Habitat use and foraging behavior of the Chestnut-sided Warbler (Dendroica pensylvanica), Canada Warbler (Wilsonia canadensis), Mourning Warbler (Oporornis philadelphia) and Nashville Warbler (Vermivora ruficapilla) were studied in a second-growth forest in northern Wisconsin (USA) during the breeding season. Our objective was to determine if interspecific and intersexual foraging differences exist among these species. The warblers concentrated foraging activities on a few species of trees and shrubs. Compared to other species, Canada Warblers more frequently used coniferous trees that were scattered through the primarily deciduous habitat. All species most frequently used branches of trees, but male Nashville Warblers also made considerable use of leaves and twigs. Warblers foraged at a variety of heights in the vegetation, with male Nashville Warblers foraging highest in the midstory and canopy of mature trees. Females of all species tended to be active lower in the vegetation than males. Gleaning was the most frequently used foraging method, except for male Canada Warblers, which used foraging methods involving flight in about 70% of prey pursuits. When 30 foraging-related characters were considered simultaneously in a multivariate analysis, it was found that intersexual differences within a species were less pronounced than interspecific differences, except Nashville Warbler males, which were distinct in their foraging. When designing a conservation strategy for these species, biologists must recognize that these birds require heterogenous second-growth forests.

Department of Biological Sciences Univeq'sity of Albeq'ta Edmonton, Albeq'ta T6G 2E9, Canada UTILIZACION DEL I-[Jd31TAT Y CONDUCTA DE FORRAJEO DE CUATRO ESPECIES DE PARf_JLIDOS EN UN BOSQUE SECUNDARIO Sinopsis.--Se estudi6, en un bosque secundario al norte de Wisconsin, el uso de hfbitat y la conducta de forrajeo de Dendroica pensylvanica, Wilsonia canadensis, Oporornis philadel- phia y Ib'mivora ruficapilla, durante le &poca reproductiva de dichas aves. E1 objetivo fue determinar si habla diferencias interespecificas e intersexuales entre las diferentes especies. Las aves concentraron su actividad de forrajeo en un grupo reducido de especies de firboles y arbustos. En comparaci6n a otros estudios Wilsonia utiliz6, con mayor frecuencia, conlferos que estaban esparcidos a trav6s de un hfbitat de deciduos. Todas las especies utilizaron con mayor frecuencia las ramas de los frboles, aunque los machos de Vermivora utilizaron con- siderablemente alas hojas y alas ramitas. La altura de forrajeo fue variable con Vermivora utilizando la parte mrs alta del docel y el estrato medio de la vegetaci6n madura. A menor altura en la vegetaci6n, las hembras de todas las especies tendieron a estar m/rs activas que los machos. E1 rebusque result6 ser el m6todo de forrajeo utilizado con mayor frecuencia, a excepci6n de los machos de Wilsonia que utilizaron m6todos asociados al vuelo en el 70% de los casos. Cuando se consideraron en una anflisis de nmltivarianza, simultfneamente, 30 caracteres relacionados al forrajeo, se encontr6 que las diferencias intersexuales dentro de una especie eran menos pronunciadas que las diferencias interespecificas, excepto para los machos de Vernfivora que fueron distintos en su forrajeo. Cuando se planifique disefiar una estrategia de conservaci6n para estas especies, los bi61ogos deben reconocer que estas aves requieren de bosques secundarios heterog6neos. Current address: Laboratory of Wildlife Conservation, National Institute for Environmental Studies, Onogawa 16-2, 7]ukuba, Ibaraki 305, Japan. This study describes the habitat use and foraging behavior of four wood warblers (Emberizidae, Parulinae), the Chestnut-sided Warbler (Dendroica pensylvanica), Canada Warbler (Wilsonia canadensis), Mourning Warbler (Oporornis philadelphia) and Nashville Warbler (Vermivora ruficapilla) on their breeding grounds in the upper Midwest, USA. Relatively little is known about the foraging ecology of these four species. Our objectives were to determine how the four