The Condor 89~282-292 0 The Cooper Omithologml Society 1987

A DESCRIPTION OF MIXED-SPECIES INSECTIVOROUS FLOCKS IN WESTERN ’

RICHARD L. HUTTO Department of Zoology, Universityof Montana, Missoula, MT 59812

Abstract. Insectivorousbird flockswere observed in all typesof forestedhabitats during the nonbreedingseason in westernMexico. The speciescomposition of flockschanged markedlyand predictablyamong five categoriesof type. The averagenumber of speciesper flockin lowlandhabitats was 4.7, while a mean of 18.6 speciesparticipated in highlandflocks, ranking the latter amongthe most species-richflocks in the world. The meanproportion of the localinsectivorous species that participatedin mixed-speciesflocks wassignificantly greater in the highlands(6 1.3%)than in the lowlands(24.6%). About half of the flock participantsin both undisturbedlowland and highlandhabitats were north temperatemigrants, ranking west Mexican flocks among the mostmigrant-rich in the world as well. In highlandflocks, the maximum numberof individualsper attendantspecies was generallytwo to three,but therewere often six to twelveindividuals belonging to eachof severalnuclear species. The lowlanddeciduous forest flocks seemed to lack nuclearspecies. Key words: Mixed-speciesflocks; insectivorousbirds; Mexico; migratory ;- woodlands;tropical deciduous forests.

INTRODUCTION mixed-speciesflocks in 26 sites(Appendix I) that Mixed-speciesinsectivorous bird flockshave been were distributed among various described from temperate and tropical areas throughout western Mexico. The habitat types worldwide (Rand 1954), and are known to occur that I surveyed can be roughly classified (after in practically every habitat type (Powell 1985). Pesman 1962) as belonging to either lowland Although mixed-species flocks are quite com- (tropical deciduous and tropical evergreen) or mon in north temperate regions during the non- highland (oak, pine-oak, and boreal) forests. In breeding season (Morse 1970) they generally addition, I visited one lowland and one highland reach their greatest sizes and speciesdiversities site for more extended periods either to conduct in tropical regions(Moynihan 1962, Powell 1985). more complete censusesso that the level of par- A noteworthy exception to this temperate-to- ticipation by winter residentscould be estimated, tropical region trend in flock complexity involves or to gather information about the daily pattern the highland mixed-species flocks of western of flocking behavior. These sites included the Mexico. Except for a brief note on the species pine-oak woodland at the La Michilia Interna- composition of a few flocks from (Short tional Biosphere Reserve, Durango (7 to 20 Jan- 1961), there is no published information on the uary 1984; 23”30’N, 104”15’W) and the tropical flocks of Mexico. In the present paper, I (1) de- deciduousforest at the Estacion de Biologia Cha- scribe the relationship between flock composi- mela, (21 to 28 February, 7 November tion and habitat type from locations throughout to 5 December 1984; 19”30’N, 105’03’W). western Mexico in winter, (2) describethe degree of flock participation by wintering speciesin both FLOCK SURVEYS highland and lowland habitats, and (3) compare I defined a mixed-species flock as a group of two the organization of Mexican flocksto that of flocks or more speciesoccurring within 25 m of one from elsewherein the world. another and moving in concert. These were not chance aggregations of individuals-they were STUDY SITES AND METHODS foraging groups of individuals that maintained STUDY SITES group cohesion.I typically observeda single flock I spent January through February, and July of on a given date within a survey site (Appendix 1975 and 1976 surveying the composition of I) for 1 to 2 hr to determine its speciescompo- sition, but I never moved more than about 200 I Received9 December1985. Final acceptance11 m during this period because I wanted to mea- December1986. sure the number of speciesthat were participat- WEST MEXICAN MIXED-SPECIES FLOCKS 283 ing at the same time. This constraint therefore minimized speciesaccumulations that might have accrued becauseof speciesturnover as the flock progressed through space (Austin and Smith 1972, Powell 1979, Gradwohl and Greenberg 1980). Gibb (1960) noted that it generally took an average of about a minute per individual to be fairly confident that all the specieswithin a flock had been identified, and I concur. On oc- casion, it was difficult to distinguish between a flock participant and a nonparticipant. In such instances, the decision to include or exclude an individual as a participant was based on obser- vation of that individual for several minutes, until I was able to discern whether it was moving in response to, or at least in concert with, the movement of others. This survey of flock com- positions which included a broad range of ge- ography and habitat types was, therefore, bought at a cost in terms of information on variance in flock composition within a site. Nonetheless, by grouping the flocks into habitat categories,I was able to investigate the degree to which habitat FIGURE 1. Average linkage cluster diagram of west types and flock compositions were interrelated Mexican bird flocks based on the similarity of their through cluster analysis. The clustering proce- bird speciescompositions. Geographic location of each dure involved use of the correlation coefficient flock (coded numerically) is given in Appendix I. as a measure of similarity in flock composition among sites and the average linkage algorithm (Hartigan 1981). I have used English names of sealevel to the highestforested elevations (Table the bird speciesthroughout; scientific nomencla- 1). The speciescomposition of theseflocks varied ture is provided in Appendix II. markedly among sites, but there was a clear as- THE LEVEL OF PARTICIPATION IN FLOCKS sociation between flock composition and habitat type (Fig. 1). Even though flock compositions for I walked a straight path through the forest until a site were based on the identification of mem- I encountered an individual bird and recorded bers of a single flock, the cluster analysis shows whether that individual was participating in a that the within-habitat variance in flock com- mixed-species flock (as defined above). I then position was generally much less than the be- continued on until the next individual was en- tween-habitat variance in flock composition. The countered. No individual birds were skipped or least species-richflocks were observedin the low- ignored, so my estimate of the proportion of time land tropical deciduous forests (mean = 4.7 that a given speciesparticipated in flockswas not species;n = 7) while the most diverse flockswere biased by the conscious selection or avoidance observed in the pine-oak woodlands (mean = of particular species. 18.6 species;II = 11). The mean values reported here for speciesrichness do not representthe pool RESULTS AND DISCUSSION of speciesthat might participate in such flocks, HABITAT DISTRIBUTION AND SPECIES they are the values based on numbers of species COMPOSITION OF FLOCKS actually observed foraging together within a re- Just as Moynihan (1979) and Powell (1980) have stricted space and a limited period of time (see describedfor other geographiclocations, mixed- Methods). The diversity of west Mexican flocks speciesinsectivorous bird flockswere ubiquitous is, therefore, noteworthy in light of the fact that in western Mexico. During the winter months the average number of species found to forage and in July, I found such flocks in nearly every simultaneously in mixed-species flocks is gen- habitat type (save deserts and grasslands)from erally lessthan 10, anywhere in the world (Moyn- 284 RICHARD L. HUTTO

