A Description of Mixed-Species Insectivorous Bird Flocks in Western Mexico’
Total Page:16
File Type:pdf, Size:1020Kb
The Condor 89~282-292 0 The Cooper Omithologml Society 1987 A DESCRIPTION OF MIXED-SPECIES INSECTIVOROUS BIRD FLOCKS IN WESTERN MEXICO’ 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 habitat 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 birds;pine-oak woodlands;tropical deciduous forests. INTRODUCTION mixed-speciesflocks in 26 sites(Appendix I) that Mixed-speciesinsectivorous bird flockshave been were distributed among various habitats 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 Oaxaca (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, Jalisco (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 Parula + + + + + + + + Blue-gray Gnatcatcher ++++++++++ ++++ + Wilson’s Warbler + + + + + + + + + + + + + + Nashville Warbler + + + + + + + + + + + + + Orange-crownedWarbler + + + ++++ +++ + + Yellow-rumped Warbler + + + + + + + Warbling Vireo ++++++ + + + + Black-throated Gray War- + + + + + bler Solitary Vireo + + + + + Black-and-white Warbler + + + Virginia’s Warbler + + Ash-throated Flycatcher + Hepatic Tanager Rufous-cappedWarbler Painted Redstart Tufted Flycatcher White-striped Wood- creeper Ladder-backed Wood- pecker White-breasted Nuthatch Bridled Titmouse Greater Pewee Ruby-crowned Ringlet Hutton’s Vireo Slate-throated Redstart Townsend’s Warbler Bushtit Empidonax sp. Spotted Wren Olive Warbler Crescent-chestedWarbler Hermit Warbler Grace’s Warbler Brown Creeper Mexican Chickadee Red-headed Tanager Red-faced Warbler Strickland’s Woodpecker Golden-browed Warbler Gray-barred Wren Red Warbler 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 woodpeckers,wood-