species differ in their use of habitat elements and in their foraging patterns, and to compare behavior be- tween the sexes within species. Numerous previous studies have reported interspecific or intersexual foraging differences among co-occurring wood warblers (e.g., Holmes 1986, MacArthur 1958, Morse 1968). The ultimate causes of these differences, and the role of processes such as competition for food, remain unresolved (Martin and Karr 1990, Morri- son 1981, Price 1991). The species considered here characteristically breed in heterogeneous second-growth forests that are mixtures of deciduous and coniferous tree species. Bird populations using some early successional stages may be more vulnerable to decline than previously realized (Hagan 1993, Martin 1993). Information on the ecological requirements of species using early successional stages is thus needed so that effective management decisions can be made regarding these birds. METHODS Data were collected in an area centered at the University of Wisconsin Trout Lake Station and the adjacent Mann Creek Wildlife Area, Vilas County, Wisconsin (46ø01'N, 89ø40'W) between 6 Jun. and 27 Jul. 1977. The area supported second-growth forest, which had been logged and/ or burned repeatedly during the previous 100 yr. Stands were a mixture of aspen, northern hardwoods and conifers, that correspond to Curtis's (1959) "northern dry-mesic forest" (for further description see Pasz- kowski 1984). Data were collected from 3-10 individuals of each sex of each species. Sexes of all these species show some plumage differences and can be distinguished in the field. Total observation times (in hours) were as fol- lows: Chestnut-sided Warbler male = 9.5, female = 1.8; Canada Warbler male = 1.3, female = 0.2; Mourning Warbler male = 3.0, female = 0.9; Nashville Warbler male = 0.3, female = 0.2. Data were collected by locating an individual and composing a timed, continuous record of the bird's movement until it was lost (maximum 10 min). We recorded: (1) plant species or genus (plants which could not be rapidly identified to species, e.g., Ac Alnus, Corylus, Populus and Prunus) where a bird searched for and/or captured prey, (2) foraging height (visually estimated), (3) substrate used during locomotion (leaf, twig [wood (1 cm diameter], branch [1.0-2.5 cm diameter], limb [2.6- 12 cm diameter], trunk [)12 cm diameter], ground and air), (4) direc- tion moved relative to the trunk or center of a plant (up, down, towards and away), (5) flight distances (to a new substrate or plant; if direction changed dramatically while a bird was flying, the move was considered a new flight) and (6) foraging method used for prey capture. We recog- nized the following foraging methods: (1) glean (picked prey from sub- strate while perched), (2) pounce (flew towards prey located on a sub- strate, alighted on substrate while capturing prey), (3) hover (picked prey from substrate while in flight without landing) and (4) sally (captured prey in mid-air while in flight). It was not always possible to determine if a foraging maneuver resulted in a prey capture, as many prey were small (e.g., Diptera), thus foraging methods encompass both successful and unsuccessful capture attempts. From our continuous records we extracted a number of other variables for analysis. We defined a "sequence" as a series of moves that began when a bird moved to a new tree or shrub and ended when it left that plant. We used sequences that were observed completely to calculate movement rate (number of hops and flights/s; cf. Robinson and Holmes 1982). As heights within a movement sequence can be correlated (Sherry 1975), we used only the height of the first move within a sequence when analyzing foraging height. Visual inspection of data revealed that warblers usually did not change their foraging heights appreciably during a for- aging sequence. We used chi-squared tests of independence to compare the distribution of movement sequences among vegetation types, the distribution of moves among substrates, and the frequency of foraging behaviors