TABLE 1. The composition of mixed-speciesinsectivorous bird flocks over a seriesof five habitat types. Each column representspresence (+)/absence (blank) information as determined from following a single flock for an extended (1 to 2 hr) period. Data were taken from western Mexican sites in winter (see Appendix I for key to geographiclocations of flocks).

Flock n,,mher Tropical evergreen forest edge Tropical deciduous forest Species I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17

Tropical + + + + + + + + Blue-gray Gnatcatcher ++++++++++ ++++ + Wilson’s + + + + + + + + + + + + + + Nashville Warbler + + + + + + + + + + + + + Orange-crownedWarbler + + + ++++ +++ + + Yellow-rumped Warbler + + + + + + + Warbling ++++++ + + + + Black-throated Gray War- + + + + + bler Solitary Vireo + + + + + Black-and-white Warbler + + + Virginia’s Warbler + + Ash-throated Flycatcher + Hepatic Rufous-cappedWarbler Painted Redstart Tufted Flycatcher White-striped Wood- creeper Ladder-backed Wood- pecker White-breasted Bridled Titmouse Greater Pewee Ruby-crowned Ringlet Hutton’s Vireo Slate-throated Redstart Townsend’s Warbler Bushtit sp. Spotted Olive Warbler Crescent-chestedWarbler Hermit Warbler Grace’s Warbler Brown Creeper Mexican Chickadee Red-headed Tanager Red-faced Warbler Strickland’s Golden-browed Warbler Gray-barred Wren Golden-crowned Ringlet

ihan 1962, 1979; Powell 1985). West Mexican The highland flocks are also unique in the level flocksare clearly extreme in the number of species of participation by the pool of available species. that participate, especially in the case of pine- Using data from Hutto (1980:table l), the mean oak woodland flocks, where ,wood- proportion of insectivorous specieswithin a giv- creepers,flycatchers, chickadees, titmice, bush- en highland site that participatedin mixed-species tits, , creepers,, , gnat- flocks(6 1.3%, N = 12) was significantly (U = 168, catchers, , , and all P i 0.00 1) greaterthan the mean proportion that participate simultaneously (Table 1). participated in lowland flocks (24.6%, n = 14). WEST MEXICAN MIXED-SPECIES FLOCKS 285

TABLE 1. Extended.

