habitat use and conservation of migratory landbirds overwintering in the neotropics

~-~c --~

fp II./S-'17 I!) i\i\Ov.f;;(); T ~ D. NI. FI~ {e;ds.) C;C-(>I,J -I Ma,.,tl7..f;1(.eP

f)f t\J~f,rZIf'c.dl M111""1t..-y b1rd~ .r!J.)(~d U()IV. r~..s~) fJy. (f.:r45) .

6

HABIT A T USE AND CONSERV A TION IN

THE NEOTROPICS

DANIEL R. PETIT, JAMES F. L YNCH, RICHARD L. HunG, JOHN G. BLAKE,

AND ROBERT B. w AIDE

INTRODUCTION et al. 1989a), migration (Stewart 1987,Gauthreaux 1992, Russell et al. 1992), and

Habitat, the vegetative, physical, and winter periods (Faaborg and Arendt 1992).topographic features associated with the Two main alternative hypotheses, bothlocation in which an animal lives (Oduro habitat based, have been advanced to explain1971), has important ramifications for the population declines of long-distance migra-persistence of a species because it can tory species in recent years (see Askins et al.influence reproductive and mortality rates. 1990, Wilcove and Robinson 1990): (1)The evolutionary significance of habitat habitat loss and fragmentation on theselection is clearly reflected by the close breeding grounds have led to increasedassociation between habitat use and mor- predation and cowbird parasitism due tophology, behavior, and life-history traits of edge effects, and to decreased opportunitiesa species (Cody 1985). Understanding habitat to breed because of shortages of suit-requirements has motivated much of the able habitat (Briggs and Criswell 1978,study of avian ecology during this century Brittingharo and Temple 1983, Wilcoveand, consequently, habitat studies form the 1985); and (2) deforestation in Neotropicalframework for conservation efforts aimed at wintering areas has forced some individualspreserving species diversity (Cody 1985, to occupy marginal habitats, which causesProbst and Crow 1991). Wildlife ecologists increased mortality (Rappole and Mortonagree that conservation of populations of 1985, Rappole et al. 1989, Robbins et al.wild animals depends upon protection of 1989b). These two alternatives are nothabitats at appropriate spatial scales (Harris mutually exclusive. However, partitioningand Kangas 1988, Grumbine 1990). This breeding season effects from those during theprospect grows increasingly difficult, how- boreal winter (or during migration; Mooreever, due to encroachment of civilization into and Simons 1992, Moore et al., chapter 5,even the most remote ecosystems on earth. this volume) is essential for understanding

Migratory birds that breed in North population declines and, subsequently, forAmerica and overwinter in the Neotropics taking action to reverse those trends.have been affected by human alteration of Presently, most evidence implicates eventsnatural landscapes (Robbins et al. 1989a, during the breeding season (Gates and GyselSauer and Droege 1992). Declines in 1978, Wilcove 1985, Yahner and Scott 1988,populations of many Neotropical migrants Gibbs and Faaborg 1990, Wilcove andhave been documented through long-terro Robinson 1990, Martin 1992; but seestudies at continental, regional (Breeding Rappole and McDonald 1994) in populationBird Survey of the US Fish and Wildlife declines of migrants, but this may beService; Robbins et al. 1986), and local misleading because research efforts have been(Ambuel and Temple 1982, Wilcove 1988, skewed toward North American breedingJohnston and Hagan 1992) scales during grounds. Evidence of an effect of tropicalbreeding (Briggs and Criswell1978, Robbins deforestation is largely circumstantial, and is

145

146 TEMPORAL PERSPECTIVES ON POPULATION liMITATION AND HABITAT USE

based upon concurrent population declines conserve, and manage habitats for migratoryof migrants and conversion of tropical birds in the Neotropics.broadleaved forests to other land uses (Hut to1988a). Nevertheless, the overwinteringperiod could be a time of intense selective HISTORICAL VIEWS OF HABITAT USE ANDpressure on Neotropical migrants because of GEOGRAPHIC DISTRIBUTIONmortality associated with stress induced fromlong migratory flights and increased competi- Behavioral ecology of single species (e.g.,tion for food due to inflated densities of Schwartz 1964) or groups of species (e.g.,potential competitors (Morse 1980). Because Moynihan 1962, Willis 1966, Leck 1972a)these pressures should be exacerbated in the dominated the study of overwinteringface of widespread destruction of tropical Neotropical migratory birds during itshabitats, a conceptual basis exists for claims emergence in the 1960s. As a result,that events on wintering areas also contribute few published quantitative descriptions ofsubstantially to many of the observed habitat use by overwintering migrants (e.g.,population trends (Robbins et al. 1989a). Willis 1966, Lack and Lack 1972, Tramer

Knowledge of habitat requirements of 1974) were available when Karr (1976)species has greatly facilitated the ways in presented the first overview of migrantwhich conservation biologists and wildlife habitat use in the New World Tropics. Karrmanagers attempt to sustain viable popula- (1976, p. 456) relied upon several publishedtions of long-distance migratory birds. studies, reports from other ornithologists,Clearly, documentation of habitat relation- and his own work, to conclude thatships is critical but it is not the only type of migratory birds as a group were foundinformation that is necessary to formulate predominantly in "disturbed, transitory, oreffective conservation plans. To understand isolated patches of habitat," and thatwhy some species are more susceptible to interiors of mature, moist forests andanthropogenic disturbances than others, a grassland/savanna vegetation were avoidedbetter comprehension of the ecological and by migrants (also see Monroe 1970, Russellevolutionary factors governing habitat 1970). In addition, Karr noted that lowselection is required; e.g., the relative roles of ( < 1000 m) and high ( > 2500 m) elevationsfood resources, predators, and competitors, supported fewer migratory birds than middleas well as the spatial scale upon which habitat elevations (1000-2500 m), and that mainlandselection is based. This mechanistic approach areas sustained fewer migrants than islands.facilitates the decoupling of complex Those generalities remained intact throughrelationships and helps isolate causal from the scrutiny afforded by a host of studies thatspurious associations. Furthermore, be- followed several years later in a Smithsonianhavioral patterns of species may provide an Institution publication devoted to theimportant link between habitat use and ecology of migrant birds in the Neotropicshabitat suitability (Hut to 1990), and may (Keast and Morton 1980). Based upon thoseenhance the predictive capabilities for studies, Terborgh (1980) concluded thatestimating the effects of certain land use migrants were most abundant (1) at higherpractices on migratory birds. tropical latitudes (see also Keast 1980), and

This chapter provides a general overview (2) in disturbed vegetation types. Theof habitat use by long-distance migratory presumption that migrants showed anbirds in the Neotropics and examines sources affinity toward middle elevations wasof variation influencing those patterns. In generally supported, but varied somewhataddition, it addresses the importance of among geographic regions (Terborgh 1980).integrating multiple subdisciplines of con- Despite publication of numerous studies ofservation biology, for example, behavioral winter habitat use by migrants sinceand landscape ecology, into the process by Terborgh's (1980) compendium, littlewhich appropriate conservation and land-use attempt has been made to summarizepractices are determined. Finally, recom- available data and reassess the abovemendations are offered of how to best define, paradigms. Initial identification of areas

WINTER HABITAT USE AND CONSERVATION 147

most appropriate for designation as pro- atively, using only data that weretected reserves can be facilitated through collected by visual/auditory techniquesapplication of information on broad patterns biases against densely vegetated habitatsof species habitat use and geographic dis- of short stature because of the difficultytribution (Margules and Usher 1981, of detecting individuals that do notDiamond 1985). Below, a detailed examina- vocalize in those habitats comparedtion of winter habitat use by Neotropical with birds in taller, more open forests.migratory birds is used to identify those For example, in young second growth,habitats and geographic areas that are most Blake and Loiselle (1992) captured 31critical in sustaining diversity of migratory migratory species in mist nests, butspecies. detected only 17 species via visual/

auditory surveys. In nearby primaryforest, however, visual (13 species) and

HABITAT AND GEOGRAPHIC DISTRIBUTION mist nest (15 species) surveys producedOF MIGRATORY BIRDS IN THE NEOTROPICS nearly identical numbers of migratory

..species. Thus, results presented belowDat~ were cO~plIed from al?ProxImately 30 should be viewed cautiously because ofp.ubIIshe~ stu~Ies (represen~Ing > 200 st.udy the probable conservative estimates of~ttes) to Ide~tIfy those ~abttats a?d regIons species richness in young, denselyIn the tropIC~ supportIng. relatI:Vely !arge vegetated habitats. Nevertheless, in-numbers of mIgr~tory species dunng wIn~er. clusion of mist-netting results does not°!1ly. those s.tudIes that met the followIng substantially change the broad con-cntena were Included. clusions reached through these analyses

1. The study was conducted largely or (see Petit et al. 1993).entirely during the overwintering 4. Data were collected within a singleperiod (as opposed to migration). This vegetation type. Surveys conductedrestriction was necessary because over several habitats and presented asmany species are more generalized in a single value were not used because ofhabitat use along migratory routes the need to ascertain habitat-specificcompared to the overwintering (or abundances of migratory birds.breeding) period (Petit et al. 1993).Habitat selection during migration Several types of information were ex-needs to be examined separately from tracted from each study, including: (1) theother periods (Moore et al. 1993). percentage of all species surveyed that were

2. Study sites were located north of the Neotropical migrants, and (2) the totalTropic of Capricorn (23.5° S) and number of migratory species detected.south of the Tropic of Cancer (23.5° N), Absolute densities of migrants were notwith the exception of one study located possible to obtain because most surveyin Florida. This area represents the techniques used in the tropics are notgeographic limits of the Neotropics and amenable to estimation of densities. Thus, theis the region within which most number of migratory species detected onmigratory landbirds overwinter. study sites was used as an index of the

3. Data on habitat use were gathered importance of habitat types or regions tothrough formal visual/auditory surveys migratory birds. The significance of certain(e.g., point counts or transects). Simple habitats or regions to migratory birdsspecies lists usually were excluded from compared to those used by resident speciesanalyses. Estimates of species richness was evaluated by examining the relativeacquired through mist netting were not representation of migratory birds on eachincluded in these analyses because, in study site (i.e., percentage of all species thatsome taller habitats, mist-netting may were Neotropical migrants). Because residentprovide a severely biased appraisal of and migratory birds may require distinctthe bird community (Karr 1981, conservation a~d management actions (PetitMorton and Greenberg 1989). Altern- et al. 1993), this latter analysis can help

148 TEMPORAL PERSPECTIVES ON POPULATION LIMITATION AND HABITAT USE

identify those habitats that merit special relationship between percentage of migratoryattention for one group or the other. The individuals and the percentage of migratorynumber of migratory species and the species on a site was robust (r = 0.779,percentage of migratory species on a site were p < 0.001, n = 159).significantly related (Pearson's r = 0.559, Habitat and topographic descriptions andp < 0.001, n = 185). geographical information of each site were

The percentage of all individuals on a site based upon descriptions provided by thethat were migrants was also derived from author or from other published sources. Thateach study but this parameter was not information was used to assess the effects of:included in analyses because: (1) estimates of (1) habitat type; (2) vegetation height; (3)abundance in field studies are more tenuous level of disturbance; (4) elevation; (5)than determination of simple presence/ insularization; and (6) latitude on theabsence; (2) a close mathematical relation- distribution of migratory species. Study sitesship exists between probability of detecting were classified in broad categories (Table 6-1)a species and its local abundance (Wright for statistical analyses. Prior to analysis,1991); (3) siting of reserves and development percentage of species was arcsine transformedof conservation programs are often based and number of species square root trans-upon the types and numbers of species that formed to improve normality of samplingare present (e.g., Burley 1988); and (4) the distributions. Analysis of variance (ANOV A)

Table 6-1, Habitat, geographic, and environmental classes used in analyses of migratory bird distributionin the Neotropics.

Variable Category Description

Vegetation type Moist forest Semi-evergreen to evergreen forests > 10 m tall in regions generally receiving> 150 cm rain/year. Examples. include wet, moist, montane, and cloudforests.

Dry forest Semi-evergreen to deciduous forests > 5-10 m tall in regions generallyreceiving < 150 cm rain/year with a pronounced dry season. Examplesinclude dry, oak, and sclerophyll, as well as arid limestone, forests.

Pine forest Habitats dominated by coniferous trees. Examples include pine and pine-oak-fir forests and pine-savanna.

