Condor, 82:43-57 0 The Cooper Ornithologd Society 1980

FORAGING BEHAVIOR OF NEOTROPICAL TYRANT FLYCATCHERS

JOHN W. FITZPATRICK

ABSTRACT.-An overview of the foraging characteristics of tyrant flycatchers (Tyrannidae) is presented in three parts. First, the physical techniques of different prey capture methods are described, and a standardized nomencla- ture for these techniques is suggested. Second, ten predominant “foraging modes” are defined according to the comparative frequencies with which each prey capture method is used by different . Certain species use specialized foraging modes, in that their foraging behavior is confined to a single capture method. Other more generalized foraging modes are charac- terized by the use of several or many capture methods with similar frequency. Third, the distribution of these foraging modes within the family as a whole is summarized in a by genus outline, with genera grouped according to a recent systematic revision of the family. The resulting pattern shows that each of the three subfamilies contains certain behaviorally generalized genera as well as radiations into related, but more specialized, foraging modes.

Tyrant flycatchers (Tyrannidae) form one ways that different tyrannids combine these of the largest and most diverse families individual techniques into discrete “forag- in the world. The 375 species (Traylor 1977) ing modes,” each mode represented by a include one tenth of South ’ entire characteristic combination of prey capture land bird fauna (Meyer de Schauensee techniques. Certain foraging modes are 1966), and their range of body forms and shown to reflect behavioral specialization, ecological roles is rivaled in the New World while other modes are more generalized. I only by the Furnariidae (Feduccia 1973). conclude by reviewing the occurrence of This diversity, coupled with the ubiquity these foraging modes within each of the and numerical dominance of tyrannids major phyletic groups in the Tyrannidae, throughout the Neotropics, makes the fam- based on my own field studies, published ily ideally suited for a variety of studies re- information, and the recent systematic re- garding adaptive radiation in a continental view of the family by Traylor (1977). avian group. Keast (1972) presented a brief overview of the tyrannid radiation, and sev- STUDY SITES AND METHODS eral authors have analyzed evolution within Data discussed in this report were accumulated over selected flycatcher groups in greater depth ten months of field work in an Amazonian forest site in southeastern Peru (Cocha Cashu Biological Station, (e.g., Lanyon 1967, 1978, Smith and Vuil- Manu National Park, dept. Madre de Dios) where 61 leumier 1971, Fitzpatrick 1976). However, tyrannid species have been recorded; two months in the scarcity of basic information on tyrannid extreme northern Peru in a subtropical montane forest biology, particularly among tropical species, (Cordillera de1 Condor, dept. Cajamarca); four months has until now prohibited any comprehen- in semi-arid scrub and deciduous forest sites in north- western Venezuela (Falcon and Aragua); two months sive evaluation of the familys’ radiation. in cerrado, campo, and caatinga sites in southeastern With this paper I initiate a series of re- and east-central Brasil (Santa Barbara do Rio Pardo. ports on ecological and evolutionary trends Sao Paulo; Parque National de Brasilia; Exu, Pernam- in the Tyrannidae by describing and clas- buco). Field work was conducted between June 1974 and September 1977. I was present in southeastern sifying the foraging tactics that characterize Peru and northern Venezuela during both wet and dry the family (see also Fitzpatrick 1978). By seasons, but worked in northern Peru and in Brasil necessity the bulk of this report is descrip- only during their dry periods, from June to September. tive, my intent being to propose a standard I observed 167 flycatcher species in the wild. These by which tyrannid foraging behavior may be represent all three subfamilies (Traylor 1977) and 71 of the 88 currently recognized tyrannid genera. I gath- quantitatively described and analyzed from ered quantitative data on foraging behavior for 90 tyr- a variety of perspectives (see Discussion). annid species at the above-mentioned sites. I watched The paper is presented in three sections, flycatchers with the naked eye or through 9 x 35 power beginning with descriptions and suggested binoculars, and dictated observations into a portable tape recorder for later transcription. Foraging data nomenclature for each of the prey capture were recorded only from individuals I judged to be techniques used by flycatchers. This is fol- actively foraging, primarily during morning and late lowed by an outline of the predominant afternoon peaks of activity. Sluggishness, resting, or