Flock number Oak woods Pine-oak woods Boreal forest 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + t + + + + + + + + + + + + + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

+ + + + +

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + t + + + + + + + + + + + + + + + + + + + + + + + + t +

In fact, no fewer than 50% of all insectivorous the pine-oak woodlandsat La Michilia, and found bird speciesparticipated in the highland habitats; that, given a landbird of any species (not just this normally included at least 95% of the small- insectivores), the probability that another bird er, warbler-sized insectivorous bird species. would be found within 25 m and moving in con- Nonparticipant speciesgenerally included those cert with the focal bird was 0.93 (Table 2). While that foraged predominantly on the ground or in it may be true that such a method biases obser- dense understory vegetation. vations toward more easily detected (flocking) Moreover, I used focal individual sampling in individuals, this level of participation in mixed- 286 RICHARD L. HUTTO

TABLE 2. For focal birds from each of five foraging rera 1979, Bell 1980). On the surface, these re- guilds, values represent the proportion of n observa- sults may seem inconsistent with the specieslists tions for which at least one other individual was found in Table 1 because it appears as if the number in associationwith a focal individual at the time the focal bird was encountered,P (group).Data are further of participant species at a given site within a subdivided into the probability that there was an as- particular habitat is far lessthan the pool of avail- sociateof another species,P (MSF) and the probability able species within that habitat. These differ- that there was another individual of the same species, ences are more a reflection of differences in the P (conspecific). numbers of bird species among geographic lo-

P cations, however, rather than a reflection of off- P P (COllSpe-and-on participation by the speciesin any one Gudd Species n (group) WF) cific) site. Only nectarivorous specieswere unlikely to Nectarivores 3 21 0.38 0.33 0.19 be observedwith associatesat La Michilia (Table Insectivores 26 301 0.96 0.95 0.58 Frugivores 8 102 0.89 0.83 0.72 2). Omnivores 2 30 1.00 0.93 0.80 Analysis of flock compositions from six of the Granivores 7 95 0.95 0.62 0.87 highland locations immediately following the All species 46 549 0.93 0.85 0.65 breeding season (July) before the return of mi- gratory speciesrevealed that an average of 53% of the winter participant specieswere north tem- perate migrants; they were absent in July (com- pare Tables 1 and 3). Moreover, this estimate is species flocks by the pool of available landbird conservative because the populations of some speciesis unusually high relative to other pub- “resident” species(determined as such because lished reports based on the same method, which of their presence in July-Warbling Vireo, Sol- indicate that most nonnuclear speciesparticipate itary Vireo, Painted Redstart, Ruby-crowned lessthan 50% of the time (Moynihan 1962, Her- Ringlet, Olive Warbler, and Grace’s Warbler),

TABLE 3. The composition of mixed-speciesinsectivorous bird flocksfrom six highland sitesthat were visited in July, before the arrival of migrants. Presence(+) or absence(blank) of a given bird specieswas determined by following a single flock for 1 to 2 hr. Geographic locations of bird flocks are given in Appendix I.

Flock number Species 38 39 40 41 42 43 Warbling Vireo Solitary Vireo + + + Hepatic Tanager + + Rufous-cappedWarbler + Painted Redstart + + Tufted Flycatcher c + White-striped Woodcreeper + Ladder-backed Woodpecker + White-breasted Nuthatch Bridled Titmouse + + Ruby-crowned Ringlet + + + Hutton’s Vireo + Slate-throated Redstart + + Bushtit + + Empidonax sp. + Spotted Wren + + Olive Warbler + Crescent-chestedWarbler + + Grace’s Warbler Brown Creeper + + Mexican Chickadee + + Golden-browed Warbler + Gray-barred Wren + Red Warbler + + Golden-crowned Kinalet t WEST MEXICAN MIXED-SPECIES FLOCKS 281