Scrub Early successional or naturally occurring broadleaved habitats < 5-8 m tall.Examples include early successional dry and moist forests, thornscrub, andsavanna-scrub.

Open Natural or human-altered habitats low in stature with few woody plants.Examples include open field, pasture, coastal dunes, marsh, and savanna.

Artificial Vegetation types heavily altered for agricultural or residential uses, but withvegetation > 4 m tall. Examples include urban areas/parks, citrus andcoffee plantations.

Vegetation height ~5 m6-10m11-20 m>20m

Disturbance Disturbed Moderate to heavy disturbance. Examples of disturbances include logging,fragmentation, agriculture, residential, and clearing of forest understory.

Undisturbed Very slight to nonexistent disturbance.Elevation.. 200 m

201-500m501-1000m1001-2000 m> 2000 m

Insularization MainlandIsland

Latitude ..5° N6°-15° N16°-25° N>25° N


was used to assess the null hypothesis that 14 A

no differences in migratory bird abundance (/) 12exist among categories for each environ- ~mental variable. If significant differences were ~ 10

detected, Tukey's test (Zar 1984) was used to (/) 8

identify statistically unique groups. For all '5hypotheses and in all analyses, .a critical (1: 6

probability of 0.10 was set for statistical Mfsignificance to reduce the chances of ~ 4

committing Type II statistical errors, which z 2often are of greater concern than Type Ierrors in conservation research (Askins et al. 0 MOIST DRY SCRUB PINE RES/AG OPEN

1990, Petit et al. 1992).The analyses described above are subject VEGETATION TYPE

to some biases because differences in plot size, ~ 60

Bsampling intensity, and sampling technique (3among studies were largely ignored. These ~ 50uncontrolled sources of variation would (/)

b h . 11 .--140serve to o scure t e potentIa y Important ~ecological effects that were being assessed, ~ 30Thus, although the broad-scale trends that u.are documented should be viewed cautiously, O 20

these results probably represent conservative ~estimates of the environmental effects on ~ 10habitat use by migratory birds. ffi

0. 0MOIST DRY SCRUB PINE RES/AG OPEN

Vegetation Type VEGETATION TYPE

Neotropical migrants were not distributed. .uniformly across vegetation types (F = 5.89. FIgu.re 6-1. (A) Average nu~ber of mIgratory

' specIes recorded per study sIte, and (B) averagedf- 5, 206, p < 0.001). MI~ratory spec!es percentage of total species (migrants and residents)~epre~ente~ more.than one-thlrd .of all sp~cles represented by migrants in different vegetationIn resldentlal/agncultural and pIne habItats, types. MOIST = moist forest; DRY = dry forest;30% in early successional and dry forest SCRUB = scrub or early successional growth;habitats, and less than one-quarter of the PINE = pine forest; RES/AG = artificial habitats,total in open areas and moist forests; such as urban parks, residential areas, and citrushowever, definitive habitat groupings were groves; and OPEN = open habitats, such asdifficult to define (Fig. 6-1B). In contrast, grassland and pasture. ~rror bar. rep~esentsnative broadleaved vegetation types (wet one standar~ er~or. VegetatIon types ":Ith dIfferent

.letters are sIgnIficantly (P < 0.10) dIfferent. Seeforest, dry for.est, and early successIonal text and Table 6-1 for details.growth) contaIned a greater number ofmigratory species per site than more artificial,managed habitats (pine, residential/agri- distributed widely across habitat types-cultural), and open habitats with little woody natural and artificial, disturbed andvegetation (F = 4.12; df = 5, 188; p = 0.001; undisturbed-is supported by these analyses.

Fig. 6-1A). However, only open grassland/ In addition, migrants were proportionallysavanna habitats were conspicuously devoid underrepresented in moist, evergreen forestof migrants. On average, approximately 9-11 and open grassland/savanna (compared tospecies of migratory birds per site were other vegetation types) during the borealidentified in all habitat types except open winter (Karr 1976, Terborgh 1980). Com-areas, which averaged only five species parison of numbers of species found in each(Fig. 6-1A). vegetation type, however, suggests that moist

The general tenet that migratory birds are and dry forests, and early successional/scrub

150 TEMPORAL PERSPECTIVES ON POPULATION IMITATION AND H

habitats contain the most migratory species. 14Residential/agricultural habitats and pine (/)forests were not statistically distinguishable ~ 12

from those broadleaved vegetation types but ~ 10contained, on average, 14-22% fewer species. (/)

These analyses provide a general pattern ofhabitat use by migrants. However, becausesites within each vegetation category variedgreatly in such characteristics as level ofdisturbance, age, and geographic location,these trends do not identify specific habitattypes in which migrants were most abundant.A more detailed examination of migrant useof major vegetation types will be presentedlater in this chapter. B

Vegetation Height

Migrants represented a greater proportion ofthe bird community in intermediate vegeta-tion (5-20 m tall) than in relatively short ortall vegetation (F = 16.66; df = 3, 208;p < 0.001; Fig. 6-2B). Analysis based onnumbers of migratory species in each heightcategory showed that, generally, more speciesoccurred on sites with canopies that were> 10 m tall than on sites where canopies wereof shorter stature (F = 7.44; df = 3, 190;p < 0.001; Fig. 6-2A).

This analysis cannot directly equatevegetation heights with particular habitattypes. However, typically only moist forestsites exceeded 20 m and open grassland/savanna habitats were < 5 m tall, a furtherindication that migrants are underrepre-sented within the bird community occupyingtall, moist forests and open grasslandhabitats. However, migrant species richnessclearly increased with canopy height, which trend was not significant (F = 245; df = I,supports the theory that availability of 192; p = 0.11 ; Fig. 6-3A). T ese resultsadditional resources (niches) allows for further suppor the paradigm tha migratorypersistence of a greater number of species birds have a tronger affinity th n resident(MacArthur and MacArthur 1961, Orians species to isturbed habita s in the


12mIgratory birds than habitats where most

A native woody vegetation has been cleared10 ( e.g., pasture, row crops, sugarcane; Sa ab and

{1) Petit 1992). One exception to this generaliza-~ tion may be Citrus groves, where migratorywe bird abundance (richness and density) can3; reach high levels (Mills and Rogers 1992,

~ 6 Robbins et al. 1992).

!1:w~ 4 Elevation

~ Migratory birds were well-represented inlowland ( < 1000 m) habitats in both relativeand absolute terms. The proportion ofmigratory species within local communities

YES NO declined linearly with elevation (Spearman'scorrelation coefficient; r = -1.0, n = 5,

40 DISTURBANCE p < 0.001), although Tukey's multiple

comparison test showed broadly overlapped35 B elevational groupings (F = 7.32; df = 4, 207;

~ p < 0.001; Fig. 6-48). The number ofu 30 migratory species per site also decreased with~ an increase in elevation (r = -0.90, n = 5,~ 25 p = 0.037) but this difference was not~ significant with analysis of variance (F = 0.77;~ 20 df = 4, 189; p = 0.548; Fig. 6-4A). On

11- average, approximately 18% more migratory0 15 species were detected per site in habitats

~ below < 1000 m (9.9 species) compared with~ 10 middle and high elevations (8.1 species).ffi Thus, the often-cited generalization thatDo 5 migrants occur disproportionately at middle

elevations (e.g., Leck 1972b, Karr 1976)appears to be unwarranted.

YES NO

DISTURBANCE I ..Insu arlzahon

Figure 6-3. (A) Average number of migratory . b . d . d I .h Id d t dt d (8) Migratory Ir s comprIse a s Ig t y greater

specIes recor e per s u y SI e, an average. ..percentage of total species (migrants and residents) prop~rtlon of the total .avlfauna o~ Islandrepresented by migrants in disturbed and un- (32.4%) compared to maInland (28.4%) sItesdisturbed habitats Error bar represents one (F = 2.75; df = I, 210; p = 0.098; FIg. 6-58),

standard error See text and Table 6-1 for details. but this result was due to a latitudinal effect(see below) associated with the fact that mostCaribbean islands used in this study are

vegetation type (see below). Although located at high latitudes (> 15" N) within theNeotropical migratory birds are slightly tropics. When analyses were restricted to sitesmore abundant in disturbed habitats, not between 16° and 25° N, the trend wasall disturbed habitats are suitable for reversed; mainland sites (34.6%) containedmigrants. Land-use practices that allow some significantly (F = 3.40; df = I, 124; p = 0.068)

of the natural canopy and subcanopy trees greater proportions of migratory species thanto remain, such as selective logging, island sites (28.1%).low-density residential areas, or cacao Nearly twice as many migratory speciesplantations, provide habitat more suitable for (F = 13.90; df= I, 192; p < 0.001) were


14 1212 A A A

(I)!!:! 10010W (I)

fu w8 O 8

u. W0 Q.a:6 (I)

W u.[Q4 06~::> a:Z 2 W

~ 40 ::>

0- 200 200 -500 500 -1000 1000 .2000 > 2000 Z

ELEVATION (m) 2

(I) 40

!!:! B 0~ 35 MAINLAND ISLAND

fu 30 LOCATION

--1 40~25

B'01- 20 (I) 35u. W015 830

~10 fuw° --125a: 5 ~W 0Q. 0 1- 20

0.200 200- 500 500 -1000 1000 -2000 > 2000u.

ELEVATION (m) 0151-Z

Figure 6-4. (A) Average number of migratory ~ 10species recorded per study site, and (8) average ffipercentage of total species (migrants and residents) Q. 5

represented by migrants at different elevations.Error bar represents one standard error. 0Elevations with different letters are significantly MAINLAND ISLAND(P < 0.10) different. See text and Table 6-1 for LOCATIONdetails.

Figure 6-5. (A) Average number of migratory..species recorded per study site, and (8) average

reco~ded on contl?ental ~tudy sItes (1~.6 percentage of total species (migrants and residents)specIes) than on Island sItes (6.4 specIes; represented by migrants on islands and mainland.Fig. 6-5A), contrary to patterns described by Error bar represents one standard error.Karr (1976) and Leck (1972b). These results See text and Table 6-1 for details.suggest that migrants and residents may beinfluenced in similar ways by ecologicalfactors thought to dictate the composition of islands (see below), habitat availabilityinsular faunal communities (MacArthur and probably cannot account for the greaterWilson 1967), for example, distance from the number of migratory species in each of themainland (see also Terborgh 1980), avail- mainland habitats. Rather, the paucity ofability of habitats, predators, and com- migratory species on islands may be relatedpetitors. However, because migrants in moist, to the relative energetic costs associated withdry, and pine forests on continents still were flying long distances over unsuitable (water)significantly more abundant than their habitat in search of relatively small landcounterparts in the same habitat types on masses (Moreau 1972, Terborgh and


Faaborg 1980). In support of this, Wunderle 12and Waide (1993) found that migratory bird (/)abundance on Caribbean islands was ~ 10

negatively related to the distance from the ~North American mainland. Nevertheless, the (/) 8

importance of Caribbean islands in providing ~ 8migratory stopover areas and overwintering 0:habitats for migratory birds should not be ~ 4underrated (e.g., see lack and lack 1972, ~Pashley 1988, Arendt 1992). z 2

0. d < 5 6-15 16-25 > 25

Lahtu e LATITUDE ( o N)

Migratory birds represented an increasing 50proportion of the total overwintering f!Javifauna from lower to higher latitudes @in the Neotropics (F = 46.34; df = 3, 208; fu 40

p < 0.001; Fig. 6-6B). Migrants comprised ..Jabout 5% of the overwintering species in 9 30

South America (excluding the northern 1-coastal areas ), 15% in southern Central ~ 20America (south of 15° N), 34% between 1-Belize/Guatemala and central Mexico ~ 10(25° N), and nearly 40% in northern Mexico ~and southern Florida (see also Keast 1980). ~The northern two regions, however, did not 0 < 5 6- 15 16- 25 > 25

di~er significantly in their .proportions ,of LATITUDE (0 N)mIgrants. The number of mIgratory specIesper. study site peaked ~etween 16° and 25°, Figure 6-6. (A) Average number of migratorywhIch represented twIce the number of species recorded per study site, and (8) averagespecies recorded per site in South America percentage of total species (migrants and residents)(F = 5.40; df = 3, 190; p = 0.001; Fig. 6-6A). represented by migrants at different latitudes.Fewer migratory species at lower latitudes Error bar represents one standard error,may be related to the increased costs Latitudes wi,th different letters are significantlyassociated with flying longer distances during (P <: 0.10) dIfferent. See text and Table 6-1 formigration (Greenberg 1980, Terborgh and detaIls.