1431 44 JOHN W. FITZPATRICK

RESULTS Aerial Hawk METHODS OF PREY CAPTURE Perch.to.Ground The descriptions in this section are limited to the physical acts involved in each type of prey capture by flycatchers. I present them in the context of a classification whereby each individual foraging maneuver can be assigned to one and only one category. This Ground affords a means of quantitatively comparing the foraging repertoires of various species, as discussed in the next section. Given the familys’ size and diversity, tyr- annids seem more remarkable for the for- Sally. Gleans aging techniques they do not employ than @@is- for the ones they do. Creeping along bark, scratching in leaf litter, searching among dead leaves, gaping and probing, all com- mon techniques in other sub-oscine fami- lies, are absent from the Tyrannidae. The family shows comparatively few modifica- tions on a stereotyped search-and-capture technique, characterized by stationary, 2‘ often long search periods on a perch, fol- @QG lowed by either an attempt at prey capture Perch.Glean Fruits (usually through an approach flight or “sal- ly”) or a move to a new perch. For this rea- son I stress certain, seemingly minor dis- FIGURE 1. Important prey capture techniques used tinctions between capture techniques. by tyrant flycatchers. Each capture is preceded by a These subtle differences appear to account stationary search period on the perch. for many of the morphological and ecologi- cal features of this diverse family (Fitzpat- rick 1978). The following descriptions are frequent preening were assumed to indicate less than diagrammatically summarized in Fig. 1. active foraging. I followed easily observable species for periods ranging from 15 min to one hour. At Cocha Aerial . Often referred to as Cashu I followed each of four species of large-bodied “true flycatching” (e.g., Skutch 1960, Slud tyrannine flycatchers for full days during each of five 1964), this encompasses all sallies in which months, as part of a longer-term study of their social aerial prey is pursued and captured in systems. Less conspicuous species were followed until flight. Aerial prey is located while the bird lost from sight. Each time a foraging bird stopped on a new perch, searches from an exposed perch. Prey is I attempted to record the , perch height, vege- snapped from the air after a rapid, direct tation characteristics surrounding the perch, and the flight off the perch. Especially after long sal- search time on the perch. Search times were measured lies the bird may hover, glide, or float brief- with wrist- or stop-watch, or counted by using a port- able, audible timer that clicked at one-second inter- ly at the point of capture. Escaping prey vals. For search periods ending in a foraging maneu- often are pursued with quick turns or a tum- ver, I recorded the sally type (see Results), its distance bling series of acrobatic maneuvers. Large and angle from the horizontal, and the distance moved prey items are carried to the perch for han- after the sally. If the perch was abandoned I noted the dling. No tyrannid is known to habitually distance to the new perch. In practice, not all measures could be accurately noted while following rapidly make multiple prey captures while search- moving , but I tried to include representative ing on the wing, although this does occur samples of all measures for each species. I did not at- occasionally in certain genera (e.g., Conto- tempt to distinguish between successful and unsuc- pus, Tyrannus). This distinguishes Aerial cessful sallies, as prey items could be seen and iden- tified only rarely. Hawking from true aerial foraging that char- The descriptions and data presented herein, while acterizes (Hirundinidae) and swifts intended as a review of the known foraging habits of (Apodidae), for which I agree with T. C. flycatchers, are primarily taken from my own notes. A Moermond (unpubl. data) in preferring the few published studies describe tyrannid foraging term “screening” (Emlen 1977, p. 103). movements with- enough detail to be usable in this analysis, and results from these accounts are incorpo- Perch-to-Ground Sallying. Terrestrial prey rated where they supplement my own. is located during a stationary search period FORAGING OF TYRANT FLYCATCHERS 45 on an elevated perch (exposed twigs, earth while snapping the prey. The follow-up or rock mounds, utility poles or wires, roof- flight is a continuation of the sally, carrying tops). The bird , glides, or hops to the the bird well away from its former perch. ground after spotting prey, which is picked (2) Upward Hover-. The bird up as the bird lands or, more frequently, searches from an enclosed position within while the bird stands on the ground. The the vegetation. The approach flight is either bird occasionally lands a short distance from horizontal or upward. While the approach the prey item and hops toward it before cap- may be rapid, the hovering bird does not ture. A short chase along the ground may move forward at the point of capture. The ensue. Larger prey are handled after the hover may begin well before the capture, bird returns to an elevated perch. and may last several seconds as the bird po- Ground Foraging. Prey is found while the sitions itself almost vertically under the bird stands, walks, hops, or runs on the prey item. Capture occurs with a quick ground. Prey may be picked from the vege- snap, and repeated attempts may occur dur- tation during a short jump or flight (Ground ing a single hover. The follow-up flight may Sally-gleaning) or snatched from the air be a momentary return to the former perch, (Ground Hawking), after which the bird re- but active searching usually begins only af- turns to the ground. Prey also may be ter a move to a new perch. (3) Outward picked from the ground or vegetation Striking. The bird snaps stationary prey off (Standing or Running Ground Gleaning). an exposed surface during a direct, horizon- The bird may search for prey some distance tal or downward approach flight. No hover away and rapidly to capture it. This fre- is used during prey capture and the follow- quently terminates in a short, fluttering pur- up flight usually carries the bird away from suit along the ground or into the air as prey the original perch. (4) Upward Striking. attempts to escape (Flutter-pursuit). Searching occurs among enclosed, often Perch-to-Water Sallying. Aquatic prey is dense vegetation. The approach is explo- spotted while the bird searches from an ex- sively rapid, and prey is snapped or scraped posed vantage-point over or very near the off the under-surface of leaves without hov- water. In Surface Gleaning, the approach ering. Only one capture attempt is made as flight is directed toward a position several the bird moves rapidly through to a new centimeters over the surface orev. which is oerch. picked from the water with a smooth down- Perch-gleaning. This category includes ward head motion, during a brief hover. foraging maneuvers in which stationary Rarely, prey is taken from well below the prey is taken from the substrate while the water surface during a Dive straight down bird remains perched. (1) Simple Perch- from the perch. Prey is captured in the bill, gleaning. Searching occurs between rapid and usually only the head submerges before movements through the vegetation, with the bird instantly returns to an exposed perches chosen in a variety of exposed and perch to handle prey or resume searching. enclosed situations. Prey is located only Several species have been recorded Wading with visual searching, the bill rarely being in search of aquatic prey in shallow water used for probing. Prey is removed from the (Hudson 1920: 179, Skutch 1960:356, Smith substrate with no approach flight. The bird 1971:261). may employ body movements during both Sally-gleaning. This category encompas- search and capture. Usually this involves an ses several widespread capture techniques upward or downward lean, but occasionally in which prey is picked off a substrate dur- even results in a somersault around the ing a flight from the perch. Four types of perch to retrieve prey from below. Escaping Sally-gleans are recognized here, although prey are rarely pursued. (2) Landing-and- even finer subdivisions could be made. (1) gleaning. After spotting prey too far away Outward Hover-gleaning. The bird search- for a Simple Perch-glean, the bird perches es from an exposed, usually well-lighted within reach of the prey and picks it off the site, looking primarily outward or down- substrate immediately after landing. ward. The approach flight is rapid and di- Frugivory. I place all instances of frugi- rect, and prey is snatched from an exposed vory by tyrannids into a single category, al- surface (upper side of a leaf, a twig, or weed though the actual feeding techniques vary top) during a short hover in which the bird considerably. Some individuals or species may still be moving forward. Often, during hover at fruit clusters, some strike the fruit the hover and capture, the bird crashes into and continue to a new perch, some perch the substrate, slowing down only slightly and pick at fruit without sallying, and a few 46 JOHN W. FITZPATRICK

30 techniques as that species ’ foraging mode. FORAGING MODE Foraging “specialists” use only one of the 20 ” : 224 f h sallying methods for most prey captures, while “generalists” regularly employ sev- eral of them, without specializing in any

AH P.G R S W one (see Appendix). Although intermedi- ;1.G; ates exist between these two extremes, the Tyrannidae actually are charactertized by a relatively small set of distinct foraging modes. In this section I describe these pre- dominant methods, drawing examples from the prey capture distributions of some Sally ProportIon = .29 species I have studied.

20 Besides the frequency distributions of DISTANCE f prey capture techniques, additional mea- 10 surements are useful in quantitatively de- scribing the foraging habits of a given species. Taken together, these measures 0 124 16h comprise what I call the foraging mode pro- returmto.perch :31x 20 FLIGHT file for a species. A sample of such a profile is demonstrated in Fig. 2. In Fig. 2 and be- f low, I refer to several terms that require def- 10 initions at this point. 10 20 30 40 Frequency Search Time. Nearly all tyrannids forage I 2 4 B 30 120 ft. by pausing on each perch to search for prey. FIGURE 2. Foraging Mode Profile for Ochthoecn The duration of each pause, which may pre- (=Ochthornis) littoralis, a Near-ground Generalist, in cede either a foraging maneuver or a flight Amazonian Peru. The species ’ “foraging mode” is to a new perch, is easily measured and con- shown by the frequencies with which each prey cap- ture technique is used (top of figure). Prey capture stitutes the search time on that perch. Each codes: AH = Aerial Hawk; P-G = Perch-to-ground; species has a characteristic, positively R = Running; S = Standing Ground Glean; W = Perch- skewed frequency distribution of search to-Water; H-GL = Hover-glean (Upward and Outward, times, which varies only slightly between shown by arrows); STR = Strike-(Upward and Out- foraging bouts or geographic sites (e.g., Fig. ward): PGL = Perch-glean: FR = Fruit. See text for definitions of terms. Sampie sizes (n) are shown for 2; Fitzpatrick, in press; Davies 1977). each measure. Sally Angle frequencies: hatched = .05- I have explored this important aspect of .15; stippled = .16-.30. tyrannid foraging behavior in greater detail elsewhere (Fitzpatrick 1978, in press). Sally Proportion. This measure is ex- species eat only a portion of the fruit with- pressed as the fraction of the total number out removing it from the plant. Several fea- of observed search periods that resulted in tures common to these techniques argue for an attempted prey capture, as opposed to a their inclusion into one behavioral category. give-up flight to a new perch. Species that The “prey items” are immobile and con- give up many more times than they attempt spicuous, and are available to any bird that a capture (proportions below 0.4) may be passes close enough to find them. Individ- pictured as “searchers” (MacArthur 1972), uals occasionally find fruit by being attract- while those that search at nearly every ed to other birds at fruit concentrations. perch until they sally (proportions above Whether a fruit is taken as food depends on 0.7) represent “sit-and-wait” predators. the micro-habitat and dietary preferences of Sally Distance. The distances travelled the bird species, and not on the species ’ between perch and prey form an approxi- searching behavior and sallying technique. mately log-normal distribution for every Interspecific differences in techniques of species. The median may be taken as a mea- frugivory have little functional significance sure of the average radius within which a in the Tyrannidae, and usually reflect each species customarily searches for prey. species ’ typical capture techniques during Sally Angle. The angle above or below foraging. the horizontal followed during a sally to- ward a prey item. TYRANNID FORAGING MODES Give-up Flight. During active foraging, I refer to the frequencies with which a the distances moved from one unsuccessful species uses each of the above prey capture perch to a new perch also are distributed FORAGING OF TYRANT FLYCATCHERS 47

