Search

We explored the evolution of geographic distributions in archipelagos by comparing mitochondrial DNA (mtDNA) sequences and morphometric haracters within and among conspecific populations of Adelaide's Warbler (Dendroica delaidae), Plumbeous Warbler (D. plumbea), and Olive-capped Warbler (D. pityophila). Phylogenetic reconstructions were based upon 1,455 nucleotides of protein-coding mtDNA sequence from 53 individual warblers; morphological analyses employed three external measurements from a larger number of museum specimens. Of the three taxa studied, Adelaide's Warbler occupied the broadest and most fragmented geographical distribution and exhibited the greatest interpopulation differentiation in both mtDNA and morphology. Phylogenetic analyses demonstrated that the three Adelaide's Warbler populations are each reciprocally monophyletic with the Puerto Rican lineage basal to sister clades on Barbuda and St. Lucia. Genetic distances among these populations were comparable with those between some continental species. In contrast o the mtDNA pattern, the Puerto Rican and Barbudan Adelaide's Warbler populations were most similar in morphometry. We observed considerably ess mtDNA and morphometric differentiation among populations of the two species with more restricted and less fragmented distributions, the Plumbeous Warbler of Dominica and Guadeloupe and the Olive-capped Warbler of the Bahamas and Cuba. High levels of molecular and morphological differentiation among the geographically disjunct Adelaide's Warbler populations and low differentiation in the two species with less fragmented ranges suggest hat range disjunctions indicate the long-term evolutionary independence of geographically isolated island populations. Received 21 August 1997, accepted 18 February 1998. INTERPRETATIONS OF BIOGEOGRAPHIC PATTERNS in avian taxa traditionally have relied on estimates of relationship based upon morphological characters such as size and plumage coloration. Morphological divergence among lineages is summarized in the hierarchical taxonomic classification of a group, and such classifications have often been accepted as indices of phylogenetic relationships (e.g. Ricklefs and Cox 1972). Molecular phylogenies now provide an alternative source of information on biogeographic patterns and the evolutionary processes that have molded them (Avise 1994). Species that occupy archipelagos are particularly amenable to biogeographic and phylogenetic investigations. The general distribution of a species s Address correspondence to: Smithsonian Tropical Research Institute, Unit 0948, AP0 AA 34002, USA. E-mail: ilovette@sas.upenn.edu and the boundaries of particular populations are readily defined by a taxon's presence or absence on a given island. Water barriers or unoccupied islands situated between occupied islands might retard gene flow between populations and thereby enhance their genetic subdivision. Finally, archipelagos offer the opportunity to make comparisons among taxa with different evolutionary histories superimposed upon a relatively simple common geography. To explore evolutionary and geographic patterns in one archipelago, we investigated intraspecific variation in mitochondrial DNA (mtDNA) sequences and external morphometric characters in three West Indian Dendroica species, Adelaide's Warbler (D. adelaidae), Plumbeous Warbler (D. plumbea), and Olive-capped Warbler (D. pityophila). Adelaide's Warbler has both the broadest and most fragmented distribution, with populations on Puerto Rico, Bar

For 1,442 Belgian Blue bulls performance-tested at the Centre de Sélection de la Race Blanc-Bleue Belge, nine traits were observed: height at withers at 7 mo, height at withers at 13 mo, weight at 7 mo, weight at 13 mo, average feed consumption of concentrates, average daily gain, average feed consumption of concentrates per average daily gain, average feed consumption of concentrates per mean metabolic weight, and price per kilogram of live weight. This price is based on muscle conformation and is therefore used as muscle conformation score. Restricted maximum likelihood with a derivative-free algorithm was used to estimate (co)variance components because there were different models and missing values per trait. Estimates of heritabilities were above .50 except for average feed consumption per average daily gain (.16) and average feed consumption per mean metabolic weight (.33). Estimates of genetic and phenotypic correlations between height at withers and weight traits were positive and moderate to high. Average daily gain showed a negative genetic correlation with weight at 7 mo (-.68) but had positive correlations with height at withers at 13 mo and weight at 13 mo (.22 and .43). Muscle conformation expressed as price per kilogram of live weight was related to low average feed consumption per average daily gain. Average feed consumption showed high correlations with weight at 7 mo and weight at 13 mo. Average feed consumption per average daily gain had a high negative genetic correlation with average daily gain (-.89).