Faaborg 1980). However, because the centersof winter distribution for many migratoryspecies lie between 10° N and 20° N, and not param~ters~ such that severa! dozen naturallyfurther north other considerations such as occurnng lIfe zones (Holdndge et al. 1971)habitat availability and historical factors may are distin~~ishable wit~in the ~e.otropics.partly define overwintering ranges. Many addItIonal vegetatIve assocIatIons also

could be defined if natural combinations wereconsidered in the context of anthropogenic

IMPORTANCE OF VARIOUS HABITAT TYPES disturbance (Hartshorn 1983, p. 120). Below,TO NEOTROPICAL MIGRATORY BIRDS a more detailed examination of the relative

use of major vegetative associations byKnowledge of general vegetation relation- overwintering migratory birds is presented.ships of migratory birds is critical during the The trends identified are meant to serve asinitial phases of conservation efforts in the general justification for conservation prior-Neotropics. Vegetative associations can itization of habitat types. However, bird-be defined by unique combinations of habitat relationships are far more complexclimatic, topographic, edaphic, and biological than discussions presented here would


indicate, and identification of actual sites that undisturbed forests. This trend was evidentare appropriate for a reserve system need to even when forest canopy height wasbe evaluated at a finer scale of resolution. statistically controlled, suggesting thatSpecies listed under each habitat type were small-scale forest fragmentation, forestderived largely from a quantitative sum- openings, and other minor-to-moderatemarization of the literature (see "Geographic disturbances enhance the suitability of moistvariation in habitat use by species " below; forest habitats for many species of over-

see also Petit et al. 1993), with additional wintering migratory birds.information taken from several key studies Overall, percentage of migrant speciesconducted within each vegetation type. (r s = -0.90, p = 0.037, n = 5) and absolute

number of species (r s = -0.60, p = 0.285,M . t/W t F t n = 5) inhabiting moist forests declined

015 e ores ho h I . I h h d..at Ig er e evatlons, a t oug Istmct

Moist forests, popularly referred to as rain altitudinal preferences were difficult to detectforests, occur where the dry season is with multiple comparison tests (percentrelatively short and annual precipitation species: F = 8.82; df = 4, 61; p < 0.001;exceeds 150 cm (Whitmore 1991 ). The mostly number of species: F = 2.92; df = 4, 52;evergreen forests supported under these p = 0.029). In disturbed forests theseclimatic conditions vary greatly in phy- elevational relationships were weak (P valuessiognomy and tree species composition >0.10). Alternatively, migratory birds in(Sutton et al. 1983), thereby making undisturbed moist forests exhibited a stronggeneralizations about the associated avifauna decline in proportional representation andtenuous unless certain distinctions are made absolute numbers up an elevational gradientamong forest associations. The most widely (P values <0.05). For example, undisturbedrecognized formations of moist tropical forest coastal and lowland evergreen forest sitesare based upon Holdridge's (1947) life-zone contained twice as many migratory speciesconcept and applied to elevational gradients as undisturbed, high elevation (> 1000 m)(also see Whitmore 1991). Here, five forest plots.broad categories of moist forest are In nearly all moist forest associations,considered: coastal ( < 200 m), lowland disturbed sites supported significantly moreevergreen (200-500 m), lower montane migratory species than relatively pristine(500-1000 m; includes premontane), mon- sites, although small sample sizes in sometane (1000-2000 m), and upper montane cases hindered interpretation of results. In( > 2000 m). Coastal forests were considered general, migrants as a group show a strongseparately from lowland evergreen forests affinity for lowland ( < 500 m) evergreenbecause of differences in human settlement forests compared to montane forests, but thispatterns in these two areas (e.g., see Myers relationship is lessened when forests are1980, p. 150; Lynch 1992). disturbed. Migrants may increase use of

Proportional representation of migrants disturbed forests at higher elevations ifwas inversely related to vegetation height, slight-to-moderate levels of disturbancesuch that migrants comprised only 16% of enhance suitability of moist forests for thosethe total species in mature forests greater species (Petit 1991). Nevertheless, higherthan 20 m tall compared with 40-50% of the species diversity is often obtained at theavifauna in shorter moist forests (F = 22.99; expense of species that are sensitive to forestdf = 2, 63; p < 0.001). In contrast, the disruption-those species currently most innumber of migratory species detected per site danger of population declines or extinctionwas greatest in forests taller than 15 m (Whitcomb et al. 1976, Petit et al. 1993).(F = 4.05; df = 2, 54; p = 0.023). Thiollay (1992) and Johns (1989) found that

Disturbed moist forests contained more selective logging of moist forest decreased themigratory species, both proportionally (28% local diversity of resident bird species invs 21%; F = 3.09; df= 1,64; p = 0.084) and French Guiana and Malaysia, respectively.in absolute terms (13.0 vs 9.9 species; Migratory birds inhabiting moist forestsF = 2.87; df = 1, 55; p = 0.096), than exhibited distributional patterns similar to


those found for the entire migratory bird effects of the unique geographic distributionscommunity, as they were best represented on of these forest types are removed (that is, drymainland sites and at middle latitudes forests included in this study were mainly(6°-25° N), compared to islands and more found in northern Central America andextreme latitudes within the Neotropics, Mexico, whereas moist forests were dis-respectively. tributed throughout Latin America). For

Although the general pattern of wet forest example, in northern Latin America, moistuse suggests that disturbed and lowland forest sites (13 species) supported approxi-forests support the greatest number of mately 40% more migratory species onmigratory birds, all types of moist forest average than dry forest sites (nine species)provided suitable habitats for overwintering Disturbance appeared to have a weakmigrants. On average, 11 migratory species effect on numbers of species occupying drywere detected within each plot of evergreen forests (F = 0.94; df = 1, 23; p = 0.343);forest (Fig. 6-18), which represented about disturbed sites contained nearly 40% moreone-quarter of the total local winter avifauna. species, but small sample size in thisMoreover, mature evergreen forest types are comparison obscured clear interpretation ofprincipal winter habitats for numerous results. Numbers of migratory species weremigratory landbirds, such as Olive-sided and greater in relatively tall (> 10 m) dry forestsYellow-bellied Flycatchers, Wood Thrush, compared to forests 5-10 m tall (F = 3.71;Yellow-throated Vireo, Scarlet Tanager, df = 2, 22; p = 0.041). Mainland dry forestsNorthern Waterthrush, and 8lackburnian, contained more migratory species (F = 3.10;Worm-eating, Cerulean, Prothonotary, and df = 1, 23; p <0.092), but comprised aKentucky Warblers. In summary, con- similar proportion of the overall avifaunaservation of many migratory species, in (F = 0.23; df = 1, 25; p = 0.639), comparedaddition to hundreds of species of resident to similar island sites.tropical birds, is highly dependent upon Dry forests represent an important habitatpreservation of moist evergreen forests. The to migrants during winter, especially atcurrent emphasis among conservationists on more northerly latitudes (Petit et al. 1993).preserving moist forests, therefore, seems Species commonly found in these habitatsappropriate. include Wood Thrush, Least Flycatcher,

8lue-gray Gnatcatcher, Solitary, white-D F eyed, and Warbling Vireos, and Orange-

ry orest crowned, Nashville, 8lue-winged, 8lack-and-

Tropical dry forests occur most frequently in white, 8lack-throated Gray, Magnolia, andareas of western Mexico and central America, Hooded Warblers.northern and western South America, and on Retention of deciduous and semideciduousthe Yucatan Peninsula and several Carib- forests is especially important for migratorybean islands. Low precipitation (usually bird conservation because much of Mexico< 150 cm/year) and stressful edaphic con- and the Pacific slope of Central America areditions often produce climax forest types composed of dry forest formations, andrepresented by short ( < 20 m), sclerophyllous because a large proportion of all migratorytrees or thorn forests (Murphy and Lugo birds (especially western North American1986). Dry, semi-deciduous forests are species) overwinter in that region (Hut tousually more open and less structurally 1980, Terborgh 1980). Little is known aboutcomplex than moist, evergreen forests migrant use of dry forest habitats in South(Holdridge et al. 1971). America (8osque and Lentino 1987).

The number of migratory species within Furthermore, some types of dry forest areeach dry forest study plot (ca. 10) was similar often the first to be cut for agricultureto that found in moist forests (11), and (Holdridge 1970, Hartshorn 1992; see alsomigrants constituted similar proportions of Sader and Joyce 1988). This selectivity ofthe local avifauna in both forest types tropical dry-forest removal during the past(Fig. 6-1). The relative importance of moist century has resulted in a proportionalforest becomes evident, however, when the reduction exceeding that of nearly all other

,,~o", ,,"",'MOO, ~"",'", ,-",,'", ~" ."",, ",:'07;;" ",",M" ""'""0"' ""m""


extreme climatic or edaphic conditions (e.g., Open Habitatslack of moisture). Low food (arthropods andfruit) abundance may be a corollary of those Open habitats represented a wide variety ofclimatic regimes, such that few migrants can natural (e.g., marsh, grassland, coastal dune)be supported. Additionally, the relatively and modified (e.g., pasture, open agriculturalhigh net primary productivity of early field) vegetation types that were charac-successional forests may support a large terized by few broadleaved, woody plants andbiomass of herbivorous insects (Janzen 1973), low canopy closure (usually < 10%). As such,which make those habitats attractive to these habitats might be expected to supportinsectivorous migrants. few bird species. Sample size (19 sites) was

Migrant species richness increased at more small for this vegetation type, so trendsnortherly latitudes (> 15° N), and peaked should be viewed with caution.between 15° N and 25° N (F = 4.22; df = 2, Migrants were relatively uncommon in26; p = 0.026). Approximately four more open habitats, although they did comprisemigratory species were found in the typical one-fifth of all species detected there (Fig.mainland scrub habitat (x = 11 species) 6-1B). The average number of migratorycompared to island sites (seven species), but species recorded within disturbed, open plotsthis disparity was not significant (F = 1.12; (seven species) was more than twice thatdf = 1, 27; p = 0.299). found in natural, open habitats (three species;

Early successional habitats are important F = 2.95; df = 1, 17; p = 0.104). These datafor several dozen migratory species, such as indicate that, whereas open pastures andBell's Vireo, Yellow-breasted Chat, Common agricultural fields support relatively fewYellowthroat, Ovenbird, Gray Catbird, Least species, marshes, dunes, and natural grass-Flycatcher, Blue Grosbeak, and Indigo lands may be used even to a lesser extent byBunting (e.g., Karr 1976, Waide 1980, Martin migratory landbirds (see also Karr 1976).and Karr 1986, Hut to 1989, Lynch 1989, Although not a critical habitat for migratoryBlake and Loiselle 1992, Petit et al. 1992). In land birds as a group, natural open habitatsfact, often > 50% of the local overwintering should not be characterized as "expendable "

migratory species are found in early because marshes and grasslands are importantsuccessional habitats, although most species habitats for overwintering shorebirds (Myersare usually detected within other habitats as 1980) and many resident species. Migratorywell (Waide 1980, Lynch 1989, Petit 1991, land birds often found in these low statureBlake and Loiselle 1992, Hut to 1992, Kricher habitats include Scissor-tailed Flycatcher,and Davis 1992). Thus, many migrants Yellow-rumped Warbler, Palm Warbler,appear capable of using forested habitats in Common Yellowthroat, Indigo Bunting,many stages of succession, especially in Painted Bunting, Savanna Sparrow, Grass-northern areas of the Neotropics (e.g., Lynch hopper Sparrow, Bobolink, and Dickcissel.1989, Hut to 1992). Wildlife management practices used in the