are migratory in the northern parts of their breed- also enigmatic because the time period during ing ranges. These species may, therefore, have which bird diversity is greatest coincides with been composedof both migrant and resident in- the time during which food resourcesare appar- dividuals in winter. ently at their lowest levels of the year (Hutto In the lowland forestsin July, I did not detect 1980: 193). This finding would seem to run mixed-species flocks in three sites that were re- counter to the conventional wisdom that diver- visited (sites of flocks 1, 11, and 16) because sity and resourceproductivity are directly related virtually all of the insectivorous bird speciesthat (MacArthur 1972, Brown 1973, Cody 1974). Mi- were observed to participate in flocks there in gratory specieshave often been said to arrive in winter were migrants ( 10 of the 12 lowland species time to exploit a flush of food that becomesavail- listed in Table 1). It would be unsafe to conclude able only during the winter residence period that the lowland flocks are predominated by mi- (Willis 1966, Karr 1976, Herrera 1978, Sinclair grants, however, becausemany of these lowland 1978, Feinsinger 1980, Hutto 1980, DesGranges sites were disturbed, and there is evidence that and Grant 1980). This is apparently not the case migrants are uncommonly abundant relative to for insectivores that winter in the undisturbed residentsin the disturbed lowland forestsofwest- lowland and highland habitats of western Mexico em Mexico (Hutto 1980). In November 1984, I becausethese highly integrated and cooperative revisited the undisturbed tropical deciduous for- flocking units occur at a time period during which est at Chamela, Jalisco(site of flock 17), and data the seasonallydry lowland deciduousforests lose from detailed observations of flocks there indi- their leaves, and the temperate-like highland cated that 25% of the 60 resident species and habitats experience the winter cold season. Be- 55% of the 20 migratory speciesthat were de- cause an obvious flush in productivity does not tected at the site in winter participated in mixed- coincide with the arrival of migrants, this leaves speciesinsectivorous bird flocks. Considering in- open the question of whether migrants and res- sectivorous species only, this included 50% of idents act to limit one another’s populations. the 30 resident speciesand 55% of the 18 mi- gratory species.The average flock size at Cha- ORGANIZATION OF FLOCKS IN WESTERN mela was 7.7 (n = 57) and long distance migrants MEXICO comprised at least eight, and as many as 13 of A typical daily cycle begins with the formation the 20 speciesthat commonly participated (40 of a large, cohesive flock. Flocks are apparently to 65%). Thus, in undisturbed lowland or high- not fully formed and moving as a single unit until land habitats, about half of the individuals in an hour or so after sunrise. It is as if it takes a any given flock are migrants. The proportions of while for individuals to “find” one another. After migratory species in mixed-species flocks re- OS:00 or so the flocks are as tight as they will get ported elsewhere in the literature are less than throughout the day. This is similar to the for- lo%, on average (Davis 1946, Moynihan 1962, mation of highland flocks in as de- Croxall 1976, Greig-Smith 1978, Herrera 1979, scribed by Buskirk et al. (1972) and highland Bell 1980), although Cuban (Eaton 1953) and flocks in Kashmir (MacDonald and Henderson Colombian (Chipley 1977) flocks have been re- 1977). A typical flock will not move rapidly (< 10 ported to have greater than 50% migrants. cm/set) and will be a rather loose association The unusually large, species-rich,and migrant- encompassingan area of 1,000 to 2,500 m2, de- rich flocks in the west Mexican highlands may pending on flock size. Foraging activity remains be primarily a consequenceof the fact that the high throughout the day in the highlands, but highlands experience a tremendous influx of drops off in more typical insectivore fashion northern temperate migrants which accumulate (Hutto 198 1) in the lowlands. there, rather than continuing farther south into In the undisturbed lowlands, I never observed Central and South America. Reasonsfor such an monospecificforaging groups of insectivores,and accumulation of migrants are probably associ- nuclear species- sensu Moynihan ( 1962) -were ated with the contiguity of what are essentially apparently absent. In contrast, highland flocks temperate habitats, and with the benefits accrued often consistedof a coreof passivenuclear species, by using the same habitat year-round (Hutto as evidenced by the fact that individuals of such 1985). Whatever the reason, this phenomenon speciescould occasionally be observed foraging of large and diverse flocks in western Mexico is in monospecific groups (never as solitary indi-