PREY CAPTURE METHODS Foraging H.GL STR PGL AH P,G R S W t ~ ,t 4 s ~ FR N Mode SPECIALISTS Tyrannus melancholicus Aer. Hawk cinerea Xolmis velata Perch to Gr. Muscisaxicola f luviat Ground assimilis U.Strike obsolet. Perch-GI.

GENERALISTS Pitangus lictor Near Gr. swainsoni 0. H-Glean fasciatus Encl. P Hawk similis Fruit/Hawk gaimardii Fruit/U.H-GI.

FIGURE 3. The ten predominant foraging modes found in the Tyrannidae, illustrated with data from the five kinds of specialists (over 50% use of one capture technique) and five kinds of generalists. Prey capture codes as in Figure 2, with Perch-gleans separated here into Simple Perch-gleans (S) and Land-and-gleans (arrow). Foraging mode codes in Figures 3 and 4: Aer. Hawk = Aerial Hawking; Perch-to-Gr. = Perch-to-Ground; Ground = Ground Foraging; U. Strike = Upward Striking; Perch-Gl. = Perch-gleaning; Near Gr. = Near-Ground Generalist; 0. H.-Glean = Outward Hover-gleaning; Encl. P. Hawk = Enclosed Perch Hawking; Fruit/Hawk = Frugivory and Aerial Hawking; Fruit/U. H.-Gl. = Frugivory and Upward Hover-gleaning.

log-normally, with a characteristic median grounds of sky or water. A few Aerial Hawk- for each species (Flight Distance after ers (, Pyrrhomyias, certain Search in Fig. 2). and Contopus) occupy true for- Follow-up Flight. This measures how far est , where they hunt from exposed an individual moves before landing on perches near small forest openings, cliffs, a new perch following a prey capture stream margins, or in a broken canopy. Ae- (Flight Distance after Sally in Fig. 2). rial Hawkers remain still for long periods, Return-to-perch Frequency. The per- searching all angles for prey and rarely centage of sallies that are followed by a re- changing positions. Their search times and turn to the former perch. sally proportions are the highest in the Figure 3 depicts frequencies of different family (Fitzpatrick, in press). Pursuit types of prey capture techniques used by flights are fast, powerful, and on the average some sample tyrannids, illustrating most of longer than those of other equivalent-sized the foraging modes described below. The tyrannids. Many Aerial Hawkers employ a following comparisons are based on com- long, floating glide immediately preceding plete foraging mode profiles for 90 tyran- the prey capture, apparently in preparation nids, together with less complete notes and for the abrupt aerial maneuvers required to published accounts of additional species. pursue and capture an escaping insect. The numerical data for 43 representative Perch-to-Ground Sallying. One large as- genera are presented in the Appendix. semblage of open country tyrannids Aerial Hawking. Specialists in this for- (Agriornis, Xolmis, Neoxolmis, Myiother- aging mode may use the Aerial Hawk for up etes, certain Ochthoeca) consists primarily to 100% of their sallies (Appendix). Such of Perch-to-Ground specialists, exemplified species habitually use a small number of fa- in Fig. 3 by Xolmis velata. Various authors, vored perches, to which they return after summarized by Smith and Vuilleumier each sally (Fitzpatrick, in press). In most (1971), indicate that some of these species cases (e.g. Tyrannus spp., Fig. 3; also forage almost exclusively in this fashion. Contopus, Colonia, , Hir-undi- These “ground gazers” (Hudson 1920) hunt nea) these species inhabit edge or canopy in open pastures, rocky slopes, or low brush situations, where light is very bright and and scan the ground from exposed, slightly prey can be spotted against solid, light back- elevated perches. As reported by Smith and 48 JOHN W. FITZPATRICK