Early successional habitats are not the highly agricultural areas of North Americatarget of conservation efforts in the tropics and Europe also may be applicable tobecause of the current plethora of these agricultural landscapes in the Neotropics.vegetation types, a result of rapid expansion For example, the maintenance of hedgerows,of shifting cultivation during this century. woodlots, stream-side (wooded) buffer zones,However, while the current conservation and scattered overstory trees could provideemphasis is on species associated with mature some migratory species with suitablemoist forests (Wilson 1988), conservationists habitats, even when most of the originalmust recognize that development of perm- forest in an area had been converted toanent pasture and agricultural plots con- pasture or cropland (Lynch 1989, Gradwohlsumes both mature forest and young second and Greenberg 1991, Sa ab and Petit 1992).growth. Thus, the current conservation status Gradwohl and Greenberg (1991) suggestedof early-successional birds could be easily that, for conservationists, these practicesreversed as disturbed second-growth sites are represent "making the best of a badconverted to permanent agriculture. situation ". Indigenous farmers in Latin

158 TEMPORAL PERSPECTIVES ON POPULATION LIMITATION AND HA81TAT USE

America have incorporated these principles requirements of some migratory species.into their land-use practices for centuries, but Twice as many migratory species were foundwith mounting population pressures and a on mainland sites compared to island sites.switch from family gardens to large-scale Overall, migrants represented a highmonocultures, these conventions are likely to proportion of the total species (43%) inbe reduced or eliminated in many areas residential and agricultural habitats (Fig.(Alcorn 1990). 6-1B), which suggests that these habitats may

be poor for resident birds. As an example,Artificial Habitats Mills and Rogers (1992) captured 25 mi-

gratory and 37 resident species during 2 yearsArtificial habitats in this study were of mist-netting in a Citrus grove in Belize. Inpredominantly residential yards and agri- comparison, Petit et al. (1992; D.R. Petit,cultural plantations dominated by tree crops unpublished data) detected 20-25 migratory(e.g., Citrus, Coffea), thereby representing species, but approximately 2.5 times ashighly disturbed areas. The unnatural many resident species within each of threevegetation structure (e.g., hedges, widely broadleaved habitat types < 50 km awayspaced trees) and sometimes exotic tree from Mills and Rogers' Citrus grove. Millsspecies result in artificial habitats, which may and Rogers (1992) suggested that migrantsor may not resemble native vegetative were more abundant than residents in Citrusassociations. The discussion that follows is groves because migrants were attracted bysupplemented by published accounts based the abundance of insects associated withupon mist-netting because few studies based rotting fruit, and that nearly all insects foundupon visual surveys were available. Trends there were too small (nearly all insects wereidentified here should be considered pre- < 5 mm) for large-billed residents to exploitliminary and are not based upon statistical effectively. Likewise, Champe (1993) recordedanalyses. similar numbers of long-distance migratory

Residential and agricultural habitats species in Citrus groves and native,supported numbers of migratory species broadleaved forests in south Florida duringcomparable to the other habitat types winter. However, she found that densities ofconsidered in this paper (Fig. 6-1A). Within most of these species were lowest in Citrus,these habitats, Citrus groves supported highlighting one of the limitations of usingapproximately 25% more species, oriaverage, only species lists in conservation planning.than did residential or agricultural areas with Coffee plantations cover extensive areas ofoverstory trees. That Citrus plantations seem tropical America (Leonard 1987) andto harbor disproportionate numbers of represent potentially important wintermigratory species has been noted by others habitats for migratory birds (Wunderle and(e.g., Mills and Rogers 1992, Robbins et al. Waide 1993). Coffee can be grown under1992), but reasons for this pattern remain management regimes that vary in theirobscure. In the present study, relative suitability for some migratory species. Formigrant abundance in residential/agricultural example, "shade coffee" is cultivated as anhabitats may be misleading because most of understory crop with an overstory of canopythose studies were conducted at higher trees. "Sun coffee," on the other hand, islatitudes (e.g., Mexico, Belize), where grown without the benefit of shade trees.geographic location could have accounted Shade coffee often supports more migratoryfor the inflated numbers (see above). species because the presence of additionalHowever, comparison of migrant species vegetative layers provides habitat for canopyrichness at sites located between 16° and and subcanopy foragers (Robbins et al.25° N confirmed that residential/agricultural 1992, Wunderle and Waide 1993; L. J. Petit,habitats supported numbers of migrants unpublished data). Canopy trees that flowersimilar to those found within native or fruit during the boreal winter also providebroadleaved vegetation types. Thus, certain resources for nectarivorous and frugivorousresidential and agricultural habitats do not migrants in coffee plantations (L. J. Petit,

appear to be incompatible with habitat unpublished data).


Other crops, such as corn (Zea), rice because high local densities could reflect rigid(Oryza), and sugarcane (Saccharum), appear dominance relationships (i.e., ideal despoticto provide few migratory species with distribution; Brown 1969, Fretwell and Lucasappropriate habitat (Robbins et al. 1992; L. J. 1970) and not distributions based uponPetit, unpublished data). The lack of relative habitat suitabilities (i.e., ideal free;vegetative structure and the often heavy use Fretwell and Lucas 1970). However, in someof pesticides in these fields result in low- tropical habitats, such as undisturbed, moistquality habitat for most species. Several forest, density of migrants can be especiallyspecies, however, can occur in high low (Hut to 1980, Willis 1980, Mabey andabundance in rice (Indigo Bunting; Robbins Morton 1992) and winter territories ofet al. 1992) and sugarcane (Barn Swallow; individuals of some species often are notL. J. Petit, unpublished data). contiguous (Bennett 1980, Chipley 1980),

A broad spectrum of migratory birds use which suggests that some species are not atagricultural and residential habitats (e.g. their carrying capacities in these habitats. InTramer 1974, Mills and Rogers 1992, other areas, however, migrants do occur atRobbins et al. 1992, Baltz 1993), including high densities (Rappole and Warner 1980,species more typical of forests (e.g., Wood Holmes et al. 1989), and dominance (i.e.,Thrush, Hooded Warbler, Black-and-white territorial) relationships among individualsWarbler, Northern Waterthrush), edges and could force some birds into less suitablesecond growth (White-eyed Vireo, Magnolia habitats. For example, young birds or femalesWarbler, American Redstart, Orchard may be forced into suboptimal, disturbedOriole) and grasslands (Common Yellow- habitats, where they overwinter in highthroat, Indigo Bunting). Because many areas densities relative to undisturbed forest (seein the tropics are being converted to Citrus below). If so, the mere distributions ofgroves, coffee plantations, agricultural fields, individuals across habitat types does not offerand residential areas, determination of the an accurate reflection of the relativesuitability and importance of these habitats suitabilities of those habitats, or indeed offor migrants will be an important component habitat "preference" patterns. Rather, over-of future conservation research. winter survival and other potential indicators

of habitat quality, such as physical conditionS f H b " t t U of individuals and age and sex ratios, need

ummary o a I a se b . d . h . f I .to e Integrate WIt estImates o re atIve

Analyses presented here provide an indica- abundance into an assessment of the valuetion of the relative use of different tropical of each habitat to migratory birds (Holmeshabitats by migratory birds. However, et al. 1989).caution must be exercised when these results On the other hand, distributions ofare applied to specific situations because the individuals among habitats may indeedrelative importance of each environmental provide a reliable indication of the relativevariable (e.g., vegetation type, vegetation suitabilities of those vegetation types (Oriansheight, latitude) in the context of all other and Wittenberger 1991). For example,variables could not be fully assessed in this physical condition (Greenberg 1992a) andstudy (small sample sizes prohibited use of survival rates (estimated indirectly throughmore sophisticated statistical analyses). recapture rates; Rappole and Warner 1980,Clearly, all tropical habitats contained some Robbins et al. 1987, Blake and Loiselle 1992,migratory birds. Disturbed habitats, whether Conway et al. 1995) are not necessarilynatural (e.g., regenerating or selectively improved for "forest-dwelling" species inlogged wet forest) or unnatural (Citrus undisturbed forests relative to disturbedplantation), typically supported a greater habitats (but, see Rappole et al. 1989). If thisnumber of migratory species than similar classical interpretation of habitat use byundisturbed habitats. Some arguments populations is correct (that is, density reflectscaution against using relative abundance of suitability), then results presented here andorganisms as an index of suitability (e.g., Van those of others suggest that many migratoryHorne 1983, Gibbs and Faaborg 1990) species are not adversely affected by at least


some current land-use practices in Latin community in unburned hurricane-damagedAmerica and the Caribbean. Indeed, migrants forest were similar to pre-hurricane valuesoften reach their greatest abundance in (Lynch 1991, 1992). The recovery of thecertain disturbed natural and artificial migrant community to pre-hurricane con-habitats. The task for students of migratory ditions was slower in hurricane-damagedbird ecology is to ascertain the level (and forest that was further disturbed by wildfires.form) of disturbance tropical forests can However, even in tracts where >90% ofwithstand and still be suitable as winter the trees had been killed and the groundhabitat for Neotropical migrants (also see cover of vegetation completely eliminated byLynch 1992). Clearly, however, wildlife fire, a diverse assemblage of migrants wasconservation plans also must consider present several months later. Only tworesident tropical birds (and other taxa), previously common forest-associated mi-which often are more adversely affected by grant species (Wood Thrush and Kentuckyalteration of natural forests (e.g., Hutto 1989, Warbler) were absent from burned hur-Kricher and Davis 1992). ricane-damaged forest 17 months after

At present, only a few general types of land wildfires razed the area, and one of theseuse appear to be particularly incompatible (Wood Thrush) reappeared during thewith habitat use by all but a few species of second winter after the fire.migratory landbirds: (1) actively grazed In contradiction to the relatively minorpastures that contain few woody plants (Karr and transitory effects of the hurricane and1976, Sa ab and Petit 1992; also see Lynch fire on migratory species, the effects of the1992, p. 192; Robbins et al. 1992, p. 216; hurricane and wildfire on resident birds wereChampe 1993); and (2) large monocultures both more severe and longer lasting. Manyof monocotyledonous crops, such as bananas resident species, particularly frugivores and(Robbins et al. 1992), rice (Robbins et al. nectarivores, completely disappeared from1992; L. J. Petit, unpublished data), and the study area when it burned, and still hadsugarcane (L. J. Petit, unpublished data). not recolonized the forest 3! years laterMost migrants occupying those agricultural (Lynch 1991; J. F. Lynch, unpublished data).lands are habitat generalists (e.g., Yellow- As a result, migrants made up a higherrumped Warbler) or are characteristic of proportion of both species and individuals inhighly disturbed vegetation types (e.g., burned than in unburned forest for at leastCommon Yellowthroat, Indigo Bunting) in several years after the disturbance occurred.their wintering areas, sucG that winter habi- Askins and Ewert (1991) also showed thattats are not presently limited for those species. resident birds on a Caribbean island were

Analyses presented above indicate that more adversely affected by Hurricane Hugomigratory birds as a group show a stronger than were overwintering migratory species.affinity toward disturbed habitats than do Likewise, in Los Tuxtlas, Mexico, a greaterresident species, which suggests that some percentage of the individuals and species thatforms of tropical forest alteration may affect were mist-netted after part of a forestmigratory species to a much lesser extent fragment was cleared for cattle pastures werethan residents. Examination of avifaunal migrants (individuals = 62%, species = 42%)response to direct habitat destruction compared with their representation beforedemonstrates that, indeed, migratory birds as alteration of the forest (54% and 33%,a group may be more resilient to natural and respectively; Rappole and Morton 1985).human-caused habitat disruption than areresident tropical species. For example, on theYucatan Peninsula, dry forests that were REGIONAL PATTERNS OF HABITAT USEcompletely defoliated by Hurricane Gilbertin September 1988 were recolonized by The above summary of habitat use can benumerous migrants within a few weeks used to formulate general recommendationsafter the storm. Less than 2 years after as to which vegetation types and areas arethe hurricane, the abundance, diversity most important in general as overwinteringand species composition of the migrant habitats for migratory birds. However,


because vegetative and historical factors than that found overall in this habitat typeunique to certain regions may greatly (10 species). In contrast, numbers of speciesinfluence the habitats used by migratory in east Mexican moist (12 species) and drybirds, as well as the species found there, these (10 species) forests were similar to thosegeneral patterns may not be fully applicable observed throughout the Neotropics inacross all regions in the Neotropics. Below, general (see Fig. 6-1A). Migrants in easternhabitat use by migrants in six broad Mexico represent a slightly more substantialecopolitical regions in Latin America and the proportion of the local avifauna in broad-Caribbean is briefly reviewed. These over- leaved (moist and dry) forests (31% ofviews also are meant to identify those regions all species) compared with other broad-and habitats most in need of further leaved forest communities throughout theornithological investigation. Neotropics (26%; see also Fig. 6-1B).