WEST MEXICAN MIXED-SPECIES FLOCKS 289 viduals) and, when they were part of a mixed- observesuch predictable patterns in composition speciesflock, werejoined and followed more often over a habitat gradient; the flocks are not hap- than they joined or followed others. The nuclear hazard groupings that vary widely in composi- speciesgenerally included Bridled Titmice, and/ tion through space or time. Systematic study of or Mexican Chickadees, which is similar to the banded individuals is needed to establishthe na- situation found in the pine-oak woodland flocks ture of any compositional changesassociated with of southern , as describedby Austin and particular flocks, if such changesoccur at all. Smith (1972). Given a flock, one or more of these Regularity in the absolute abundances of specieswas most always present (Table 4). Each speciesseems to be the proximate result of in- of these species,plus Bushtits, consistedof a half traspecificaggression. Such aggressionappears to dozen or more individuals, while all the remain- be common in most mixed-species flocks (Bock ing attendant speciesconsisted of no more than 1969, Morse 1970, Greig-Smith 1978, Munn and two or three individuals. The mean number of Terborgh 1979, Wiley 1980, Powell 1985). When individuals per speciesfor each of 21 attendant members of a flock meet individuals of another specieswas less than 2.0, while it was greater flock, aggressiveencounters occur between in- than 2.0 for the Bushtit, Mexican Chickadee, and dividuals of the same species.Some of these en- Bridled Titmouse (Table 4). counters result in individual birds grappling to The flock composition data taken from La the ground, as described for Bushtits by Ervin Michilia were somewhat atypical because this (1977). While limits on the number of individ- site was located far enough north to be only mar- uals per speciesseems to be set by interference ginally within the area of concentration of win- competition (through such intraspecific aggres- tering migrants. In particular, while the low sion), my observations of interspecificaggreSSiOn probabilities of coexistencewith conspecifics(as were much rarer, suggestingthat the number of shown by values in the diagonal of Table 4) ac- speciesin a flock may be set proximately by hab- curately reflect the situation at La Michilia, these itat structure (Fig. 1) and/or exploitative com- values would be considerably higher for most petition (Croxall 1976, Powell 1979) rather than attendant speciesif calculatedon the basis of data interspecific aggression. taken from farther south, where most attendant Seasonalchanges in flock composition accom- species are only rarely seen solitarily. In most panied the departureof migratory speciesin April locations, the attendant speciesconsist of pairs or May. During the breeding seasonin May and of individuals-an observation that is conspic- June, cohesive flocks probably no longer formed uous because of the close proximity of individ- on a daily basis because flocking and breeding uals of theseforaging pairs. Greenberg and Grad- activities are generally incompatible (Sedgwick wohl (1980) have described a similar situation 1949, McClure 1967, Morse 1970, Chipley 1977, for Warblers in Panama. Whether these Powell 1985). By mid-July, the resident insec- are male-female pairs, and whether these short- tivores (Table 3) foragedin mixed-species flocks lived specieshave permanent pair bonds is un- that contained about as many individuals as win- known and would be an unusual and unexpected ter flocks, and there were five to 10 individuals finding. belonging to speciesthat averaged two to three The species composition of a given flock is individuals in winter. predictable from knowledge of the habitat type From the standpoint of avian community ecol- (Fig. 1). My impression is that, although there is ogy, western Mexico appears to be unique not some turnover of individuals as the flock pro- only because of the unusually large geographic gressesthrough the smaller, individual territories and habitat overlap among temperate migrant of some species(e.g., Western Flycatcher, Wil- bird species(Hutto 1985) and the comparatively son’s Warbler, Rufous-capped Warbler, Painted high densitiesof migrants in many habitats (Hut- Redstart), the majority ofparticipants occupythe to 1980) but because of its unusually large, di- entire range traversed by the flock. If true, this verse, and migrant-rich mixed-species flocks as would be similar to the situation described in well. Peru by Munn and Terborgh (1979) where 12 specieswere found to defend a common terri- ACKNOWLEDGMENTS torial boundary. Hence, the speciescomposition I would like to dedicate this paper to the memory of is very stable and is undoubtedly the reason that Harry Bell, who offered me both stimulation through I could record one-time flock compositions and his work on Australian and New Guinean flocks, and 290 RICHARD L. HUTTO encouragement.Financial support during various phases mation of antwren flocks on Barro Colorado Is- of this study was provided by the University of Cali- land, Panama. Auk 97:385-395. fomia, World Wildlife Fund-U.S., the SmithsonianIn- GREENBERG,R., AND J. GRADWOHL. 