across warbler species. Data from males and females were handled separately. For these tests, plants were pooled into four categories: "shrubs" (alder, Alnus crispa and rugosa; hazelnut, Corylus americana and cornuta), "hard- woods" (black oak, Quercus borealix cherry, Prunus virginiana, serotina and pensylvanica; paper birch, Betula papyrifera; and maple, Acer sacchar- um and rubrum), "aspens" (Populus tremuloides and grandidentata), and "conifers" (white pine, Pinus strobus;, red pine, P.. resinosa; jack pine, P.. banksiana; black spruce, Picea mariana; and balsam fir, Abies balsamea). Substrate categories were also combined (twig with leaves, and limb with trunk). Movement rates and heights, along with flight distances, were compared for males and for females of the four species using one-way analysis of variance (ANOVA) coupled with Tukey's test. To compare foraging behavior and habitat use between sexes and among species simultaneously, we performed cluster analysis on a data matrix consisting of eight rows (species x sexes) and 30 characteristics (columns), following Holmes et al. (1979). Analysis used the unweighted pair-group clustering method with arithmetic averages (UPGMA) and Eu- clidian distance coefficients (NTSYS-pc; Rohlf 1987). Twenty-seven vari- ables were percentages that were log transformed (log0[x + 1]) to re- duce skewness. These were: (a) percentage use of four foraging methods, (b) percentage of moves in four directions, (c) percentage of moves on seven substrates and (d) percentage of movement sequences in 12 plant species/genera. Specific categories were the same as described for uni- variate analyses. The remaining three variables were: mean foraging height, standard error of foraging heights and sex-specific body masses (obtained from Dunning 1984). All columns were standardized by sub- tracting the column mean from each value in the column (see Holmes et al. 1979). Our data are non-independent and may be autocorrelated (Hurlbert 1984). Such data may give inaccurate estimates of variance and should be interpreted with care (Hejl et al. 1990). RESULTS Males and females of each warbler species concentrated their activities in relatively few of the available tree and shrub species. The five woody plants most frequently used by all warbler species were: trembling/big- toothed aspen, sugar/red maple, paper birch, hazelnut and white pine. When woody plant species were combined into four categories, interspe- cific comparisons revealed significant differences in the use of these veg- etation types among males (X = 80.4, df = 9, P < 0.001) and among females (X' = 28.5, df = 9, P < 0.001) of the four warbler species (Fig. 1). Male Chestnut-sided and Nashville Warblers foraged more frequently in hardwood trees than the other species, whereas male Canada and Mourning Warblers foraged more frequently in aspens. In addition, male Canada Warblers made greater use of conifers than the other species. Female Canada Warblers also foraged more frequently in conifers than females of the other species (Fig. 1). Female Chestnut-sided, Mourning, Canada and Nashville Warblers foraged frequently in aspen trees, shrubs, aspen trees and hardwood trees, respectively. All species were active at a wide range of heights in the vegetation (Table 1), but mean foraging heights differed significantly among species for both males and females (F = 38.9, df = 3,521, P < 0.001; F = 3.60, df = 3,187, P = 0.015, respectively). For males, all species differed sig- nificantly from each other in foraging heights, with Nashville > Chestnut- sided > Canada > Mourning Warbler (Tukey pair-wise comparisons; P 0.05). For females, Nashville Warblers foraged significantly higher than Chestnut-sided or Mourning Warblers (Tukey pair-wise comparisons; P 0.05). All warblers travelled most frequently by hopping along branches (Fig. 2). Comparisons across species, however, revealed significant differences in substrate use for males (X' = 78.1, df = 9, P < 0.001) and females = 52.1, df = 9, P < 0.001). Nashville Warbler males used twigs and leaves more