Vuilleumier (1971), most Perch-to-Ground in only a single gleaning technique, while Salliers frequently take some aerial prey, a others use several with equal frequency. In habit that varies with local conditions. At Fig. 3 the foraging modes of three Amazo- two Brasilian sites I found Xolmis cinerea, nian, canopy-dwelling flycatchers illustrate elsewhere described as a Perch-to-Ground the three most common modes of foliage specialist, taking more prey from the air gleaning: than from the ground (Fig. 3). In addition, (1) Outward hover-gleaning. Nearly all certain open country species that are pri- species of Myiarchus, and the related marily Aerial Hawkers (e.g., Gubernetes genera of medium-sized tyrannids (, yetapa, Knipolegus spp., , , ), habitually rubinus) make facultative use of Perch-to- use the Outward Hover-glean. Forty to 80% Ground Sallies, and occasionally forage ex- of their foraging maneuvers are of this type, clusively in this manner where locally the remainder consisting largely of other heavy concentrations of ground prey are sally-gleaning variations. These species fa- found (e.g., recently burned pastures, tidal vor leafy micro-habitats with high light in- flats, rich, wet meadows). Thus a continuum tensity, including scrub, forest openings exists in the degrees to which various and subcanopy, and river or lake margins. species use the two “sit-and-wait” foraging They choose exposed perches close to fo- modes of Aerial Hawking and Perch-to- liage, peering in all directions with a char- Ground Sallying (see Appendix). Some evo- acteristic head-bobbing motion. Prey tend lutionary consequences of this continuum to be large, including caterpillars, moths, are discussed in Fitzpatrick (1978). orthopterans, and arboreal lizards. Many of Ground Foraging. The use of the ground these generalists regularly eat fruit. as a foraging substrate has evolved indepen- (2) Upward striking. One of the most dently in three major tyrannid groups (Fitz- stereotyped foraging modes in the Tyran- patrick 1978). True ground specialists ha- nidae, Upward Striking is especially char- bitually walk or run, hopping only where it acteristic of the medium- to small-bodied is required by a broken terrain. In most species formerly comprising Hellmayrs’ cases (e.g., Muscisaxicola, Fig. 3; also Mus- (1927) subfamily Euscarthminae (see Tray- cigralla, Machetornis, nengeta) lor 1977). These include the genera Platy- the bird picks prey from open ground or rinchus, Tolmomyias (Fig. 3), , grass clumps while standing or after a short (including Idioptilon), and rel- run. Flutter-pursuits are frequently em- atives (cf. genera in Appendix). Species us- ployed by rufa and Fluvicola nen- ing this foraging mode occupy nearly every geta. Smith (in Smith and Vuilleumier foliated habitat in the Neotropics, and mi- 1971) reported occasional aerial sallies up cro-habitat subdivisions between species to 35 m long by certain species of Musci- are extremely sharply defined (Fitzpatrick, saxicola, which otherwise are pipit-like unpubl. data). Perches are chosen close to ground specialists. One aberrant genus leaves and searching is directed upward. (Corythopis) forages by walking along the Sixty to 90% of the sallies are Upward ground in the forest interior, picking prey Strikes, and the remainder mostly are Up- from leaf litter and, more frequently, Up- ward Hover-gleans. Search times are com- ward Striking off the undersides of ground paratively long, and sally proportions (around story vegetation. Search times for ground 0.3) place these species toward the “search- foragers are extremely brief, apparently re- er” end of the foraging mode spectrum. flecting their tiny searching radius at any (3) Perch-gleaning. Several groups of one stopping point (Fitzpatrick, in press). tyrannids in the subfamily Elaeniinae Pausing to search often gives way to (Traylor 1977) specialize in Perch-gleaning continuous walking, especially in Mache- (e.g. Camptostoma, Fig. 3), a foraging mode tornis and Corythopis. The more characteristic of (Vireonidae) (Machetornis rixosus) regularly follows and warblers (Parulidae) than of flycatchers. livestock in pastures, picking or sallying af- Perch-gleaners move rapidly with short ter prey under their feet, and even riding hops or flights, searching only the area they their backs while searching for can reach without sallying. They choose scrub, kicked up from the grass. canopy, or edge habitats where foliage is Foliage Gleaning. The vast majority of dense and light is bright. Most Perch-glean- tyrannids forage by moving through vege- ers are highly frugivorous during appropri- tation, pausing briefly to search on each ate seasons, and they frequently associate perch and gleaning stationary prey from with mixed species flocks in the forest can- nearby surfaces. Certain species specialize OPY. FORAGING OF TYRANT FLYCATCHERS 49

Enclosed Perch Hawking. A number of groups of genera feed on it heavily during medium- to small-bodied generalist species most or all of the year. Each group has its frequently use the Aerial Hawk while for- own repertoire of insect foraging to supple- aging inside dense vegetation. These ment frugivory, and three discrete foraging species forage by perching in relatively modes can be recognized. (1) The most fru- small openings within the foliage, where givorous appears to be Le- light intensity and dappled backgrounds are gatus leucophaius (e.g., Morton 1977). less favorable for spotting aerial prey than Many of its large-bodied tyrannine relatives are the open habitats of true Aerial Hawk- (e.g., Myiozetetes similis, Fig. 3) are forag- ers. For this reason, these species use sally- ing generalists in the truest sense: all are gleans (mostly Upward Strikes and Upward largely frugivorous, they use the Aerial Hover-gleans) at least as often as Aerial Hawk extensively, and they facultatively Hawks (Appendix). Aerial Hawking be- employ every other prey capture technique comes more frequent when they wander regularly. This group of frugivorous gener- into larger openings or move along the alists includes Empidonomus, , edges of clearings. Enclosed Perch Hawk- and Myioxetetes as well as Legatus. (2) The ers include Myiophobus (e.g., M. fasciatus; genus shows a similar repertoire, Fig. 3), , Cnemotriccus, Teren- also taking fruit extensively during much of otriccus, Myiotriccus. the year, but is more restricted to densely Certain species of use a unique foliated edge or canopy micro-habitats. Ae- variant of this foraging mode by moving rial Hawking is less important than in the through forest understory with extremely former group, and is replaced by Perch- and rapid flights, short pauses, and active wing Sally-gleaning techniques. One assemblage and tail flashing in the manner of redstarts of strictly canopy-dwelling, small-bodied (Setophaga and Myioborus). Skutch (1960) Elaenia relatives (, , interpreted these rapid, erratic flights as Ae- Tyrannulus, ; Traylor 1977) uses rial Hawks. However, my experience shows Perch-gleans and Upward Hover-gleans ex- the most common sally techniques to be clusively during insect foraging, but feeds Outward and Upward Strikes, followed by mostly on small berries, especially of mis- rapid pursuits of dislodged, escaping prey. tletoes (Loranthaceae). (3) A third highly This peculiar, poorly understood foraging frugivorous group, primarily represented by mode, characterized by very brief search the expanded genus (includes times, portrays Myiobius as a behaviorally Pipromorpha), uses the Upward Hover- aberrant member of the Tyrannidae (Fitz- glean when foraging for insects in the forest patrick 1978, in press). understory. These species feed heavily on Near-ground Generalists. This heteroge- melastome fruit where available, and are neous group includes a number of species often associated with forest understory mixed whose foraging repertoires are generalized flocks of (Formicariidae) and oven- but depend upon being close to the ground. birds (Furnariidae; Munn and Terborgh Included are many of the species in Och- 1979). thoeca (including 0. littoralis, Fig. 2), Say- Pitangus sulphuratus. The Great Kiska- ornis, Pyrocephalus, Arundinicola, Fluvi- dee, a supreme generalist, is so uniquely cola, Pitangus lictor (Fig. 3), , variable in its foraging behavior that it rep- and the cinerea species group resents its own foraging mode in this clas- (Smith 1971). These species, many of which sification. The species is widespread habitually forage near or over water, use throughout tropical Central and South most of the capture techniques described America, and shows tremendous versatility above. They typically employ moderate to throughout its distribution. I have seen it high percentages of Aerial Hawks and scavenging dead fish, hunting for live fish, Perch-to-Ground Sallies (Appendix). Perch- preying on tadpoles, , and terrestrial es are low, permitting the bird to scan the lizards, Aerial Hawking, Perch-gleaning, ground, water surface, and low vegetation. Sally-gleaning with all variations, foraging Sallies are usually outward or downward. over army (see also Slud 1964), and Home ranges are often linear along streams, feeding on a vast array of fruits, from small river banks, or lake edges. Search times, sal- melastomes up to large figs. Other investi- ly proportions, and return-to-perch frequen- gators have found the species exclusively cies vary according to body size and the Aerial Hawking (R. Bailey, pers. comm.), amount of Aerial Hawking employed. wading in shallow ponds (Hudson 1920), Partial Frugivory. No tyrannid is known preying on nestlings (Skutch to rely exclusively on fruit, but several 1960, ffrench 1976), eating refuse from gar- 50 JOHN W. FITZPATRICK