Migratory species that occupy a broad rangeE M .of habitats during winter in the Yucatanastern exlco . 1 .

d Fl h GPemnsu a mclu e Least ycatc er, rayThe ecology of overwintering Neotropical Catbird, White-eyed Vireo, Northern Parula,migrants has been studied intensively in two and Magnolia Warbler (Lynch 1989).areas of eastern Mexico (defined here as sites Second, overwintering migrants are botheast of the continental divide). Rappole and diverse and abundant in eastern Mexico. Incoworkers (Rappole et al. 1989, Rappole and the Yucatan, despite the fact that the speciesWarner 1980; Rappole and Morton 1985, pool of residents far exceeds that of migrantsWinker et al. 1990) have surveyed migrants (Paynter 1955), migrants commonly con-at lower elevations in Los Tuxtlas, a humid stitute 30-50% of the individual birds, and(annual rainfall = 400-500 cm), mountainous 30-40% of the species surveyed in a givenarea along the Gulf coast of southern habitat (Lynch 1989, 1992). In Los Tuxtlas,Veracruz. Several investigators have studied about 30% of the bird species present duringNeotropical migrants in the northern half of winter are Neotropical migrants (Coates-the Yucatan Peninsula (states of Campeche, Estrada et al. 1985), and migrants made upQuintana Roo, and Yucatan). This area > 50% of the individuals that were mist-supports natural vegetation from xeric netted in both heavily and lightly disturbedthorn scrub in Yucatan (annual rainfall = moist forest (Rappole and Morton 1985).50- 75 cm), to semi-evergreen tropical forest Finally, many migrants are more abundantin southern Campeche and Quintana Roo in forest than in heavily disturbed or open(annual precipitation = 110-140 cm). Mi- habitats, although few migratory species aregrants in the Yucatan have been surveyed in totally restricted to mature forest in easterndry and moist forests, various stages of Mexico. Disturbed vegetation is, indeed,successional growth, pastures and agricul- heavily used by some common migratorytural fields, residential areas, mangrove species in eastern Mexico, notably Com-swamps, savannas and coastal dune scrub mon Yellowthroat, Yellow-breasted Chat,(Tramer 1974, Waide 1980, Waide et al. 1980, Indigo Bunting, and Painted Bunting.Lynch et al. 1985, Lynch 1989, 1991, 1992, However, migrants such as Yellow-belliedLopez Ornat 1990, Lopez Ornat and Lynch Flycatcher, Wood Thrush, Yellow-throated1991, Greenberg 1992a). Vireo, Black-and-white Warbler, Kentucky

Several general patterns have emerged Warbler, Hooded Warbler, American Red-from these and other studies in eastern start, Black-throated Green Warbler, andMexico. First, migrants occur across Blue-winged Warbler are more common inessentially the entire available spectrum of mature forest than in second-growth habitatsvegetation types in eastern Mexico, but (Rappole and Warner 1980, Lynch 1989,species richness is often greatest in early Greenberg 1992a).successional growth/scrub. For example, the Many ornithological investigations havenumber of migratory species recorded per site been conducted in eastern Mexico, especiallyin early successional growth in eastern compared to other regions, but theMexico (16 species) was substantially higher geographic scope is still relatively limited.


More detailed attention is needed, for MacGillivray's Warbler in cloud forest;example, in the dry forest and scrub habitats Broad-tailed and Calliope Hummingbirds,in northeastern Mexico, and in the moist Red-naped Sapsucker, and Red-faced, Towns-montane and cloud forest associations of end's, and Hermit Warblers in mixedthe eastern highlands of Veracruz and coniferous forest; and Bell's Vireo, Lucy'sChiapas. Warbler, and Grasshopper Sparrow in young

second growth. However, many other speciesW t M .are habitat generalists (Hut to 1992), such as

es ern exlco ..WarblIng VIreo, Orange-crowned Warbler,

Western Mexico is characterized by relatively Nashville Warbler, Black-throated Graylow annual precipitation (typically < 150 cm) Warbler, Black-and-white Warbler, andand varied topographic relief associated with Northern Oriole. Because populations oftall mountain ranges. These factors have many long-distance migratory species thatcontributed to the predominance of four breed in western North America aremajor forest types characteristic of dry concentrated in western Mexico during theclimates and/or high elevations in that winter, maintaining intact forests in thisregion: (1) thorn forest, (2) deciduous and region is particularly critical to conservationsemideciduous broadleaved forest, (3) cloud efforts. Hut to's (1980, 1992) work hasforest, and (4) mixed (pine-oak-fir) forest. provided a wealth of information onThe overwintering migratory bird fauna is migratory bird use of forested habitatsderived almost exclusively from North throughout much of western Mexico.American breeding populations west of the However, a gap exists in our knowledge oflOOth meridian, which results in only the use of heavily modified habitats (e.g.,moderate overlap in winter bird community residential and agricultural areas, pasture) bycomposition between western Mexico and those species, limiting predictions of thesites on the Caribbean slope of Mexico (e.g., effects of future land use practices oncompare Hut to 1992 and Lynch 1992). The migrants.majority of survey work on Neotropicalmigrants in western Mexico has been N rth C t I A ..O ern en ra merlcaconducted durIng the past 15 years by Hut to(1980, 1989, 1992), who has demonstrated Northern Central America (here defined asthat, although migrants are most abundant Belize, Guatemala, Honduras, and Elin disturbed habitats, many species can be Salvador) supports a diverse array offound in undisturbed natural vegetation. migrants, both because of the great

In dry deciduous forests of western variability in vegetation types presentMexico, disturbance does not appear to within the region and because of the closeinfluence migrant richness greatly, much as proximity to North American breedingin dry forests of the Yucatan, but in contrast grounds. The dry Pacific slope (annualto patterns observed in moist forests rainfall typically 100-200 cm) of the northernthroughout the Neotropics (see above). The isthmus is partitioned from the moistgreatest number of migratory species in Caribbean lowlands (rainfall 150-500 cm)western Mexico overwinter in deciduous, by a cordillera often exceeding 2500 m.cloud, and mixed coniferous forests (Hut to Consequently, vegetation communities vary1992); cloud and coniferous forests appear to from cloud, montane and pine forests atbe particularly important for migrants here higher elevations to thorn, semideciduouscompared to those vegetation types in other and wet-forest associations in lowlands.regions (Petit et al. 1993). Some migratory Disturbed and artificial habitats (e.g.,species exhibit close associations with plantations, residential areas, forest frag-specific vegetation types in western Mexico: ments) are, as elsewhere in the Neotropics,Western Kingbird and Virginia's Warbler commonplace in northern Central America.in thorn forest; Western Flycatcher Oven- The Pacific slope harbors many migratorybird, and Summer Tanager in deciduous species from western North America, therebyforest; Black-throated Green Warbler and further enhancing the already rich regional


avifauna originating from eastern North Lowland pine savanna and upland pineAmerica. forest cover large areas in Northern Central

Few studies designed to examine habitat America and represent important winteruse by migratory birds have been conducted habitats for at least several migratory species.in northern Central America, especially In the pine-savanna scrub of Belize, thecompared to the large-scale surveys in Black-and-white Warbler, Black-throatedMexico, for example. Most of the following Green Warbler, Yellow-throated Warbler,discussion is based upon quantitative surveys and Common Yellowthroat form mixed-conducted in Belize and from qualitative species flocks with residents, while narrowaccounts of habitat use in Guatemala, bands of broadleaved vegetation along smallHonduras, and El Salvador. However, water courses in pine-dominated habitatsbecause of the limited data on habitat use by often contain American Redstarts, Magnoliamigrants on the dry Pacific coast, this general Warblers, and Ovenbirds (Petit 1991, D. R.summary may not be applicable to El Petit, unpublished data).Salvador, or certain areas of Honduras and Because little research in northern CentralGuatemala. America has been conducted outside of

Habitat use by migratory birds in northern Belize, future efforts to quantify migrant useCentral America is generally similar to of native and disturbed habitats in Honduras,patterns for the Neotropics as a whole (see Guatemala, and El Salvador would improveFig. 6-1), although migrants in northern our understanding of migratory bird-habitatCentral America do not appear to be as well relationships in this region (e.g., see Whitacrerepresented in residential habitats. Migrants et al. 1993). Indeed, the relatively high densityoften reach their greatest abundance in of migrants in northern Central Americadisturbed moist forests, including early requires that intense research and monitoringsuccessional growth and forest edges (e.g., activities be integral components of anKricher and Davis 1992, Petit et al. 1992), overall conservation plan for migratoryand in Citrus, cacao, Coffea, and other tree birds.crop plantations (Mills and Rogers 1992,Robbins et al. 1992, Vannini 1994, D. R. S th C t I A ...ou ern en ra mencaPetIt, personal observatIon). Pastures man-aged by repeated mowing or burning, and Lower Mesoamerica (Nicaragua, Costa Rica,active agricultural fields support few and Panama), like the northern region,migratory species (Robbins et al. 1992, Sa ab exhibits a general dichotomy in rainfalland Petit 1992). Undisturbed moist forests between Pacific and Caribbean slopes. Thecontain about 25% fewer migratory species distribution of moist and dry forests reflectsand individuals than disturbed forests in this climatological difference. Althoughnorthern Central America (e.g., Robbins et broadleaved forests are well representedal. 1992, Petit et al. 1992), although several in this region, pine-savanna, grasslands,species ( e.g., Wood Thrush, Kentucky agriculture, pasture, and early second growthWarbler, and Worm-eating Warbler) often also cover extensive areas. Holdridge et al.reach their greatest local densities in the (1971), Howell (1971), Janzen (1973), andinterior of undisturbed forest (Russell 1964, Hartshorn (1983) provide descriptions ofMonroe 1968, Land 1970, Petit 1991, some of the ecological life zones and habitatsWhitacre et al. 1993). Migrants comprise a of this region. Extensive research has beensmaller proportion of the total avifauna in conducted on migratory birds in certain areasundisturbed compared to disturbed forests. of Panama and Costa Rica, but not inFor example, in Belize (this study; n = 18 Nicaragua.sites), migrants represented 38% of species Whereas many migratory species indetected within disturbed forest habitats, but southern Central America can be found inonly 21% in undisturbed forests. Moreover, early second growth, forest edge, and openmangroves also appear to be underutilized forest, only a small ( < 20%) proportionby migrants in this region compared to areas appear to be dependent upon undisturbedoutside of Central America (Petit et al. 1993). mature forest (e.g., Willis 1980, Stiles 1985,


Blake et al. 1990, Blake and Loiselle 1992). received inadequate attention from avianFor example, based upon 28 moist forest ecologists.stands used in this study (incorporating bothvisual and mist net surveys), slightly-to- S th A .d 1 d . bed . d .au merlcamo erate y Istur sites supporte twiceas many migratory species (i = 8) than Natural vegetative formations in Southrelatively pristine forests (four species). America are highly diverse (for example,However, because more migratory species desert, open pampas, dry thorn forest,were found in mature moist forest (disturbed deciduous forest, rain forests, cloud forest,and undisturbed; i = six species) than in and barren paramo above 3000 m), so theeither early successional growth (4) or in low density and richness of migratory birdsnatural grasslands and pasture (3), the in this region probably are not due to lackimportance of forested tracts to migrants in of suitable habitats per se. General habitatsouthern Central America should not be relationships of many species inhabitingunderestimated. In fact, at least a dozen South America are reasonably well knownspecies use heavy forest cover in this region, but, because few formal surveys haveincluding Chuck-will's-widow, Olive-sided been conducted, a quantitative summary ofFlycatcher, Acadian Flycatcher, Wood habitat use by migratory species is difficult.Thrush, Northern Waterthrush, and Black- In an extensive survey of forests along anburnian, Worm-eating, and Kentucky elevational gradient in Peru, RobinsonWarblers (Ridgely and Gwynne 1989, Stiles et al. (1988) found that Neotropical migrantsand Skutch 1989, L. J. Petit, unpublished were most abundant in early successionaldata). and edge habitats along river systems.