frequently than any other species. Chestnut-sided Warbler females made minimal use of tree limbs and trunks (Fig. 2). Flights were relatively rare, accounting for <20% of observed moves for each sex and species (Fig. 2, "air"). The were also short, typically <1 m (Table 1). Flight distance varied significantly among male warblers (F = 4.34, df = 3,685, P = 0.005), but not among females (F = 0.39, df = 3,244, P = 0.76). Male Chestnut-sided Warblers made significantly longer flights than male Mourning Warblers (P < 0.05, Tukey test). Frequency of movement differed significantly among species for both A 20 o Shrub Hardwood Aspen D Chestnut-sided male rT1 Canada male F Mourning male E Nashville male Conifer D Chestnut-sided female [[1 Canada female ['q Mourning female E Nashville female Shrub Hardwood Aspen Conifer Vegetation FIGURE 1. Vegetation use by fbur warbler species in a second-growth fbrest (% occurrence of movement sequences). Total numbers of sequences used in the analysis were: Chest- nut-sided male = 267, female = 107; Canada Warbler male = 129, female = 39; Mourn- ing Warbler male = 92, female = 30; Nash411e Warbler male = 31, female = 16. males and females (F = 58.3, df = 3,648; F = 9.03, df = 3,223, respec- tively, P < 0.001; Table 1). Among males, movement rates differed sig- nificantly between all pair-wise combinations (P < 0.05, Tukey test) ex- cept the species with the highest rates, Canada and Nashville Warblers. Similarly, female Canada and Nashville Warblers had high movement TABLE 1. Activity heights, flight distances and movement rates of male and female warblers recorded during the breeding season in a second-growth forest. Values reported as mean + SE. Range and sample sizes in parentheses. Movement Movement Flight frequency Warbler height distances (hops and species Sex (m) (m) flights/s) Chestnut-sided M 5.47 _ 0.16 0.98 4-_ 0.05 (0.3-11.7; 269) (0.3-10.0; 334) F 2.59 _ 0.23 1.01 _+ 0.06 (0.7-12.0; 103) (0.3-6.7; 172) Canada M 4.06 _+ 0.25 0.98 _ 0.04 (0.3-23.0; 122) (0.3-4.0; 210) F 3.18 _ 0.27 0.85 _+ 0.10 (1.0-8.3; 39) (0.3-2.7; 33) Mourning M 3.07 _+ 0.26 0.74 _+ 0.06 (0.3-11.0; 105) (0.3-5.0; 118) F 2.68 4-_ 0.46 0.90 _+ 0.18 (0.7-10.0; 32) (0.3-6.7; 32) Nashville M 8.22 - 0.57 0.57 _ 0.05 (1.3-13.0; 29) (0.3-1.0; 27) F 4.49 ñ 0.74 0.91 ñ 0.19 (1.7-11.7; 17) (0.3-2.7; 11) 0.13 _+ 0.01 (0.0-0.9; 375) 0.20 _+ 0.01 (0.0-1.0; 150) 0.40 _ 0.03 (0.0-2.0; 150) 0.33 q- 0.04 0.0-0.9; 30) 0.23 _+ 0.02 0.0-1.0; 101) 0.20 _+ 0.03 0.0-6.7; 34) 0.38 _ 0.06 (0.0-1.2; 26) 0.36 ñ 0.07 (0.1-1.0; 13) rates and differed significantly (P < 0.05, Tukey test) froIn Chestnut-sided and Mourning Warbler females which had low movement rates. Overall, gleaning was the most commonly observed foraging inethod except in male Canada Warblers (Fig. 3). Males of the four species dif- fered significantly in their use of gleaning versus non-gleaning behaviors that required flight (X = 37.6, df = 9, P < 0.001). Male Nashville War- biers gleaned more frequently and male Canada Warblers less frequently than the other species. Females of three species (Mourning Warbler could not be included because of low sample size) did not differ significantly in use of gleaning versus inethods involving flight (X ' = 5.86, df = 6, P > 0.10). Cluster analysis, which integrated detailed measures of habitat use and foraging patterns, indicated that for Chestnut-sided, Canada and Mourn- ing Warblers, males and females were more similar to each other in their behavior than to either sex of the other species (Fig. 4). Male Nashville Warblers proved to be the most distinct of all sex-species categories, and females of this species were more similar to males and females of the other three species than to conspecific males. DISCUSSION The predominance of activity in Populus species and understory ele- ments, such as Corylus, reflected the fact that warblers in our study were using relatively young, disturbed forests which characteristically have high A Tw & Le B r Li & Tr -"D Chestnut-sided male ITI Canada male I1 Mourning male I1 Nashville male B 80 0 4O 