19. timing and number of molts for Baltimore and LANGSTON,N. E., AND S. ROHWER.1996. Molt-breed- Bullock’s Orioles: imnlications to hvbrid fitness ing tradeoffs in albatrosses:life history implica- and theories of delayed plumage ma&ration. Con- tions for big birds. Oikos 76:498-510. dor 92: 125-140. LLOYD-EVANS,T. L. 1983. Incomplete molt of juvenile RUSSELL,S. M., AND G. MONSON.1998. The birds of White-eyed Vireos in Massachusetts.J. Field Or- Sonora. Univ. Arizona Press, Tucson, AZ. nithol. 54:50-57. SKUTCH,A. E 1980. Arils as food of tropical American MURPHY,M. E., AND J. R. KING. 1992. Energy and birds. Condor 82:31-42. nutrient use during moult by White-crowned Spar- SZAREK,S. R. 1979. Primary production in four North rows Zonotrichia leucophrys gambelii. Ornis American deserts: indices of efficiency. J. Arid Stand. 23:30&3 13. Environ. 2: 187-209. MURRAY,B. W., W. B. MCGILLIVRAY,J. C. BARLOW, THOMPSON,C. W. 1991. The sequence of molts and plumagesin Painted Buntings and implications for R. N. BEECH,AND C. STROBECK.1994. The use of theories of delayed plumage maturation. Condor cytochrome b sequencevariation in estimation of 93:209-235. phylogeny in the Vireonidae. Condor 96:1037- UNDERHILL,L. G., AND W. ZUCCHINI.1988. A model 1054. for avian primary moult. Ibis 130:358-372. NIELSON,R. P 1986. High-resolution climatic analysis VOELKER,G., AND S. ROHWER.1998. Contrastsin the and southwestbiogeography. Science 232~27-34. scheduling of molt and migration in eastern and PHILLIPS,A. R. 1991. The known birds of North and western Warbling-Vireos. Auk 115:142-155. Middle America. Part II. A. R. Phillips, Denver. YOUNG,B. E. 1991. knnual molts and interruption of PIMM, S. 1976. Estimation of the duration of bird molt. the fall migration for molting in Lazuli Buntings. Condor 78550. Condor 93:236-250. PYLE,P 1997. Identification guide to North American YURI, T., AND S. ROHWER.1997. Molt and migration in birds. Part 1. Slate Creek Press, Bolinas, CA. the Northern Rough-winged Swallow. Auk 114: ROHWER,S. 1971. Molt and the annual cycle of the 249-262.