Few studies or quantitative surveys have Orejuelaet al. (1980) and Pearson (1980) alsobeen conducted in mangroves in this region, found that migrants were underrepresentedbut the current literature review suggests that in undisturbed moist forests comparedfewer species use this habitat type in southern to dry forest and early second growth. EvenCentral America than in regions outside of within moist forest associations in SouthCentral America (Petit et al. 1993). In America, significantly more migratoryPanama, Northern Waterthrushes and species occurred on average in disturbed (sixProthonotary Warblers are highly dependent species) than undisturbed (four species)upon mangroves, whereas American Red- forests (this study, using both mist-nettingstarts and Black-and-white Warblers occur and visual survey data).there in small numbers (L. J. Petit, Most migratory species that overwinter inunpublished data). Natural grassland/ South America are associated with relativelysavanna and cattle pastures are avoided by tall, moist/wet forests, although, as alreadymigratory birds in Panama (Karr 1976, L. J. mentioned, these include moderately dis-Petit, unpublished data) and Nicaragua turbed sites, gaps and forest edges. Thus, the(Howell 1971). Many intensive studies on importance of maintaining these forests ismigratory and resident birds have been obvious; conversion to agricultural lands orconducted within several sites in this region pasture would remove the critical structural(e.g., Barro Colorado Island and Pipeline components of these habitats. In fact, aRoad, Panama; and Estacion Biologica La relatively large proportion of the species thatSelva, Costa Rica), but few simultaneous overwinter in South America have beensurveys of native habitat types have been identified in a companion paper (Petitinitiated. Moreover, Nicaragua is in dire et al. 1993) as being particularly suscept-need of ornithological investigation, and few ible to alteration of broadleaved forestsquantitative studies have surveyed highly [e.g., Chuck-will's-widow, Olive-sided Fly-disturbed habitats (e.g., agriculture, pasture, catcher, Veery, Prothonotary Warbler,forest edge, fragments) in any country in (mangroves), Blackburnian Warbler, Bay-lower Mesoamerica. Furthermore, the most breasted Warbler, Cerulean Warbler Black-remote areas in this region, such as Oarien poll Warbler, and Scarlet Tanager], high-and Bocas del Toro in Panama, have lighting the urgency of maintaining large


tracts of such habitats. Despite the large Preservation of native habitats in thisland area of tropical South America, region is particularly critical for severalrelatively few quantitative investigations of migratory species whose winter populationsoverwintering migratory birds have been are largely or entirely restricted to Caribbeanconducted there. Studies of the effects of islands; for example, Kirtland's, Prairie, Capedeforestation on migratory birds are urgently May, and Black-throated Blue Warblers.needed because many species that over- Because single natural or human-inducedwinter in South America are highly catastrophies, such as hurricanes, droughts,dependent upon undisturbed forest (Petit or forest conversion, can severely affectet al. 1993). species with limited geographic ranges,

migrants whose winter populations areC .bb I I d limited to the Caribbean (or are similarlyarl ean s an s ..

restncted on the maInland) may be the mostWith few exceptions, all major vegetation likely to undergo severe population declinesassociations found on the mainland are or extinction (Wunderle and Waide 1994).represented on Caribbean islands (see Lugo For example, Kirtland's Warbler, confined toet al. 1981). Arendt (1992) provided a several islands in the Bahamas and Greatersynopsis of the major "bird habitats" in the Antilles (Arendt 1992), exhibited majorregion. High human population densities on declines in breeding populations in Michiganmany Caribbean islands have led to severe following severe drought in the Caribbeandeforestation during the past two centuries during the previous winter (Radabaughand this situation poses special conservation 1974). Similarly, the presumed extinction ofproblems for the region, i.e., there are few Bachman's Warbler, also limited to severallarge tracts offorest remaining and developed Caribbean islands during winter, may beareas comprise a large portion of the total linked to relatively local events on theland area (e.g., McElroy et al. 1990). Two wintering grounds (Hamel 1986). Ornith-major reviews of habitats used by migratory ologists have devoted considerable attentionbirds in the Caribbean have been published to many of the habitats and islands in the(Arendt 1992, Wunderle and Waide 1993). Caribbean, such that ample informationThese and other studies are summarized is now available on the importance ofbelow. many habitat types to migrants. Because

Overall, migrant species richness and most forested habitats on many islandsabundance were higher in disturbed than in have already been extensively logged ormore natural habitats, but this trend varied converted to other land uses, perhapsgreatly among islands and sites. For example, conservation efforts and research in theon Cuba and the Dominican Republic, moist Caribbean should be directed at assessingforest habitats contained the greatest number the effects of agricultural and urbanof species (as estimated by rarefaction), but land uses on populations of migratoryon Jamaica, dry forest and shade coffee birds (see also Arendt 1992). Protectionplantations supported the most species of migratory (and resident) birds on densely(Wunderle and Waide 1993). The most populated Caribbean islands depends inconsistent pattern from the Caribbean large part on identifying and implementingislands is the high use of coastal mangroves land-use practices that are compatibleby migrants. A survey of ornithologists by with both economic goals and preservationArendt (1992) revealed that > 50% of of natural vegetation types. Arendt (1992),migratory landbirds on Caribbean islands in collaboration with other bird experts fromused mangroves to some extent during the region, outlined four major require-winter. In the Greater Antilles, Yellow- ments necessary to achieve that goal in thebellied Sapsucker, Black-and-white Warbler, Caribbean: establishment of protectedAmerican Redstart, Magnolia Warbler, and reserves; restoration of degraded habitats;Northern Waterthrush reached their greatest legislation and education, and ecologicaldensities in mangroves (Wunderle and Waide research (also see Wunderle and Waide1993). 1994).


GEOGRAPHIC VARIATION IN COMPOSITION migratory species are shared between forest

OF MIGRANT ASSEMBLAGES sites in western Mexico (Hutto 1980, 1992)

and sites in the northern Yucatan Peninsula

The similarity of winter migrant assemblages (Lynch 1989), but the latter area shows a

occupying broadleaved forest tends to relatively high faunal resemblance with

decrease with increasing distance between migrant communities in Costa Rica and

sites (Fig. 6- 7). However, similarity of Panama (Table 6-2, Fig. 6- 7). In general,

assemblages is not always a simple function similarities between the composition of

of geographic proximity. For example, few winter migrant communities in northern

Central America and Mexico, where most

migratory species overwinter, are much

100 greater from NW-SE along the Gulf-

~ Caribbean coastal plain than they are from

OC 80 E- W between the Atlantic and Pacific

~ coastal plains (Table 6-2). In these regions,

~ 60 distinct migratory bird communities may

C;; arise because of restrictions placed by tall

1- 40 mountain ranges on movement of in-

ffi dividuals. Alternatively, differences in bird

~ 20 community composition among these and

~ other regions in the Neotropics may reflect

variation in vegetation composition as-

02.5 3.0 3.5 4.0 sociated with climatic changes due to major

LOG DISTANCE (km) montane barriers or to latitudinal gradients.

10 For example, migratory species that breed in

Figure 6-7. Relationship between the similarity the dry and montane portions of western

(coefficient of community; Pielou 1975) in the North America tend to overwinter in the dry

composition of migratory bird assemblages in and montane regions of western Mexico

forest and the distance (km) between sites. and Guatemala. In contrast, the humidEquation for regression line: percent similarity = Atlantic coastal slope of Mexico and Central

265.7- 70.1 (I.oglo ?ist.ance);.R2 = 0.74. Squ~res America tends to be inhabited by over-

repr~sent relatlonshll?s mvolvln.g western Mexlc?, wintering migrants that breed in mesic

and Illustrate that migratory bIrd assemblages In h b. t f t N th A .(H ttwestern Mexico differ from other locations in a Ita s o eas er~ or menca u o

Latin America and the Caribbean more than 1985). Thus, specIes may tend to occupy

would be expected simply based on their similar veget~tio.n. types throug~out the

geographic proximity. year, and avaIlabIlIty of those habItats may

Table 6-2. Coefficients of community (Pielou 1975) for pairwise comparisons among nine forested sites in

the Neotropics. Possible values range from 0 (no species in common at the two sites) to 100 (all species

occur at both sites).~ ,

Site Site Number-

1 2 3 4 5 6 7 8 9

,St John's, Virgin Islands (1) -

Western Mexico (2) 16Yucatan, Mexico (3) 35 46Los Tuxtlas, Mexico (4) 37 46 88Belize (5) 46 25 78 75La Selva, Costa Rica (6) 25 32 65 68 68Santa Rosa, Costa Rica (7) 25 22 61 60 68 77Panama (8) 27 29 55 54 71 81 71 -Peru (9) O 8 6 5 6 6 6 7 -

'-' Data sources: Stiles (1980), Willis (1980), Coates-Estrada et al. (1985), Lynch (1989, 1992), Terborgh et al. (1990), Hutto (1992), Askins

et al. (1992), and D. R. Petit (unpublished data).


restrict the areas where a species can spend ecosystem-level approach (e.g., Hut to et al.the winter. 1987, Grumbine 1990) has been proposed.

Factors associated with the breeding range Despite the widely divergent practicalof a species may also constrain where it application of these three concepts, however,overwinters in the Neotropics. For example, all are based upon the idea of minimizationmost species that breed in western North of the number of management prescriptionsAmerica tend to overwinter in the dry (hence monetary cost and labor) necessaryhabitats of western Mexico and the Pacific to protect habitats and species. Typically, asversant of northern Central America despite the ecological (or taxonomic) similaritythe great variability of habitats used by those increases among species assemblages, so toospecies during the breeding season. This does the specificity and effectiveness ofobservation argues against vegetation being management plans. Hence, development ofthe sole determinant of winter ranges. conservation plans on a region-by-habitatVariation in bird communities among basis may be a cost-effective and ecologicallyregions may also relate to historical factors, sound approach to the protection ofand not to present-day characteristics of overwintering migratory birds. Althoughthese landscapes. For example, Kirtland's formation of a single plan to cover eachWarbler overwinters exclusively on Carib- habitat type might be more efficient thanbean islands, often in coastal scrub and dry multiple regional strategies, disparities inforest habitats similar to those found on the bird species composition among regionsYucatan Peninsula of Mexico. Furthermore, could seriously reduce the utility of such anoverwintering Kirtland's Warblers are approach because of the disparate ecologicalrestricted to several islands in the Bahamas requirements of species comprising thoseand Greater Antilles (Arendt 1992), although assemblages. An argument could be madedry scrub and forest are found on many that a single plan might be effective for allnearby islands. moist forests if a "conservative" approach

Composition of winter migrant assem- were adopted; i.e., one that was implementedblages in South America appears to be for the most "disturbance-sensitive" species.particularly distinct from that in other However, conservation biologists mustregions of the Neotropics (Table 6-2). The justify their recommendations convincingly,similarity between the migratory avifauna especially in countries where there is greatoccupying forested sites in Central America/ pressure on the land base. A conservativeCaribbean and the migrant assemblage at plan that is not fully applicable in someCocha Cashu (Peru) was independent of regions, although it may be in others, is notgeographic proximity. In contrast, migratory constructive in merging conservation effortsbird composition on St John's in the with long-term economic and social needs ofCaribbean Sea was more similar to that developing nations (Geerling et al. 1986, Gowfound on the mainland, varying as a function 1992).of distance between sites (Table 6-2).