Two plumage forms were observed in redpolls from Churchill, Manitoba, which correspond to the putative taxa Carduelis fiammea fiammea (dark-plumaged birds) and C. hornemanni exilipes (pale-plumaged birds). in a sample of breeding adults (n = 277), we examined whether morphometric differentiation parallels this plumage polymorphism. Males and females were analyzed independently, and age, date of capture, and status at measurement (study skin vs. live bird) were considered as covariates in univariate analyses. We used a separate phenetic analysis of plumage characters, rather than our subjective field identifi- cations, to establish the groups to be compared statistically; this ensured that mensural characters (e.g. bill shape) were not utilized as classification criteria for the present investi- gation of metric traits. in both sexes, redpoll plumage forms differed significantly in three of seven external mensural characters. Discriminant-function analyses, based on the same characters, showed that the forms can be distinguished morphometrically with great confi- dence (jackknifed estimate of correctly classified individuals was 87% in both sexes). in males and females, the distribution of discriminant scores of typical individuals and of a set of unidentified birds is bimodal. in a bivariate-reduced space of plumage and morphometric variability, pale- and dark-plumaged ASY (after-second-year) males form distinct groups. in SY (second-year) males and in females, plumage forms are not strictly distinct, but in no case was there an abundance of intermediates as predicted under the hypothesis that the forms are distinct species that frequently interbreed. Redpoll types may be specifically distinct, as has frequently been suggested, but they also may be examples of intraspecific genetic or ecophenotypic polymorphism. Experimental breeding and an assessment of mating patterns in the field are required to test these possibilities. Received 3 October 1991, accepted 5 September 1992.

Redpolls (Carduelis flammea and C. hornemanni) are well known for their extensive phenotypic variability. Various explanations have been proposed for the phenomenon, ranging from extreme variability within a single inclusive taxon, to more modest variation within frequently hybridizing or strictly isolated taxa. The predominant view favors the recognition of two largely sympatric and only slightly differentiated species: the Common Redpoll (C. flammea), comprised of the four subspecies flammea, cabaret, islandica, and rostrata; and the Hoary Redpoll (C. hornemanni) comprised of the two subspecies hornemanni and exilipes. We investigated the possible existence of sympatric redpoll forms in a large sample (n = 209) of breeding individuals from Churchill, Manitoba, where the putative forms flammea and exilipes were reported to co-occur. Groups of individuals of specific age and sex were analyzed independently. The sample was not divided a priori into flammea and exilipes subgroups. The pattern of variation within each age-and-sex group is described, and the differences between these groups are analyzed. We show that several pigmentation characters in specific age-andsex groups are not distributed normally, and that pale and dark subgroups of second-year (SY) and after-second-year (ASY) males can be identified in a principal-components analysis (PCA) of seven characters reflecting the extent of melanin and carotenoid pigmentation. The pinkand red-breasted subgroups of ASY males match almost perfectly the pale and dark groups found in the PCA analysis. Further, when ASY males of different breast color (pink vs. red) are analyzed independently, all plumage characters are distributed normally. The analyses of SY and ASY females failed to provide good evidence for a plumage polymorphism. Our results support the idea that two relatively distinct redpoll forms breed at Churchill, which differ in several plumage characteristics involving both melanin and carotenoid pigmentations. These forms are more clearly differentiated in males than in females. They may be specifically distinct (C. f. flammea and C. hornemanni exilipes), as has frequently been suggested, but they also could be the product of different types of intraspecific genetic or ecophenotypic polymorphisms. Preliminary observations are presented that support the idea that plumage variants are largely genetically determined, but experimental work is needed to confirm this suggestion. Received 28 May 1991, accepted 19 February 1992. ORNITHOLOGISTS HAVE long been intrigued by the extensive plumage variability seen in redpolls (Carduelis [flammea] spp.). Based roughly on the same type of data and analytical approach (i.e. colorimetric and morphometric studies), different investigators have described that variation as continuous (e.g. Harris et al. 1965, Troy 1985) or dimorphic (e.g. Todd 1963, Molau 1985, Knox 1988), and have divided the "complex" into one to four species (see Knox 1988:table 1). The taxonomic disagreement comes mainly from the different interpretations given to the origin of birds having intermediate attributes in sympatric populations. Some re1 Present addresses: Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA; and Smithsonian Tropical Research Institute, Unit 0948, APO AA 34002, USA. searchers hold that intermediate birds are common in many populations and represent either hybrids and backcross individuals (e.g. Williamson 1961, Harris et al. 1965), or merely central points in a homogeneous distribution of phenotypic variants within a single highly variable species (Troy 1985). Other investigators (e.g. Molau 1985, Knox 1988) have suggested that, when sexual and age dimorphisms are carefully taken into account, very few redpolls qualify as intermediate, and these probably are convergent variants of distinct sympatric species. The most commonly encountered classification of redpolls is one in which two polytypic species are recognized, the Common Redpoll (C. flammea) comprised of the subspecies cabaret, flammea, and rostrata, and the Hoary Redpoll (C. hornemanni) comprised of the subspecies exilipes and hornemanni (Knox 1988). The taxonomic affinities of redpolls breeding in Iceland are un

eaters (Meropidae) and cooperative breeding in hot-climate birds. Ibis 114:1-14. GIROUX, J. F. 1985. Nest sites and superclutches of American Avocets on artificial islands. Can. J. Zool. 63:1302-1305. GROVES, S. 1984. Chick growth, sibling rivalry, and chick production in American Black Oystercatchers. Auk 101:525-531. LAURO, B., AND J. BURGER. 1989. Nest-site selection of American Oystercatchers (Haematopus palliatus) in salt-marshes. Auk 106:185-192. MAYR, E. 1939. The sex ratio in wild birds. Am. Nat. 73:156-179. NOL, E. 1985. Sex roles in the American Oystercatcher. Behaviour 95:232-260. NOL, E. 1989. Food supply and reproductive performance of the American Oystercatcher in Virginia. Condor 91:429-435. NOL, E., A. J. BAKER, AND M. D. CADMAN. 1984. Clutch initiation dates, clutch size, and egg size of the American Oystercatcher in Virginia. Auk 101:855-867.