0 ß Tw & Le B r Li & Tr '[] Chestnut-sided female ITI Canada female F":I Mourning female I Nashville female Substrate FIGURE 2. Substrates used by four warbler species in a second-growth forest (% of moves observed). Tw & Le = twig and leaves, Br = Branch, Li & Tr = limb and trunk, and Ai = Air. Total numbers of movements used in the analysis were: Chestnut-sided Warbler male = 1189, female = 695; Canada Warbler male = 598, female = 93; Mourning Warbler male = 729, female = 260; Nash511e Warbler male = 142, female = 103. light penetration and a well-developed layer of shrubs and herbs. Al- though this habitat was quite heterogenous, males and females of the four warbler species did not use a particularly wide variety of the shrub and tree species present. The Ganada Warbler was the most distinctive of the 100 20 Glean :D Chestnut-aided male ITI Canada male I Mourning male I1 Naahville male Pounce Hover Sally 100- 60- 20 Glean Pounce Hover Foraging Method :':'D Chestnut-aided female ITI Canada female ] Mourning female I1 Naahville female Sally FIGURE 3. Foraging methods used by four warbler species in a second-growth forest. Total numbers of prey capture attempts observed were: Chestnut-sided male = 116, female = 47; Canada Warbler male = 89, female = 30; Mourning Warbler male = 30, female = 6; Nashville Warbler male = 34, Ismale = 11. four as it foraged in coniferous trees scattered through a forest dominated by deciduous species (Fig. 1). Mourning Warblers frequently moved along wider branches, limbs and trunks, where they walked (alternating footfalls) rather than hopped (si- multaneous footfalls). In contrast, male Nashville Warblers often captured prey while perched on flexible, unstable surfaces such as leaves and twigs 9.00 7.50 Warbler Forag Euclidean Distance 6.00 ] 4.50 [285 3.00 [ CHM CHF CAM CAF MOM MOF NAF NAM FIGURE 4. Cluster dendrogram showing intersexual and interspecific relationships of four warbler species in a second-growth forest, based on muhivariate analyses of habitat use and foraging behavior. CHM = Chesmut-sided Warbler male; CHF = Chesmut-sided Warbler female; CAM = Canada Warbler male; CAF -- Canada Warbler female; MOM = Mourning Warbler male; MOF = Mourning Warbler female; NAM = Nashville War- bler male; and NAF = Nashville Warbler Female. (Fig. 2). The ability to exploit these foraging locations may be an outcome of the small body size of the Nashville Warbler; males on average weigh 8.9 g (vs. 9.8-13.0 g for the other species; Dunning 1984). Airola and Barrett (1985) also found that Nashville Warblers foraged more frequent- ly on foliage than Yellow-rumped Warbler (D. cororata) or Hermit War- bler (D. occidentalis) and depended almost exclusively on gleaning. Flight was used moderately by the other species to capture prey. Similar to the results of Martin and Karr (1990), we found Chestnut-sided War- blers used flight in about 50% of prey capture attempts. Tramer and Kemp (1980) reported that Wilson's Warbler (147../9usilla), a congener of the Canada Warbler, used flight in about 40% of foraging maneuvers. We observed that male and female Canada Warblers flew in 70% and 43% of their prey capture attempts, respectively. Movement rates differed among species for both sexes, with Nashville and Canada Warblers displaying higher rates than Chestnut-sided and Mourning Warblers (Table 1). What ultimate or proximate causes were responsible for these differences is difficult to determine. Vegetation structure could be a factor limiting speed, as could morphology (Moer- mond 1990). Both factors could also contribute to interspecific differ- ences in foraging methods (Robinson and Holmes 1984). Our cluster analysis indicated that males and females were most similar to each other in their foraging patterns for Chestnut-sided, Canada and Mourning Warblers. Morse (1971) and Holmes (1986) likewise reported that foraging behavior of the parulid warbler species that they studied were most similar between males and females of the same species. Our study indicated that male and female Nashville Warblers were the