These analyses suggest that geographicproximity is a good predictor of similarity in HABITAT USE AT THE SPECIES LEVELthe composition of migratory birds in-habiting tropical broadleaved forests, es- Documentation of geographic and regionalpecially within Central America and eastern patterns of habitat use for migratory birds asMexico. As such, it may provide justification a group provides critical information onfor a region-wide approach to conservation those habitats and areas that are mostand management of forest-dwelling birds in important for overwintering migrants, but itthe Neotropics. Most current nongame offers little insight into which habitats aremanagement efforts in North America used by individual species. Clearly, speciesare aimed at large assemblages of species, exhibit unique responses to available habitatsoften indirectly through indicator species during winter, much as they do during the(Salwasser et al. 1983) or ecological guilds breeding season. Recognition of habitat(Severinghaus 1981). Recently, a direct associations of individual species allows


identification of those species not adequately Table 6-3. Forty-five long-distance migratoryprotected under general conservation and landbirds most likely to be negatively affected bymanagement plans. While many researchers destruction of tropical forests. Information washave noted the urgent need to document taken from Rappole e~ al. (1983), Diamond (199.1),habitat needs of migratory species during ~eed (1992), and Petit et al. (1993). Onl~ species

lIsted by more than one author are IncludedwInter, surprisIngly .little. quantItatIve data (but see footnote). The list of species ishave been collected In this area (Greenberg incomplete because little information on winter1986, Hut to 1989, Lynch 1989). The lack of habitat use exists for certain families (e.g.,an objective, quantitative approach to Apodidae, Trochilidae, Hirundinidae) and becauseclassification of winter habitats by avian all authors did not necessarily consider all speciesecologists has resulted in confusion over the of migratory land birds in their summaries (e.g.,habitat associations of migratory species, Reed 1992).

thereby making conservation recommenda- Black-bi1Jed Cuckoo Black-throated Bluetions dubious (see, for example, Martin 1992). Chuck-will's-widow WarblerFuture research on habitat associations of Whip-poor-will Townsend's Warbler

. h Id ..Yellow-bellied Sapsucker Black-throated Green

migrants s ou co~centrat~ on quantitative Olive-sided Flycatcher Warbler

assessments of habItat requirements, follow- Western Wood-pewee Golden-cheeked Warblering, for example, the approach of Blake and Yellow-bellied Flycatcher Blackburnian WarblerLoiselle (1992), Petit (1991), Powell et al. Acadian Flycatcher Kirtland's Warbler(1992) and Wunderle and Waide (1993). Great Crested Flycatcher Bay-breasted Warbler

, Veery Blackpoll Warbler"A review of habItat use by each species IS Gray-cheeked Thrush Cerulean Warbler

beyond the scope of this chapter. However, Swainson's Thrush American Redstartmany avian ecologists believe that the Wood Thrush. Prothonot.ary Warblerlong-distance migratory species currently Bla~k-capJ:1ed VIreo Wo~m-ea~mg Warbler

..SolItary VIreo Swamson s Warblermost susceptible to severe population Yellow-throated Vireo Ovenbird

declines are those that depend upon Philadelphia Vireo Northern Waterthrush"undisturbed forests during the boreal winter, Blue-winged Warbler Louisiana Waterthrushas well as during the summer (e.g., Terborgh Gol~en-Winged Warbler Kentuck.y Warbler

...Collma Warbler ConnectIcut Warbler1989). In light of the. need t? Identify th?se Lucy's Warbler Hooded Warbler

species that may require special conservation Chestnut-sided Warbler Canada Warblerattention, a list of those migrants that are Magnolia Warbler Scarlet Tanagergenerally thought to be in the greatest danger Black-headed Grosbeak

of population declines due to destruction of "Includedonlyonlhe.highlyvulnerable"listorpetitetal.(1993).forests on the wintering grounds is presentedbelow; the quantitative methods used toderive this list are presented in Petit et al. population declines of migratory birds: many(1993). The 45 species identified in Table 6-3 species tend to breed and overwinter in therepresent 42% of the 107 Neotropical same general habitat type, thereby hamper-migratory species that overwinter primarily ing partitioning of habitat-based events (e.g.,in mature phases of tropical forests (Rappole mortality) between seasons. Petit et al. (1993)et al. 1983). This is a conservative list, showed that migratory species highlyhowever, because not all taxa were equally restricted to undisturbed broadleaved forestsevaluated because of lack of information during winter had average breeding seasonabout winter habitat requirements (e.g. population declines more than eight timesswifts, swallows). Members of the families that exhibited by forest-dwelling migrantsTyrannidae, Muscicapidae, Vireonidae, and that were more tolerant of winter habitatEmberizidae (Parulinae) comprised nearly disturbance. Robbins et al. (1989a), also usingthe entire register, the same taxa (and species) Breeding Bird Survey data, suggested thatthat typically are included on similar lists events on the wintering grounds were thederived from habitat studies during the main cause of population declines. However,breeding season. Evident, then, is one of the Hussell et al. (1992), using an independentmajor problems confronting ecologists data set collected during migration, found noattempting to unravel the cause(s) of firm evidence to support that conclusion.~


Continuation or initiation of large-scale individual migratory species (see Petit et al.studies designed to monitor bird populations 1993). Reports derived from study of singleand, especially, habitat conditions in vegetation types were not used, as thosebreeding, migration, and wintering areas are provided no information on habitat "selec-needed to understand better the relative tion," which is defined operationally aseffects of seasonal events on populations of differential abundance of a species in one ormigratory birds (e.g., Powell et al. 1992, more habitats relative to others. Data fromSherry and Holmes, chapter 4, this volume, some studies were presented qualitativelyMoore et al., chapter 5, this volume). Equally (e.g., Land 1970, Stiles and Skutch 1989) andimportant are intensive, small-scale studies others quantitatively (e.g., Waide 1980,that examine the relative seasonal effects on Lynch 1989, Blake and Loiselle 1992, Hut tolocal population dynamics (e.g., Holmes et 1992). Habitats used outside the winteringal. 1986, 1989). Large-scale studies provide period (e.g., during migration) were excludedbroad, long-term trends that are important when possible. Thus, for each speciesin assessing the scope of the problem, whereas included in a given study, two types oflocal studies may provide better insight into information were extracted: (1) all "possible"the potential causes of those trends. habitats where the species could have

Even at the species level, habitat use can occurred in the study area (defined by thevary considerably among locations and author and including only those that wereindividuals, thereby increasing the difficulty actually surveyed; and (2) the habitat(s) thatin managing habitats for species that have contained a disproportionate number ofbeen recognized as susceptible to habitat individuals or that were stated by the authoralteration. Variation in habitat use and ways to be "preferred" by the species. A total ofin which that variability can be incorporated 11 habitats were identified for this analysis:into conservation efforts are discussed below. moist forest, dry forest, cloud forest,

mangroves, advanced second growth, forestG h. V. ..H b. U b S .edge, open woodlands/forest, coniferous/

eograp IC arlatlon In a rtat se y pecles mixed coniferous forest, early second

Habitat use by most migratory species is growth, residential, and grassland/pasture.believed to be consistent across winter Thus, for each species the number of bothranges. For example, a species that occurs in "available" and "occupied" (preferred) sitesmoist, mature forest in eastern Mexico will for each habitat type were derived. Thetend to occupy structurally and climatically proportion of available sites that weresimilar (though floristically different) habitats occupied (occupied/available) provided anin Belize, Costa Rica, or Panama. Verifying indication of the extent of use of each habitatgeographic consistency in habitat use is type while controlling for the differences indifficult however because of the lack of habitats available across studies and,quantitative data within different regions for geographic regions. For each habitat type formost species. each species, proportions ranged from 0

Understanding geographic variation in (species never occurred there or was lesshabitat use has obvious conservation common than expected) to 1.0 (speciesimplications. Most importantly, it allows for always occurred there with relatively highidentification of appropriate spatial scales at abundance). Six regions (identified above)which conservation and management plans were used to examine geographic variationcan be enacted (see above; also see Van in habitat use.Home and Wiens 1991). Martin (1992) Some species showed little variability inprovided one example of the danger habitat use across regions. For example,associated with assuming geographic con- Common Yellowthroat and Summer Tanagersistency in habitat use. were relatively consistent in their distribu-

To address the issue of geographic tions across habitats throughout their rangesvariation in habitat use, information was (Fig. 6-8). Other species, however, exhibitedsummarized from approximately 50 publica- extensive variation, which suggests thattions that examined habitat affiliations of disagreement among avian ecologists about


COMMON YELLOWTHROAT SUMMER TANAGER

(I)w15100 100:)t; 80 80

LLO 60 60

~ 40 40W~ 20 20Wc.. 0 0

SCRUB YNG FOR MANOR FOR EDGE DRY FOR SCRUB YNG FOR WOODS DRY FOR WET FOR

HABITAT TYPE HABITAT TYPE

Figure 6-8. Geographic variation in habitat use by Common Yellowthroats and Summer Tanagers. Theordinate represents the number of studies in which a species was found to occur in a given habitat typedivided by the number of studies in which each species could have occurred in that habitat. See text fordetails.

habitat requirements of certain migrant Potential reasons for geographic variation inspecies in the tropics (e.g., Greenberg 1986, habitat use include different availabilities ofMartin 1992) probably has been to a large habitats, and regional differences in pre-extent a result of geographic variation in ference for certain habitat types. Thetheir habitat use and not necessarily Black-throated Green Warbler, for instance,because of incompatibilities of various is commonly found in pine stands in northernsurveying methods (e.g., see Morton 1992). Central America and on Caribbean islands,For example, Gray Catbirds and Northern but occupies broadleaved forest associationsWaterthrushes both exhibited pronounced in eastern Mexico and Southern Centralregional differences in habitat use (Fig. 6-9). America, where pines are scarce (this study).

GRAY CATBIRD NORTHERN WATERTHRUSH

~ .CENTRAL AMER ~CARIBBEAN DMEXICO .CENTRAL AMER ~CARIBBEAN15100 100 DMEXICO &1SOUTHAMER:)t; 80 80

'6 60 60

~ 40 40w~ 20 20wc.. 0 0

SCRUB MANGR DRY FOR CONIF WET FOR SCRUB

HABITAT TYPE HABITAT TYPE-

Figure 6-9. Geographic variation in habitat use by Gray Catbirds and Northern Waterthrushes. Theordinate represents the number of studies in which a species was found to occur in a given habitat typedivided by the number of studies in which each species could have occurred in that habitat. See text fordetails.


Black-throated Green Warblers at the edge mendation that conservation and manage-of their winter range in western Mexico, ment plans for migratory bird habitathowever, make extensive use of both should incorporate a regional (or local)broad leaved cloud forest and mixed pine perspective.forest (Hut to 1992, R. L. Hut to, personal Geographic variability in habitat use mayobservation). Louisiana and Northern also indicate localization of individuals ofWaterthrushes provide other examples of different subspecies (see Ramos and Warnerhow availability of a given habitat type 1980). For instance, the Yellow-rumpedin a region might influence habitat use. Warbler occurs in two rather differentIn continental areas, waterthrushes generally situations within its Middle Americando not occupy residential areas during wintering range. In the interior of Mexicowinter but, on densely populated and and Central America, this species is typicallydeveloped Caribbean islands [compare encountered in high-elevation pine-oak-firland-use and population data in Leonard forests, although it occurs less commonly at(1987, Table A.2) with that in McElroy lower elevation in thorn-scrub and disturbedet al. (1990, Table 21.1)], this species uses urban and agricultural areas (Binford 1989,residential gardens (this study; see also Hut to 1992, R. L. Hut to, personal observa-Arendt 1992). Furthermore, for the migratory tion). In the Mexican Yucatan, which lacksbird community as a whole, certain habitats tall mountains, Yellow-rumped Warblers areare of high regional importance, e.g., characteristic of grassy coastal dunes andmangroves in the Caribbean and cloud forests extremely open disturbed habitats (includingin western Mexico and South America (Fig. recently burned sites) at or near sea level6-10), suggesting that regional availability (Russell 1964, Lynch 1989,1991,1992, Lopezmay have influenced the evolution, or at least Ornat and Lynch 1991). The species alsopatterns, ofbird-habitat relationships. These occurs sporadically in low-elevation grass-data corroborate the previous recom- lands in Panama (Karr 1976). Virtually all

(/) ~EAST MEXICO ~WEST MEXICO

~ 1 00 D CARIBBEAN fZJ CENTRAL AMERICA (N)

O .CENTRAL AMERICA (S) D SOUTH AMERICA:)1- 80(/)

u.O 60

1-Z 40W()c: 20Wa..