G S -- 6 fig. 4a

ELAENIINAE Sources

Phyllomyias 4 Zimmerius 4 1,2,3 1 Camptostoma 41 Phaeomyias 12 4 Suiriri 44 Tyrannulus 4 Myiopagis 41 Elaenia 4 1,2,3,5,6

Mecocerculus 47 Serpophaga 4 1,3,4,7,8 lnezia 4 4,7 Stigmatura 4 :,:,7,8 Tachuris 47,8 Culicivora 4 8 4

Mionectes 4 1,2,3,9 9 1,18 B...... 4 Pseudotriccus 10 Corythopis 4 15 41 Lophotriccus 4 1,18 Atalotriccus 4 10 Oncostoma 1,3 Hemitriccus 4 Todirostrum A 1,2,3,11,12,13 ...... 2 Cnipodectes .....,... 1 4 4 I,3 Tolmomyias 4 1,3,4 Platyrinchus 4 1,3

FIGURE 4. Characteristic foraging modes of the tyrannid genera (see text). Genera are listed in the subfamilies Elaeniinae (Fig. 4a), Fluvicolinae (Fig. 4b), and Tyranninae (Fig. 4c) according to the classification of Traylor (1977) except where noted with an asterisk (*). Foraging modes, coded as in Figure 3, are grouped into three behavioral categories in which related generalized (G) and specialized (S) modes are shown together. Black bars indicate the foraging mode most typical of each genus; stippled bars indicate less important modes for certain FORAGING OF TYRANT FLYCATCHERS 51

fig. 4b

FLUVICOLINAE Sources

Onychorhynchus 4 Myiotriccus 4 Terenotriccus 4 Myiobius 4 1,3 Myiophobus 4 1,3,4 1 Mitrephanes 1,3,17,18 Pyrrhomyias 4 *Hirundinea 4 13,14 Contopus 4 1,3,18,19,20 Empidonax 4 1,16,18- 22 Nesotriccus 26 Cnemotriccus 4

Sayornis 9 1,3,19,20 Pyrocephalus 4 8,13,20 Ochthoeca A 4,23 ...... ...... A 23 Xolmis A 4,6,8,23,24 Neoxolmis 68 ...... Agriornis ...... 4 4,8,23 Muscisaxicola 4 4,6,8,23,25 *Muscigralla 4 27 Lessonia * 4 4,6,8 Knipolegus y 4 4,8,28 Hymenops ...... 488 Fluvicola ...... 4 4,29,30 *Arundinicola iii 4 4,29 Colonia 4 1,2,3,12 ...... 4 4,8 Gubernetes ...... 3...... 4

Satrapa 3 Tumbezia ? Muscipipra 7

genera. Question marks show probable foraging modes for genera where behavioral data are lacking. Sources: A = Authors’ field data; 1: Slud (1964); 2: Haverschmidt (1968); 3: Skutch (1960); 4: Wetmore (1926); 5: Crowell (1968); 6: Crawshay (1907); 7: Smith (1971); 8: H II d son (1920); 9: Willis et al. (1978); 10: personal communication, ONeill,’ Terborgh, Willis; 11: Skutch (1930); 12: Skutch (1972); 13: Mitchell (1957); 14: Meyer de Schauensee, Phelps, and Tudor (1978); 15: 0 ni k ’i and Willis (1972); 16: Willis (1966); 17: Webster (1968); 18: Skutch (1967); 52 JOHN W. FITZPATRICK

fig. 4c

TYRANNINAE Sources

Attila Casiornis 4 Rhytipterna 4 4 Slrystes 4 Myiarchus 4 1,3,20,32 Deltarhynchus

Pitangus m ~ ...... 9 1,2,4,8,20,29,31 :“““’ Megarhynchus ...... 4 1,3,33

...... Myiozetetes ...... 4 1,2,3,34 Conopias ?!i ...... 1,12,35 ...... 9 1,3,20,38,39 Legatus ...... 4 1,2,3,12,36 Empidonomus ...... 9 4,8 Tyrannopsis ...... fil - 31,37 Tyrannus & il-4,8,20 140-4. *Machetornis 4 4,8,13,24

19: Verbeek (1975); 20: Bent (1942); 21: B eaver and Baldwin (1975); 22: Johnson (1963); 23: Smith and Vuilleu- mier (1971); 24: Gibson (1918); 25: Humphrey et al. (1970); 26: Slud (1967); 27: Short and Morony (1969); 28: Willis (1976); 29: ffrench (1976); 30: Wetmore (1972); 31: Leek (1971); 32: Lanyon (1975); 33: Skutch (1951); 34: Cracraft (1967); 35: E. T. Thomas (1979); 36: Morton (1977); 37: Smith (1966); 38: Gross (1950); 39: Ligon (1971); 40: Morton (1971); 41: Fitch (1950); 42: Hespenheide (1964); 43: Skutch (1954).