most divergent from each other and from the other species. We had the least data on the Nashville Warbler, but propose that it was truly distinctive because of its use of the overstory created by mature trees of aspen and birch. Although males and females of three of the species examined strongly resembled each other in behavior (Chestnut-sided, Canada and Mourn- ing Warblers), the sexes were by no means identical in their foraging. A number of hypotheses have been put forward to explain why such differ- ences occur between the sexes in passefine birds. It has been suggested that differences are related to the center of foraging activity during the breeding season (space-restrained foraging hypothesis; Morse 1968, 1989). Males forage at greater heights where they can alternate foraging with singing to communicate effectively inter- and intrasexually. Females forage in lower strata close to their nests, thereby saving energy while travelling between the nest and foraging locations (Morse 1968, Holmes 1986). Foraging heights in males and females are, therefore, correlated with singing perch heights and nest heights, respectively. As reported for other parulid warblers (Busby and Sealy 1979, Franzfeb 1983, Morrison 1982, Morse 1968), males of the four species examined here were active on average 0.4-3.7 m higher than females. Nests of all four species are typically on or near the ground (Ehrlich et al. 1988). Morse (1968) argued that female warblers forage faster than males be- cause they: (1) need supplemental energy to produce eggs and (2) are time-constrained as they must alternate incubation/brooding with forag- ing. The patterns of sex-based differences we observed did not support this argument, as only Chestnut-sided Warbler females moved faster than males. Rand (1952) and Selander (1966) suggested that sexual differences in foraging niche occur in response to intersexual competition. Consistent with this hypothesis, Peters and Grubb (1983) and Destochefs (1989) have shown in wintering Downy Woodpeckers (Picoides pubescens) and Black-capped Chickadees (Parus atricapillus), respectively, that females avoid microhabitats used by dominant males. Contradicting this hypoth- esis, Morton et al. (1987) found that habitat use of non-breeding female Hooded Warblers (W. citrina) did not change when males were experi- mentally removed. Morse (1989) suggested that the degree of sexual difference in foraging niche is related to the intensity of both intra- and interspecific competi- tion. He found that male and female warblers foraged more similarly on islands with low numbers of competitors (both intra- and interspecific) when compared to warblers on the mainland where the number of po- tential competitors was high (Morse 1971). We did not document diet composition or food abundance, thus the role of resource competition in shaping observed intersexual and interspecific differences in foraging patterns and habitat use cannot be addressed in a meaningful fashion. In summary, the four species of warblers displayed interspecific and intersexual differences in behavior, yet males and females obviously co- exist on the same breeding territory and species coexist in the same hab- itat. In designing a conservation strategy for this suite of birds, biologists must recognize that co-occurring species require different elements pro- vided by the heterogeneous environment of second-growth forest. ACKNO%rLEDGMENTS Thanks to the staff of the Trout Lake Research Station for use of this excellent facility. E. A. Nash and W. M. Tonn helped with data analyses. We also thank R. T. Holmes, J. D. Parrish, K. Yasukawa and an anonymous reviewer for making comments on an earlier draftß LITERATURE CITED AIROLA, D. A., AND R. H. BARRETT. 1985. Foraging and habitat relationships of insect-glean- ing birds in a Sierra Nevada mixed-conifer forestß Condor 87:205-216. BusBy, D. G., )d,D S. G. SL3/4. 1979. Feeding ecology of a population of nesting yellow warblersß can. J. Zool. 57:1670-1681. CURTIS, J. T. 1959. Vegetation of Wisconsinß Univ. of Wisconsin Press, Madison, Wisconsin. 