1980. Obser- stitution, and the University of Montana. Research vations of paired Canada Warblers can- facilities at the La Michilia BiosphereReserve and the adensis during migration in Panama. Ibis 122:509- Estacion de Biologia Chamela were provided by the 512. Instituto de Ecologia and the Universidad National GREIG-SMITH,P. W. 1978. The formation, structure Autonoma de Mexico, respectively. I extend special and function of mixed-species insectivorousbird thanks to Jorge Nocedal for suggestingstudy at La flocksin west African savannawoodland. Ibis 120: Michilia, and to Sonia Gallina and Steve Bullock for 284-297. logistic SUDDO~~.I also thank Paul Hendricks and San- HARTIGAN,J. 1981. Cluster analysisof variables, p. dra Pletschet for field assistanceat La Michilia, and 448-455. In W. J. Dixon [chief ed.], BMDP sta- Jared Diamond for helpful comments on the manu- tistical software. Univ. California Press,Berkeley. script. HERRERA,C. M. 1978. Ecologicalcorrelates of resi- dence and non-residencein a Mediterranean pas- LITERATURE CITED serine bird community. J. Anim. Ecol. 47:871- 890. AUSTIN,G. T., AND E. L. SMITH. 1972. Winter for- HERRERA,C. M. 1979. Ecologicalaspects of hetero- aging ecology of mixed insectivorous bird flocks specific flock formation in a Mediterranean pas- in oak woodland in southernArizona. Condor 74: serine bird community. Oikos 33:85-96. 17-24. HUTTO, R. L. 1980. Winter habitat distribution of BELL,H. L. 1980. Composition and seasonality of migratory land birds in western Mexico, with spe- mixed-speciesfeeding flocks ofinsectivorous birds cial reference to small, foliage- insecti- in the Australian Capital Territory. Emu 80:227- vores, p. 181-203. In A. Keast and E. S. Morton 232. [eds.],Migrant birds in the Neotropics. Smithson- BOCK,C. E. 1969. Intra vs. interspecific aggression ian Institution Press, Washington, DC. in Pigmy Nuthatch flocks. Ecology 50:903-905. HUTTO, R. L. 1981. Temporal patterns of foraging BROWN,J. H. 1973. Speciesdiversity of seed-eating activity in some wood warblers in relation to the desert rodents in sand dune habitats. Ecoloav_I 54: availability of prey. Behav. Ecol. Sociobiol. 775-787. 9:195-198. BUSKIRK,W. H., G.V.N. POWELL,J. F. WIT~ENBERGER, HUTTO, R. L. 1985. Habitat selection by nonbreed- R. E. BUSKIRK,AND T. U. POWELL. 1972. Inter- ing, migratory landbirds, p. 455-476. In M. L. specific bird flocks in tropical highland Panama. Cody [ed.], Habitat selection in birds. Academic Auk 89:612-624. Press, Orlando, FL. CHIPLEY,R. M. 1977. The impact of wintering mi- KARR, J. R. 1976. On the relative abundanceof mi- grantwood warblerson residentinsectivorous pas- grants from the north temperate zone in tropical serines in a subtropical Colombian oak woods. habitats. Wilson Bull. 88:433-458. Living Bird 15:119-141. MACARTHUR, R. H. 1972. Geographical ecology. CODY,M. L. 1974. Competition and the structureof Harper and Row, New York. bird communities. Monogr. Popul. Biol. 7: l-3 18. MACDONALD,D. W., AND D. G. HENDERSON.1977. CROXALL,J. P. 1976. The composition and behavior Aspects of the behaviour and ecology of mixed- of some mixed-speciesbird flocksin Sarawak.Ibis speciesbird flocks in Kashmir. Ibis 119:48l-49 1. 118:333-346. MCCLURE,H. E. 1967. The composition of mixed DAVIS,D. E. 1946. A seasonalanalysis of mixed flocks speciesflocks in lowland and sub-montaneforests of birds in Brazil. Ecology 27: 168-l 8 1. of Malaya. Wilson Bull. 79: 130-l 54. DESGRANGES,J. L., ANDP. R. GRANT. 1980. Migrant MORSE,D. H. 1970. Ecologicalaspects of somemixed- ’ accommodation into tropical speciesforaging flocks of birds. Ecol. Monoar. 40: communities, p. 395-409. In A. Keast and E. S. i19-168. - - Morton [eds.], Migrant birds in the Neotropics. MOYNIHAN,M. 1962. The organizationand probable Smithsonian Institution Press, Washington, DC. evolution of some mixed species flocks of Neo- EATON,S. W. 1953. Wood warblerswintering in Cuba. tropicalbirds. Smithson.Misc. Collect. 143:l-140. Wilson Bull. 65: 169-l 74. MOYNIHAN,M. 1979. Geographicvariation in social ERVIN,S. 1977. Flock size, composition, and behav- behavior and in adaptationsto competition among ior in a population of Bushtits (Psaltriparus min- Andean birds. Publ. Nuttall Omithol. Club 18:l- im&. Bird-Banding 48:97-109. 162. FEINSING~R,P. 1980. Asynchronous migration pat- MUNN, C. A., AND J. W. TERBORGH.1979. Multi- terns and the coexistence of tropical humming- speciesterritoriality in Neotropical foragingflocks. birds, p. 411-419. In A. Keast and E. S. Morton Condor 81:338-347. [eds.],Migrant birds in the Neotropics. Smithson- PESMAN,M. W. 1962. Meet flora Mexicana. Dale ian Institution Press, Washington, DC. King, Publ., Globe, AZ. GIBB,J. A. 1960. Populations of tits and POWELL,G.V.N. 1979. Structure and dynamics of and their food supplyin pine plantations. Ibis 102: interspecificflocks in a Neotropical mid-elevation 163-208. forest. Auk 96:375-390. GRADWOHL,J., AND R. GREENBERG.1980. The for- POWELL,G.V.N. 1980. Migrant participation in Neo- WEST MEXICAN MIXED-SPECIES FLOCKS 291