bage dumps (ffrench 1976), and even pirat-. the Tyrannidae as a whole. Ecologically, a ing from other birds (J. P. Myers, pers. foraging mode classification of the tyrannid comm.). The species appears to develop genera can be used in analyses of neotrop- short-term foraging specialities according to ical bird community compositions, as I will local and seasonal conditions of food avail- discuss in later papers. Evolutionarily, the ability. It occupies a spectrum of habitats range of foraging styles within groups of re- from desert to rain forest, and is the most lated species, and the morphologic adapta- common native passerine along the cen- tions and constraints associated with these tral boulevards of Sao Paulo, South Amer- behaviors, can indicate the paths through icas’ largest city. The ex- which the family has radiated (Fitzpatrick hibits an omnivorous foraging versatility 1978). To these ends, I present below, and more typical of the Corvidae than of a sub- in Fig. 4, a summary of the foraging modes oscine passerine. that are known to characterize each tyrannid genus. Figure 4 lumps the different foraging FORAGING CHARACTERISTICS OF modes, each of which is described above TYRANNID GENERA and exemplified in Fig. 3, into three cate- As shown above, prey capture techniques gories. From left to right in the figure these are combined in different ways by tyrannids correspond to foliage gleaners, aerial sal- to form behaviorally and ecologically dis- liers, and ground-related foragers. Within tinct foraging modes (see Appendix). It is each category, specialized foraging modes of both ecological and evolutionary interest (S) are distinguished from generalized ones to examine how these habits occur within (G) as defined above and statistically veri- FORAGING OF TYRANT FLYCATCHERS 53 fied in the Appendix. For each genus, a in the family. Mionectes (including Pipro- black bar is placed under the foraging mode morpha), Leptopogon, and a few Phyllo- that best describes the foraging habits of scartes combine frugivory with Upward that genus. Question marks indicate genera Hover-gleaning. The latter mode, together for which adequate information on foraging with Upward Striking, characterizes most techniques is lacking. References to the Phylloscartes. The entire remainder of the best available descriptions of foraging styles lineage, from Pseudotriccus through Platy- also are given for each genus in Fig. 4. rinchus, consists of extremely stereotyped As with any broad categorization of a com- Upward Strikers. The aberrant genus Cor- plex natural spectrum, Fig. 4 contains cer- ythopis, a strictly terrestrial form which tain simplifications. Most important, this uses this capture technique from the ground summary fails to show the behavioral vari- (Appendix), was tentatively placed in this ations present within some of the tyrannid group by Traylor (1977) in part because of genera. Intra-generic variability is especial- this behavior. ly prevalent among the generalists, where Fluvicolinae. Traylor (1977: 159) recog- two species often differ quantitatively in nized two fluvicoline lineages, as shown their dependence upon a given set of cap- with minor changes in Fig. 4b. Although I ture techniques. In certain cases these vari- follow Traylors’ linear order exactly, I agree ations appear to be ecologically important with Smith (1970) and Smith and Vuilleu- among the assemblages of congeneric tyr- mier (1971) that Sayornis and Pyrocephalus annids in . These behavioral represent primitive members of the ground- details require a fine-grained treatment that tyrant radiation (Sa yornis through Guber- is outside the scope of this review. In Fig. netes in Fig. 4b). Traylor included these 4, this problem is dealt with in part by two genera with the other fluvicoline lin- showing a major (black bar) and a minor eage (Onychorhynchus through Cnemotric- (stippled bar) foraging mode for certain gen- cus in Fig. 4b). The latter lineage contains era (e.g. Elaenia, Serpophaga, Phyllo- an assortment of forest-inhabiting genera scartes, and Fluvicola as shown in the ranging from generalized Enclosed Perch Appendix) in which site-to-site or species- Hawkers (Myiobius, Myiophobus, Empi- to-species variation is pronounced. The donax, Cnemotriccus etc.) to Aerial Hawk- Appendix presents supporting data for ing specialists that occupy forest borders many of the genera mentioned below. and openings (Pyrhomyias, Mitrephanes, Elaeniinae. This group contains small- to Contopus). As discussed in Fitzpatrick medium-sized, predominantly foliage- (1978) and Smith (unpubl. data), the most gleaning species. Three elaeniine lineages specialized Aerial Hawking flycatcher, Hi- are suggested in Traylors’ (1977) classifica- rundinea, appears to be closely related to tion (Fig. 4a). The Elaenia relatives (PhyZ- Pyrrhomyias as reflected in Fig. 4b. The Zomyias through Elaenia) are primarily gen- second fluvicoline lineage consists of Near- eralized frugivores. They forage heavily or Ground Generalists (Sayornis, Pyrocephal- exclusively on small berries during seasonal us, Ochthoeca, certain Knipolegus, Arun- fruiting peaks, and use Perch-gleans, Up- dinicola) and a large radiation into Perch- ward Hover-gleans, and short Aerial Hawks to-Ground specialists (Myiotheretes through during insect foraging. Four genera, two Agriornis) and pure Ground Foraging spe- of them monotypic, have specialized into a cialists (Muscisaxicola, Muscigralla, Les- Perch-gleaning foraging mode. Of the tit-ty- sonia, Hymenops, Fluvicola). As discussed rants and relatives ( through elsewhere (Fitzpatrick 1978, in press), Euscarthmus), nearly all are Perch-glean- sit-and-wait behavior of the open country ing specialists. These genera, along with Perch-to-Ground species leads also to the Camptostoma and Ornithion of the pre- evolution of certain Aerial Hawking spe- vious lineage, represent the most “warbler- cialists (Colonia, Alectrurus, Gubernetes; like” tyrant flycatchers. Only the aberrant see continuum between these modes in the Culicivora, and presumably its close rela- Appendix). Muscipipra may also belong tive Polystictus, shows more generalized here, but it must at present remain in the foraging, using Striking and Hover-gleaning small group of confusing genera, about at least as often as Perch-gleaning (see Ap- which we know too little, placed by Traylor pendix). These two peculiar genera are also (1977) at the end of the Fluvicolinae. unique in being restricted to purely open Tyranninae. Traylor (1977: 166) summa- grass habitats. The lineage Mionectes rized the strong evidence that two lineages through Platyrinchus includes the most be- comprise this subfamily (Fig. 4~). The haviorally stereotyped generic assemblage Myiarchus group (Attila through Myiar- 54 JOHN W. FITZPATRICK

&us) is a homogeneous assemblage of above, each major lineage contains assem- Outward Hover-gleaning generalists. The blages of behavioral generalists and radia- behavior of the monotypic genus Deltar- tions into related but more specialized hynchus is unknown, and its inclusion in modes. This pattern strongly suggests that this group remains tentative (Traylor certain lines of ecological radiation are pre- 1977: 166). The remaining genera comprise served among present-day tyrannid forms. a tremendously successful, heterogeneous This suggestion is supported by the patterns lineage in which nearly every foraging of morphological adaptations within each mode is represented by at least one species. group (Fitzpatrick 1978). The group includes an Outward Hover- This brief catalog of tyrannid foraging be- gleaning generalist (Conopias), Upward havior will be amended and refined as new Strikers of sorts (Myiodynastes), a Near- information accumulates. However, it is al- Ground Generalist (Pitangus Zictor), many ready an adequate foundation for a variety frugivorous generalists (especially Myioze- of analyses of the tyrannid radiation. Topics tetes and Empidonomus), at least one fru- that depend on this descriptive framework givorous specialist (Legatus; Morton 1977), include (1) the analysis of morphological Aerial Hawking specialists (Tyrannus), and, adaptations in relation to different foraging as argued in Fitzpatrick (1978), one Ground styles; (2) phylogenetic patterns of radiation Foraging specialist (Machetornis). Finally, in structure and behavior, and the use of as described earlier, this group contains the foraging characteristics as a tool in solving single, omnivorous species that appears to certsin taxonomic problems; (3) relation- use various foraging modes facultatively, ships between foraging mode and micro- and whose overall repertoire encompasses habitat characteristics, and the effects of nearly the entire range of tyrannid foraging these patterns on flycatcher community techniques (Pitangus sulphuratus). compositions; (4) the effects of foraging mode on home range and territorial habits; DISCUSSION (5) quantitative differences in the optimal searching behavior associated with differ- Traylors’ (1977) full scale revision of the ent foraging modes. These topics will be Tyrannidae is based upon features that are, explored in subsequent papers. so far as we know, largely independent of foraging characteristics. Traylor relied heavily upon the anatomical comparisons of ACKNOWLEDGMENTS Warter (1965) and _4mes (1971), supple- I am extremely grateful to the Peruvian Ministerio de menting these studies with comparisons of Agricultura, Direccibn General Forestal y de Fauna, nest characteristics and patterns. for permitting my fieldwork in the pristine splendor of the Gran Parque National de1 Manu. I also thank Man- For this reason the behavioral catalog pre- uel Plenge (Peru), the late Paul Schwartz (Venezuela), sented here is especially revealing, and it Paulo Vanzolini (), and Maria Tereza Jorge Padua strikingly supports Traylors’ conclusions. (Brasilia, Brazil) for their kind logistical assistance in Foraging modes are not randomly distrib- their countries. This paper is drawn from a doctoral dissertation ap- uted within his phylogenetic lineages. proved by the Biology Department of Princeton Uni- Rather, with the exception of the tyrannine versity. The study benefitted from encouragement and group of ecological generalists (Pitangus helpful criticism from all members of my dissertation through Machetornis in Fig. 4c), foraging committee, especially Henry S. Horn and John W. Ter- styles are remarkably restricted within each borgh. I thank them and the students and faculty in Princetons’ Ecology and Population Biology group. of the taxonomic groups hypothesized in- For sharing information on the behavior of certain fly- dependently by Traylor (1977). catchers, I thank Peter Alden, Eugene Eisenmann, Figure 4 presents each specialized for- Steve Hilty, Wesley Lanyon, John ONeill,’ Ted Parker, aging mode next to the more generalized Chris Parrish, Lester Short, W. John Smith, John Ter- borgh, Betsy Thomas, John Weske, and Edwin Willis. ones to which it is behaviorally related. Ae- I am grateful to Melvin A. Traylor and especially Hen- rial Hawkers, for example, specialize on a ry Hespenheide and Tom Sherry for criticizing an earlier capture technique used extensively by En- draft, and to Debra Moskovits for her tremendous help closed Perch Hawkers. Similarly, both both in the field and with all drafts of this paper. Perch-to-Ground and Ground foraging are Fieldwork was supported by a dissertatibn grant from the National Science Foundation, and research stereotyped specializations upon a few cap- grants from the Frank M. Chapman Memorial Fund ture techniques used by Near-Ground gen- and Princeton University. eralists. When these generalist-specialist re- lationships are overlaid onto the systematic LITERATURE CITED list of Traylor (1977) (Fig. 4) the distinctive AMES, P. L. 1971. The morphology of the syrinx in behavioral characteristics of each phyloge- passerine birds. Peabody Mus. Nat. Hist. Bull. netic line clearly emerge. As summarized 37:1-194. FORAGING OF TYRANT FLYCATCHERS 55