657 pp. DESROCHERS, A. 1989. Sex, dominance, and habitat use in wintering black-capped chicka- dees: a field experiment. Ecology 70:636-645. DUNNING, J. B., JR. 1984. Body weights of 686 species of North American birds. Western Bird Banding Assoc., Monogr. 1, Cave Creek, Arizona. EHPICH, P. R., D. S. DOBKIN, ^ND D. WHEYS. 1988. The birder's handbook. Simon and Schuster Inc., New York, New York. 785 pp. FRNZPO, K. E. 1983. Intersexual habitat partitioning in Yellow-rumped Warblers during the breeding seasonß Wilson Bull. 95:581-590. HAGS, J. M., III. 1993. Decline of the Rufous-sided Towhee in the eastern United States. Auk 110:863-874ß H[I L, S. J., J. V;PO4ER, .aND G. W. BELL. 1990. Sequential versus initial observations in studies of avian foraging. Stud. Avian Biol. 13:166-173. HOLM[S, R. T. 1986. Foraging patterns of forest birds: male-female differences. Wilson Bull. 98:196-213ß , R. E. BONN[Y, JR., ^ND S. W. P^C^LA. 1979. Guild structure of the Hubbard Brook bird community: a multivariate approach. Ecology 60:512-520ß HURLBERT, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecol. Monogr. 54:187-211. MACARTHUR, R. H. 1958. Population ecology of some warblers of northeastern conifer for- ests. Ecology 39:599-619. MATN, T. E. 1993. Nest predation among vegetation layers and habitat types: revising the dogmas. Am. Nat. 141:897-913. , )d,D J. R. KPO. 1990. Behavioral plasticity of foraging maneuvers of migratory war- biers: multiple selection period for nichesß Stud. Avian Biol. 13:353-359ß MO[tX4OND, T. C. 1990. A functional approach to foraging: morphology, behavior, and the capacity to exploitß Stud. Avian Biol. 13:427-430. MomusoN, M. L. 1981. The structure of western warbler assemblages: analysis of foraging behavior and habitat selection in Oregon. Auk 98:578-588ß ß 1982. The structure of western warbler assemblages: ecomorphological analysis of the Black-throated Gray and Hermit Warblers. Auk 99:503-513ß MORSE, D. H. 1968ß A quantitative study of foraging male and female spruce-woods warblersß Ecology 49:779-784. 1971. The foraging of warblers isolated on small islands. Ecology 52:216-228. ß 1989. American warblers. Harvard Univ. Press, Cambridge, Massachusetts. 406 pp. MORTON, E. S.,J. F. L7qCH, K YOUNG, AND P. MEHLHOP. 1987. Do Hooded Warblers exclude females from nonbreeding territories in tropic forest? Auk 104:133-135. PASZKOWSKI, G. A. 1984. Macrohabitat use, microhabitat use, and foraging behavior of the Hermit Thrush and Veery in a northern Wisconsin forest. Wilson Bull. 96:286-292. PETERS, W. D., AND T. C. GRUBB, JR. 1983. An experimental analysis of sex-specific foraging in the downy woodpecker, Picoides pubescerts. Ecology 64:1437-1443. PRIC[, T. 1991. Morphology and ecology of breeding warblers along an altitude gradient in Kashmir, India. J. Anita. Ecol. 60:643-664. RAND, A. L. 1952. Secondary sexual characters and ecological competition. Fieldiana Zool. 34:65-70ß ROBINSON, S. K., AND R. T. HOLMES. 1982. Foraging behavior of forest birds: the relationships among search tactics, diet, and habitat structure. Ecology 63:1918-1931. , AND '. 1984. Effects of plant species and tbliage structure on the foraging behavior of forest birds. Auk 101:672-684. ROHLF, F. L. 1987. Numerical taxonomy and multivariate analysis system. Exeter Publishing, Setauket, New York. 37 pp. SELANDR, R. K. 1966. Sexual dimorphism and differential niche utilization in birds. Condor 68:113-151. SHRY, T. W. 1975. Foraging behavior and niche relationships of breeding American red- starts (Setophaga ruticilla) and least flycatchers (Empidonax minimus): the flycatcher guild in a New England hardwoods forest. M.Sc. thesis, Dartmouth College, Hanover, New Hampshire. TRAMER, E.J., AND T. R. KEMP. 1980. Foraging ecolog 3, of migrant and resident warblers and vireos in the highlands of Costa Rica. Pp. 285-296, in A. Keast and E. S. Morton, eds. Migrant birds in the neotropics: ecology, behavior, distribution, and conservation. Smith- sonian Institution Press, Washington, D.C. Received 15 Jun. 1994; accepted 26 Aug. 1994.