tropical mixed speciesflocks, p. 477-483. In A. SHORT,L. L., JR, 1961. Interspeciesflocking of birds Keast and E. S. Morton [eds.], Migrant birds in of montaneforests in Oaxaca,Mexico. Wilson Bull. the Neotropics. Smithsonian Institution Press, 731341-347. Washington, DC. SINCLAIR,A.R.E. 1978. Factors affecting the food POWELL, G.V.N. 1985. Sociobiologyand the adaptive suvvlv and breedina seasonof resident birds and significanceof interspecific foraging flocks in the movements of Pa&arctic migrants in a tropical Neotropics. Omith. Monogr. No. 36, American African Savannah.Ibis 120:480-497. Omitholoaists’ Union. Washinaton. DC. WILEY,R. H. 1980. Multispecies societiesin RAND, A. L. i954. Social feedingbehavior of birds. lowland forests of Suriname and : stable Fieldiana: Zoology 36: l-7 1. membership and foraging differences. J. Zool. SEDGWICK,E. H. 1949. Mixed associationsof small (Lond.) 191:127-145. birds in the south-westof western Australia. Emu WILLIS, E. 0. 1966. The role of migrant birds at 49:9-13. swarms of army ants. Living Bird 5: 187-23 1.

APPENDIX I. Geographic locations of flocks that APPENDIX I. Continued. were observed in western Mexico. The numbers cor- respond with flock numbers given in Tables 1 and 3, and Figure 1. Flock 18 (23/2/75)-20 km south of Guadalajara, Jalisco. Flock 19 (24/2/75)- 100 km southwest of Guada- Tropical evergreen forest. I have included , plantation, secondgrowth, and tropical evergreenfor- lajara, Jalisco. est habitats under this lowland (sea level to 10 m) Flock 20 (8/3/76), Flock 38 (24/7/75)-40 km south category(see Hutto [ 19801for more complete descrip- of , . tions of these habitat types). In all cases, the flocks Flock 2 1 (13/2/76)-35 km southeastof Tepic, Na- generallymoved along the disturbededges or canopies yarit. of these habitats. Pine-oak woodland. These woodlands were generally Flock 1(4/2/75)-Ester0 de1Pozo, San Blas,Nayarit. composed of a diverse mixture of (Pinus spp.), Flock 2 (23/2/76)-2 km east of Barra de Navidad, (Quercus spp.), and occasionallymadrones (Ar- Jalisco. ’ butus sp.), all of which stood 15 m high, on average Flock 3 (11/2/76), Flock 10 (3/2/75)-2 km north (elevation range was 2,000 to 3,000 m). of Matanchen, Nayarit. Flock 22 (9/3/75), Flock 23 (29/2/76), Flock 39 (231 Flock 4 (31/l/75), Flock 5 (2/2/75), Flock 8 (28/l/ 7/75)-31 km west of Morelia, Michoacan. 76), Flock 9 (7/2/76)-plantation and second Flock 24 (5/3/75), Flock 40 (1 g/7/75)- 10 km south growth sites 1 km south of San Blas, Nayarit. of Carapan, Michoacan. Flock 6 (14/2/75), Flock 7 (6/2/76)-evergreen forest Flock 25 (19/1/75)-l km west of El Palmito, Si- edge 8 km east of San Blas, Nayarit. naloa. Tropical deciduousforest. These sitesincluded the short Flock 26 (19/2/76)- 50 km south of Puerto Vallarta, tree habitatsjust below the oak woodlandsin elevation Jalisco: ’ (sea level to 1,500 m), and were dominated by plants Flock 27 (l/3/75), Flock 28 (l/3/75), Flock 29 (281 2/75)-Volcan de Fueao, Jalisco. in the genera Acacia, Bombax, Bursera, Caesalpinia, Flock 30 (10/3/75), Flock3i (27/2/76), Flock41 (20/ Ficus, and Tabebuia. Flock 11 (19/l/76)- 135 km northwest of Mazatlan, 7/75)-44 km east of Uruapan, Michoacan. . Flock 32 (6/3/75), Flock 42 (21/7/75)-5 km south Flock 12 (4/3/75)- 1 km south of , Colima. of Patzcuaro, Michoacan. Flock 13 (25/2/75)-70 km south of Barra de Na- Borealforest. Theseincluded the high elevation(> 3,000 vidad, Nayarit. m) pine- forests composed largely of Pinus pseudo- Flock 14 (26/2/75)-20 km southofColima, Colima. strobus, Quercus laurina, and Abies religiosa. Flock 15 (24/2/75)-240 km southwest of Guada- Flock 33 (27/2/75), Flock 34 (27/2/75)-Volcan de lajara, Jalisco. Fuego, Jalisco. Flock 16 (18/l/75)-10 km west of El Palmito, Si- Flock 35 (7/3/75), Flock 36 (8/3/75), Flock 43 (221 naloa. 7/75)-58 km southeastof Morelia, Michoacan. Flock 17 (24/2/76)- 10 km southeastof Chamela, Flock 37 (8/3/75)-28 km west of Mexico City, Dis- Jalisco. trito Federal. Oak woodland. These habitatsconsisted mostly of oaks (Quercus spp.), were rarely taller than 10 m in canopy height, and occurredfrom 1,500 to 2,200 m. 292 RICHARD L. HUTTO