BEAVER, D. L., AND P. H. BALDWIN. 1975. Ecological LECK, C. F. 1971. Some spatial and temporal dimen- overlap and the problem of competition and sym- sions of foraging flights. Wilson Bull. patrv in the Western and Hammonds’ flycatchers. 83:310-311. Condor 77: 1-13. LIGON, J. D. 1971. Notes on the breeding of the Sul- BENT. A. C. 1942. Life histories of North American nhur-bellied Flvcatcher in Arizona. Condor 73:250- flycatchers, larks, swallows, and their allies. U.S. i52. Nat]. Mus. Bull. 179. MACARTHUR, R. H. 1972. Geographical ecology. Har- CRACRAFT, J. 1967. Comparative foraging behavior of per and Row, New York. Muiozetetes similis and Muiozetetes granadensis MEYER DE SCHAUENSEE, R. 1966. Species of birds of in-Costa Rica. Wilson Bull.-79: 115-116. South America with their distribution. Livingston CRAWSHAY. R. 1907. The birds of Tierra de1 Fueno. Pub]. Co.. Narberth. Pennsvlvania. Bernard Ouaritch, London. MEYER DE S&HAUENSE& R., W. H. PHELPS, JR., AND CROWELL, K: L. 1968. Competition between two G. TUDOR. 1978. A guide to the birds of Vene- West Indian flvcatchers. Elaenia. Auk 85:265-286. zuela. Princeton Univ. Press, Princeton, N.J. DAVIES, N. B. 1977. Prey ’ selection and the search MITCHELL, M. A. 1957. Observations on birds of strategy of the ( southeastern Brazil. Univ. Toronto Press, Ontario. striata): a field study on optimal foraging. Anim. MORTON, E. S. 1971. Food and migration habits of the Behav. 25:1016-1033. in Panama. Auk 88:925-926. EMLEN, J. T. 1977. Land bird communities of Grand MORTON, E. S. 1977. Intratropical migration in the Bahama Island: the structure and dynamics of an Yellow-green and . Auk avifauna. Ornithol Monogr. 24:1-129. 94:97-106. FEDUCCIA, A. 1973. Evolutionary trends in the Neo- MUNN, C. A., AND J. W. TERBORGH. 1979. Multi-spe- tropical ovenbirds and woodhewers. Ornithol. cies territoriality in Neotropical foraging flocks. Monogr. 13: l-69. Condor 81:338347. FFRENCH, R. 1976. A guide to the birds of Trinidad ONIKI, Y., AND E. 0. WILLIS. 1972. Studies of - and Tobago. Hardwood, Valley Forge. following birds north of the eastern Amazon. Acta FITCH, F. W., JR. 1950. Life history and ecology of the Amazonica 2: 127-151. Scissor-tailed Flycatcher, Muscivora forjicata. SHORT, L. L., AND J. J. MORONY. 1969. Notes on some Auk 67: 145-168. birds of central Peru. Bull. Br. Ornithol. Club FITZPATRICK, J. W. 1976. Systematics and biogeog- 89:112-115. raphy of the tyrannid genus Todirostrum and re- SKUTCH, A. F. 1930. The habits and nesting activities lated genera (Aves). Bull. Mus. Comp. Zoo]. of the Northern Tody-flycatcher in Panama (Todi- 147:435-463. rostrum cinereum). Auk 47:313-322. FITZPATRICK, J. W. 1978. Foraging behavior and SKUTCH, A. F. 1951. Life history of the Boat-billed adaptive radiation in the avian family Tyrannidae. Flycatcher. Auk 68:30-49. Ph.D. diss., Princeton University, Princeton, New SKUTCH, A. F. 1954. Life history of the Tropical King- Jersey. bird. Proc. Linn. Sot. N.Y. 63-65:21-38. FITZPATRICK, J. W. Search strategies of tyrant flycatch- SKUTCH, A. F. 1960. Life histories of Central Ameri- ers. Anim. Behav., in press. can birds. Vol. 2. Pac. Coast Avif. 34:1-593. GIBSON, E. 1918. Further ornithological notes from SKUTCH, A. F. 1967. Life histories of Central Ameri- the neighborhood of Cape San Antonio, Province can highland birds. Pub]. Nuttall Ornithol. Club 7. of Buenos Avres. I. Passeres. Ibis (Ser. 10) 6:363- SKUTCH, A. F. 1972. Studies of tropical American 415. ’ birds. Pub]. Nuttall Ornithol. Club 1G. GROSS, A. 0. 1950. Nesting of the SLUD, P. 1964. Birds of Costa Rica, distribution and in Panama. Wilson Bull. 62: 183-193. ecology. Bull. Am. Mus. Nat. Hist. 128:1-430. HAVERSCHMIDT, F. 1968. The Birds of Surinam. Oliv- SLUD, P. 1967. The birds of Cocos Island. Bull. Am. er and Boyd, Edinburgh. Mus. Nat. Hist. 134:261-296. HELLMAYR. C. E. 1927. Catalonue of birds of the SMITH, W. J. 1966. Communications and relationships Americas and adjacent islandsypart V, Tyrannidae. in the genus Tyrannus. Pub]. Nuttall Ornithol. Field Mus. Nat. Hist., Zool. Ser. 13:1-517. Club 6. HESPENHEIDE, H. A. 1964. Competition and the ge- SMITH, W. J. 1970. Courtship and territorial display- nus Tyrannus. Wilson Bull. 76:265-281. ing in the , PurocenhalusI HUDSON, W. H. 1920. Birds of La Plata. Vol. 1. E. P. rubinus. Condor 72:488491. Dutton, New York. SMITH. W. I. 1971. Behavioral characteristics of ser- HUMPHREY, P. S., D. BRIDGE, P. W. REYNOLDS, AND pophagine tyrannids. Condor 73:259-286. R. T. PETERSON. 1970. Birds of Isla Grande (Tier- SMITH, W. J., AND F. VUILLEUMIER. 1971. Evolution- ra de1 Fuego). Prelim. Smithson. Manual: U. Kan- ary relationships of some South American ground sas Mus. N%t: Hist. (distrib.). tyrants. Bull. Mus. Comp. Zoo]. 141:181-286. JOHNSON, N. K. I963. Biosystematics of sibling THOMAS, B. T. 1979. Behavior and breeding of the species of flycatchers in the Empidonax hammon- White-bearded Flycatcher (Conopias inornata). di-oberholseri-wrightii complex. Univ. Calif. Pub]. Auk 96:767-775. Zoo]. 66:79-238. TRAYLOR, M. A. 1977. A classification of the Tyrant KEAST, A. 1972. Ecological opportunities and domi- Flycatchers (Tyrannidae). Bull. Mus. Comp. Zoo]. nant families, as illustrated by the Neotropical Ty- 148:129-184. rannidae (Aves). Evol. Biol. 5:229-277. VERBEEK, N. A. M. 1975. Comparative feeding be- LANYON, W. E. 1967. Revision and probably evolution havior of three coexisting tyrannid flycatchers. of the Myiarchus flycatchers of the West Indies. Wilson Bull. 87:231-240. Bull. Am. Mus. Nat. Hist. 136:329-370. WARTER, S. 1965. The cranial osteology of the Tyr- LANYON, W. E. 1975. Behavior and generic status of anni and its taxonomic implications. Ph.D. diss., the of Peru. Wilson Bull. Louisiana State University, Baton Rouge. 87:441-455. WEBSTER, J, D. 1968. A revision of the Tufted Fly- LANYON, W. E. 1978. Revision of the Myiarchus fly- catchers of the genus Mitrephanes. Auk 85:287- catchers of South America. Bull. Am. Mus. Nat. 303. Hist. 161:427-628. WETMORE, A. 1926. Observations on the birds of Ar- 56 JOHN W. FITZPATRICK