Females of many avian species are able to produce more or less elaborate songs. In most passerines this behavior seems to be occasional or exceptional, and the quality of the songs produced is usually poor. Only in a few species is singing a regular feature of female behavior (e.g. Armstrong 1963, Van Tyne and Berger 1976, Ritchison 1983). The occurrence of singing by female Willow Flycatchers (Empidonax traillii) is reported here. I studied the territorial behavior of sympatric Alder (E. alnorum) and Willow flycatchers in June and July 1985 near Brighton, Northumberland Co., Ontario (44?05'N, 77?45'W), and near Montreal, Quebec (45?20'N, 73?30'W). Alder Flycatchers from allopatric populations were studied near Lac-Carre, Terrebonne Co., Quebec (46?10'N, 74?30'W), and near Jonquiere, Chicoutimi Co., Quebec (48?30'N, 71?15'W). Birds were identified by song, and playback experiments were conducted with the original intent of investigating their vocal responses toward both Alder and Willow flycatcher songs. Two playback tapes were prepared. The tape used to test Alder Flycatchers contained a series of Willow Flycatcher songs, lasting 90 s (about 20 songs); after a pause of 30 s, a series of Alder Flycatcher songs lasting 13 min followed. The tape used to test Willow Flycatchers was constructed in the same pattern, but it contained a series of Alder Flycatcher songs followed by a series of Willow Flycatcher songs. The three advertising songs of Willow Flycatchers were used on both tapes, arranged in the following sequence: 3 "fitz-bews," 1 "fizz-bew," and 1 "creet." In this paper, all onomatopoeias referring to specific songs and calls are taken from Stein (1963). The tapes were played over the speaker of a Uher 4000 or a Sony recorder directed toward the test bird. The recorder was placed 3-5 m from the bird's song perch, and the observer was located 2-3 m farther away. Playback experiments usually lasted 15 min but were stopped whenever the test bird repeatedly directed aggressive behavior toward another flycatcher. No experiment lasted less than 5 min. After playback, birds that engaged in aggressive response were collected and sexed by gonadal inspection. Specimens are deposited as study skins in the National Museum of Natural Sciences, Ottawa, under Accession No. 1986-7. Twenty-one Willow Flycatchers were collected, four of which were singing females. All females had fully developed ovaries ranging in size from 5 x 5 to 8 x 6 mm. No female songs were recorded, but I judged them to have the same pitch, pattern of syllables, and duration as male songs. Each female responded differently to playback experiments. Female 22-02 delivered only "pit" notes and 2 faint "fizz-bews" during the 10 min in which conspecific songs were played. The other three females responded more strongly to conspecific songs. Female 21-06 appeared to be paired. Before the male was collected, this female was perched in dense alder bushes about 3 m from the speaker giving "pits" and a few "churrs." Immediately after the male was collected she flew as close as 50 cm from the speaker and gave 7 loud "fitz-bews" and 3 "fizz-bews." Playback continued for 5 min, during which time she gave numerous "pit," "zweeoo," and "churr" calls. In two instances she gave a combination of "churrweeoo," recalling the "fee-bee-o" song of the Alder Flycatcher. Stein (1963: 45) similarly noted that Alder Flycatchers can produce a combination of calls that resemble the Willow Flycatcher "fitz-bew." Before the end of the experiment she uttered 5 more "fitzbew" songs. Female 25-01 was accompanied by another individual (presumably her mate) during the entire playback experiment. Both birds came repeatedly as close as 50 cm from the speaker, and each sang the three advertising songs of the species as well as most of the call notes. Female 22-05 was approached while singing at the top of a 2.5-m-high willow bush. Before playback started, she was singing "fitz-bews" at a rate of 6-10 songs/min. She responded strongly to playback of conspecific songs by giving "pit," "weeoo," and "churr" calls, as well as many "fitz-bews" and "creets." This female had a fully developed ovary (8 x 6 mm), and the oviduct contained a 10 x 10-mm ovum. No other birds were seen or heard in the vicinity during the 10 min of playback. I did not observe any singing by female Alder Flycatchers. A total of 45 singing individuals, none of which was a female, was collected from four different populations. Kroodsma (1984) noted that newly fledged female Willow Flycatchers are able to give vocalizations resembling the adult "creet." He further demonstrated, by testosterone implantation, that adult females can produce all three advertising songs of the species. My observations indicate that female Willow Flycatchers also sing under natural conditions. Kroodsma (1984) showed that female Alder Flycatchers also can be artificially induced to sing, but my observations do not support the idea that singing occurs in nature. Females of a fair number of tyrannid species are involved in duetting. Farabaugh (1982) listed 16 for