APPENDIX II. Common and scientificnames ofbirds mentioned in the text.

Common name Scientific name

Acorn Woodpecker Melanerpes formicivorus Yellow-bellied Sap- Sphyrapicus varius sucker Ladder-backed Wood- Picoides scalaris pecker Hairy Woodpecker Picoides villosus Strickland’s Wood- Picoides stricklandi pecker Northern Flicker Colaptes auratus White-striped Wood- Lepidocolaptes leucogas- creeper ter Tufted Flycatcher Mitrephanes phaeocercus Greater Pewee Contopus pertinax Pine Flycatcher Empidonax ajinis Empidonax sp. Empidonax sp. Ash-throated Flycatcher Myiarchus cinerascens Mexican Chickadee sclateri Bridled Titmouse Parus wollweberi Bushtit Psaltriparus minimus White-breasted Nut- Sitta carolinensis hatch Brown Creeper Certhia americana Gray-barred Wren mega- lopterus Spotted Wren Campylorhynchus gularis Golden-crowned King- Regulus satrapa let Ruby-crowned Ringlet Regulus calendula Blue-gray Gnatcatcher Polioptila caerulea Solitary Vireo Vireo solitarius Hutton’s Vireo Vireo huttoni Warbling Vireo Vireo gilvus Orange-crownedWar- celata bler Nashville Warbler Vermivora rujcapilla Virginia’s Warbler Vermivora virginiae Tropical Panda Parula pitiayumi Crescent-chestedWar- Parula superciliosa bler Yellow-rumped Warbler Dendroica coronata Black-throated Gray Dendroica nigrescens Warbler Townsend’s Warbler Dendroica townsendi Hermit Warbler Dendroica occidentalis Grace’s Warbler Dendroica graciae Black-and-white War- Mniotilta varia bler Wilson’s Warbler Wilsonia pusilla Red-faced Warbler rubrtjiions Red Warbler tuber Painted Redstart Myioborus pictus Slate-throated Redstart Myioborus miniatus Rufous-cappedWarbler ru&irons Golden-browed Warbler Basileuterus belli Olive Warbler Peucedramus taeniatus Hepatic Tanager Piranga flava Red-headed Tanager Piranga erythrocephala