gentina, Paraguay, Uruguay, and Chile. Bull. U.S. WILLIS, E. O., D. WECHSLER, AND Y. ONIKI. 1978. On Natl. Mus. 133: l-448. behavior and nesting of McConnells’ Flycatcher WETMORE, A. 1972. The Birds of the Republic of Pan- (Pipromor&a macconnellii): does female rejec- ama. Part 3. Smithson. Misc. Collect. 150:1-631. tion lead to male promiscuity? Auk 95: l-8. WILLIS, E. 0. 1966. The role of migrant birds at swarms of army ants. Living Bird 5:187-231. Dioision of Birds, Field Museum of Natural History, WILLIS, E. 0. 1976. Similarity of a Tyrant-flycatcher Roosevelt Road at Lake Shore Drive, Chicago, Illinois and a Silky-flycatcher: not all character conver- 60605. Accepted for publication 24 April 1979. gence is competitive mimicry. Condor 78:553.

APPENDIX. Abridged foraging mode profiles of some representative South American tyrant flycatchersa

Prey substrate: Air GIOlIlld Foliage Fruit

Aerial Perch- UP Down Up DOWII Perch Foraging mode Capture method! hawk ground Ground Watt21 hover hover stnke strike glean Fruit

Aerial Hawk/Perch-to-Ground Hirundinea ferruginea R, SJ” (24)” loo* ’ Tyrannus melancholicus M (425) 94* < 1 2 3 Colonia colonus B (53) 92* 6 Contopus sirens (winter) M (164) ss* 2 4 1 Pyrocephaalus’ rubinus M, V (249) 79* 10 1 2 1 5 < Gubernetes yetapa S, B (96) 51 18 2 3 15 11 Knipolegus lophotes B (37) 38 32 19 11 Xolmis oelata S, B (66) 35 61 2 2 2 Ground Foraging Muscigralla brevicauda P (103) 100* Muscisaxicola fluviatilis M (311) 5 1 92* < 2 1 Machetornis rixosus V, E (63) 5 8 87* Lessonia rufa c (24) 17 83* Fluvicola nengeta E (120) 11 3 73* 4 3 3 3 1 Upward Strike Platyrinchtcs coronatus M (44) 5 5 82* 8 Atalotriccus pilaris V (48) 6 2 73* 17 2 Todirostrum maculatum M (62) 6 2 63* 19 2 8 Ramphotrigon fuscicauda M (23) 13 4 65 17 Tolmomyias assimilis M (40) 15 3 63 18 Hemitriccus zosterons M (30) 10 3 63 20 Onychorhynchus coronatus M (17) 6 65 29 Myio%s ecaudatus M (34) 9 12 50 27 Corythopis torquata M (20) 30 65 FORAGING OF TYRANT FLYCATCHERS 57

APPENDIX. Continued.

Prey substrate: Air Ground Foliage Fruit

Aerial Perch- UP Down UP Down Perch Foraging mode Capture methodb: hawk ground Ground W&X hove, hover strike strike glean Fruit

Perch-glean Camptostoma obsoletum V, M (149) 1 6 2 83* 3 Suiriri suiriri 6 S (58) 7 9 79* tenuirostris v (88) 1 23 9 1 61* Serpophaga subcristata S (37) 11 14 14 11 51 Specialist ’ Alectrurus tricolor S, B (135) 45 1 10 42 Culicivora caudacuta B, S (62) 13 8 8 23 23 26 Tachuris rubrigastra P (20) 25 15 60 Near-ground Generalist Arundinicola leucocephala E (73) 41 1 22 1 10 2 21 1 Ochthoeca littoralis M (224) 29 11 17 3 15 7 8 5 2 7 Pitangus lictor M (634) 21 3 18 5 13 11 25 < 3 Fluvicola pica V, M, E (8% 14 8 31 19 7 3 9 3 4 Hover-glean Generalist Myiarchus swainsoni (winter) M (61) 2 11 46 7 25 10 Sirystes sibilator M (14) 14 14 50 21 Attila bolivianus M (17) 18 24 42 6 6 6 Sublegatus arenarum v (101) 13 10 1 33 23 12 5 1 4 Enclosed Perch Hawk Myiophobus fasciatus V, B (3% 49 11 31 Myiobius barbatus R, B (27) 37 11 44 Cnemotriccus fuscatus V (26) 27 15 19 31 Fruit/Aerial Hawk Generalist Myiozetetes similis M (881) 47 1 2 2 5 31 Fruit/Hover-glean Generalist Phaeomyias murina V, E (65) 2 20 35 Elaenia cristata M (54) 19 24 20 Myiopagis gaimardii v (75) 35 16