0 MANGROVE DRY FOR WET FOR CLOUD FOR

HABITAT TYPE

Figure 6-10. Example of geographic variation in habitat use by migratory birds as a group. Theordinate represents the number of studies in which a species was found to occur in a given habitattype divided by the number of studies in which each species could have occurred in that habitat,summed over all migratory species. See text for details.


of the Yellow-rumped Warblers that Sex- and Age-Specific Habitat Useoverwinter in the interior mountains ofOaxaca, Mexico, are members of the western Habitat use by some migrants varies by sexsubspecies, (D. c. auduboni), formerly know or age of the individual, such that overallas " Audubon's Warbler" (Binford 1989). In habitat use by a species reflects the combinedcontrast, Yellow-rumped Warblers that distribution of all age and sex classes. Lynchoccur in the Yucatan belong to the race D. et al. (1985) showed that male and femalec. coronata, formerly known as the "Myrtle Hooded Warblers in the Yucatan regionWarbler". In Honduras, the two races of differed in selection of winter habitat: malesYellow-rumped Warbler also tend to tended to occur in relatively tall, closed-segregate geographically, with D. c. auduboni canopy forest, while females were most oftenoccurring mainly in montane pine-oak found in old fields, native coastal scrub,habitats, and D. c. coronata overwintering treefall gaps, and other disturbed, low-staturealong the Caribbean coast and on offshore vegetation. Furthermore, laboratory andislands, with occasional inland occurrences field experiments suggested that this differ-(Monroe 1968). Separation of subspecies ence reflects active choice, not merelyalong geographic/habitat lines breaks down exclusion of females from forest by thein the highlands of western Guatemala, where dominant males (Morton et al. 1987) andan isolated breeding population of D. c. that sexual differences in habitat selection inauduboni coexists with migratory D. c. Hooded Warblers is innate (Morton 1990).coronata during winter. However, the In other species, however, behavioralAtlantic coastal plain of Guatemala is dominance by males appears to driveinhabited exclusively by D. c. coronata (Land sex-related patterns of habitat use (Wunderle1970). Thus, in some parts ofits winter range, 1992, Marra et al. 1993). Sex-specific habitatthe Yellow-rumped Warbler seems to offer use has been documented for about ten otherone of the few known examples of geographic migratory parulines (Lopez Ornat anddifferences in habitat use that correlate with Greenberg 1990, Wunderle 1992, Wunderlegenetic differentiation. Regional variation in and Waide 1993) and, in most cases, maleswinter habitat use by migratory birds may occur in more mature habitats. Despite itsreflect segregation of individuals from potential importance, the evolutionarydistinct breeding populations, whose winter significance of habitat partitioning by sexeshabitat preferences may be related to has not received much attention and littlebreeding habitat. attempt has been made to incorporate it into

Geographic variation in habitat use could conservation recommendations. For example,result as an outcome of selective pressures if habitat use reflects preferences of the sexesthat vary regionally. For example, geo- (Morton 1990), then all habitats used aregraphic differences in habitat use could be important to the species and should beinfluenced by regional differences in assem- maintained. However, if sex-specific habitatblages of competing species ( e.g., Bennett patterns result from dominance relationships,1980, Keast 1980), characteristics of the food then habitats used by the subordinate sexbase supported by vegetative associations might be suboptimal, and conservation(Janzen 1973), or predation pressure efforts should focus on increasing or(Buskirk 1976). Whether geographic differ- improving those habitats used by theences in habitat use reflect variation in these dominant sex.factors or others (e.g., population density, Age-related exclusion of some individualsclimate) is unknown. However, at this stage from certain habitats (or actual preferencesof conservation efforts, knowledge of for those habitats) also may occur on thegeographic variation in habitat use at least wintering grounds, although evidence isallows conservation planners to recognize sparse. In Belize, L. J. Petit et al. (unpublishedthat variability does exist so that it can be data) found that young ( ~2 years) maleincorporated into management and con- American Redstarts occupied habitats thatservation prescriptions. were intermediate between those of females

and older (>2 years) males (Fig. 6-11).


(/)-J .OLDER MALES« 50 ~YOUNG MALES~ OFEMALES

~40OZ-30u.O1-- 20

Zw010a:Wa.. 0 SCRUB PJNE-SAV FOR EDGE FOR INT PINE FOR

HABITAT TYPE

Figure 6-11. Sex- and age-specific habitat use by American Redstarts overwintering in five habitat typesin Belize, Central America. Age classes for males represent individuals ;.02 years old (older males) and<2 years old (young males). Ages of females were not determined in most cases. Data are based uponvisual/auditory surveys conducted during winters 1988-1989 and 1989-1990 (L. J. Petit et al., unpublisheddata).

Behavioral observations suggested that second growth, as compared with a ratio ofage-related habitat use may reflect the 3.4 : 1 in the eastern portion of the Peninsula,outcome of behavioral dominance of older where the predominant vegetation isindividuals, as observed during the breeding closed-canopy semi-evergreen forest. How-season (Sherry and Holmes 1989, Marra et al. ever, because the sampling scheme in that1993). study was not random across all habitats and

These findings raise two important regions, more data are needed to confirm thisquestions. First, do sexual and age differences possible geographic gradient in sex ratio.in habitat use translate into differences in the The second question is whether numbersgeographic distributions of males and of males or females are differentially limitedfemales, and age groups? Koplin (1971) by the availability of habitats on thedemonstrated that habitat differences be- wintering grounds. This possibility wouldtween wintering male and female American most likely arise in situations where a speciesKestrels were correlated with a strong exhibits a strong degree of sexual habitatgeographic gradient in the sex ratio. segregation (such as the Hooded Warbler)However, whether sexual habitat differences and human-related habitat destructionwere driving sexual differences in geographic greatly reduced the preferred habitat of onedistribution, or vice versa, was not estab- or the other sex. Recent observations oflished. Pearson (1980) also described a effects of hurricane and fire on overwinteringpossible latitudinal gradient of age classes Hooded Warblers suggest that sexual habitatfor overwintering Summer Tanagers. For differences are variable over space and timeHooded Warblers, Lynch et al. (1985) found (Lynch 1992, personal observation), but thea male: female ratio of 2.0: 1 in the generality of these observations is unknown.northwestern Yucatan Peninsula, which is Sex- and age-specific habitat use hascharacterized by lower, semideciduous important ramifications for conservation of


Neotropical migrants. First, partitioning of period (Rappole et al. 1989, Loiselle andwinter habitats could lead to differential, Blake 1991 ), such that patterns of habitat usehabitat-based mortality rates either due to may also covary. For example, migrants haveintrinsic factors or to more extensive loss of been shown to respond to changing fruitcertain habitats than others. The resulting abundance throughout the winter in theskewed sex ratios would reduce the effective tropics (Leck 1972a, Greenberg 1981). Thesize of breeding populations and could result is a shift in habitat use as the winterdramatically enhance the probability of local progresses, with migrants sometimes con-extirpation (Shaffer 1981). For example, cent rating in early second growth and otherSherry, Holmes and coworkers (Holmes disturbed habitats where fruit can beet al. 1989, Sherry and Holmes 1989, Marra abundant (Martin 1985, Loiselle and Blakeet al. 1993, Parrish and Sherry 1994) have 1991; also see Blake and Loiselle 1992).articulately documented the territorial Knowledge of the importance of suchsystem, habitat use, and survival of American "landscape complementation " (Dunning et

Redstarts overwintering in Jamaica. They al. 1992), where multiple habitats within afound that male dominance was the basis for landscape are used by an individual (orsex-specific habitat distribution, with males population) for different purposes, is criticaloccupying the more seasonally stable, to the understanding of how best to managefood-rich mangroves and females residing in and conserve wildlife populations. Forscrub habitats that decreased markedly in example, species such as Bay-breastedquality throughout the winter period. Marra Warbler and Wood Thrush may spend mostet al. (1993) suggested that differences in of the winter in mature forest, but shift towinter habitat quality, combined with sex- younger second growth during late winter tospecific habitat use, could account for the fulfill their energetic needs for migrationmale-biased sex ratio in breeding populations (Martin 1985). Thus, both habitats and theirof redstarts (Sherry and Holmes 1991). proximity to each other may be importantClearly, more investigation of this type is for overwinter survival and for returning toneeded to identify those factors that limit north temperate breeding grounds. In thispopulations of Neotropical migrants on their light, a landscape-Ievel approach to manag-wintering (and breeding) grounds (see ing tropical habitats for wildlife would seemHolmes and Sherry 1992, Sherry and Holmes, warranted. Furthermore, these observationschapter 4, this volume). also may have consequences for the way in

Second, placement of reserves must which researchers gather and examineincorporate potential sex-based dichotomies habitat data: information collected over onlyin habitat use by assuring that appropriate a portion of the winter, or collected over thehabitat for both sexes is adequately entire winter and then lumped together, mayrepresented. Finally, if innate or dominance- overlook important temporal variation inbased habitat use is manifested in distinct habitat requirements of species (Greenberggeographic winter ranges for sex or age 1992b). Clearly, the ecological and evolu-classes (e.g., Ketterson and Nolan 1983), tionary correlates of habitat use by migrantsthen development of protected reserves or is central to development of long-termapplication of appropriate land-use practices conservation plans. Incorporation of prin-must be implemented throughout entire ciples of behavioral ecology into researchwinter ranges of species. programs may heighten understanding of the

ultimate and proximate bases for patterns ofT I V . t .. H b.t t U habitat use by migrants in the Neotropics.

empora aria Ion In a I a se

Many migratory species are territorial duringwinter and, therefore, temporal stability in BEHAVIORAL ASPECTS OF HABITAT USE ANDhabitat use would be expected. However, in MIGRATORY BIRD CONSERVATIONapparent response to food availability,several species are known to range over Species-habitat relationships are manifesta-extensive areas during the overwintering tions of evolutionary processes that have


shaped the ecology of species to exploit their Freemark 1983, Loiselle and Blake 1991), theenvironments effectively. Thus, general presence of corridors (sensu MacClintockrelationships between the behavior (Holmes et al. 1977) and adjacent suitable habitatset al. 1979) and morphology (e.g., Miles and may greatly facilitate movements amongRicklefs 1984) of species and the micro- areas. Rappole et al. (1989) radiotrackedhabitats that they occupy are not unexpected. Wood Thrushes in eastern Mexico andHowever, close association between the showed that individuals that wanderedbehavior or morphology of species and broad suffered greater mortality from predatorshabitat types are much less evident due to than those individuals that held stablethe diverse assemblages of species occupying territories. They predicted that mortality ofhabitats at that scale; i.e., a relationship wintering Wood Thrushes would increase asbetween the "average" behavior or morpho- tropical deforestation continued because itlogy of local assemblages and habitat type would force more individuals to wander inare not well pronounced. Thus, basing search of food (territories). Forest destruc-conservation plans simply on the habitats tion reduces the connectivity of landscapespecies use may not adequately protect all elements (Forman and Godron 1986), furtherspecies within a given area because of the inflating the hazards faced by Woodunique ecological needs of each species found Thrushes moving from one patch to another.there. Firm understanding ofhow to conserve Additionally, several migratory species areand manage habitats for Neotropical believed to move from mature forest intomigrants during winter requires knowledge younger second growth during late spring inof the ways in which individuals search for order to exploit the abundant fruit there (e.g.,and secure insects and fruit, the types of food Martin and Karr 1986). Thus, appropriatetaken, and the significance of different social composition of landscapes may be necessarysystems. Below, these issues are reviewed and for some migratory and many residentways in which such information can be frugivores and nectarivores to completeapplied to conservation efforts are identified. their annual cycles (Loiselle and Blake

Most Neotropical migrants are highly 1991).insectivorous. The evolution oflong-distance The manner in which insectivorousmigration traditionally has been explained migrants search for and capture arthropodby year-round reliance on insects (Griffin prey varies widely among species, includ-1964; see Levey and Stiles 1992 for alternative ing gleaning from foliage (e.g., Northernview). More recently, however, research on Parula, Chestnut-sided Warbler), hoveringbehavioral ecology of migrants has revealed at leaves (American Redstart, Yellow-that up to one-third of all long-distance rumped Warbler), gleaning from trunks andmigratory landbirds consume some fruit branches (Yellow-bellied Sapsucker, Black-during winter (e.g., Leck 1972b, Morton and-white Warbler), flycatching (Acadian1980, Greenberg 1981, Rappole et al. 1983, Flycatcher, Olive-sided Flycatcher), aerialBlake and Loiselle 1992) and that fruit-eating pursuit (Chuck-will's-widow, Chimney Swift),species, in general, display winter social searching ground substrates (Wood Thrush,systems different from those of many highly Ovenbird) and aerial leaf litter (Worm-eatinginsectivorous species. Unlike many insec- Warbler, Blue-winged Warbler), probingtivores, total or partial frugivores often do flower parts (Cape May Warbler, Orange-not defend territories over extended periods crowned Warbler), gleaning from openof time, but instead range widely in search of ground (Palm Warbler, Louisiana Water-fruit (Levey and Stiles 1992). Thus, thrush) or water (Northern Waterthrush),abundances of frugivorous migrants (and and gleaning from herbaceous vegetationresidents) in a given habitat can vary greatly (Common Yellowthroat). Examination ofover time in response to fruit availability avian foraging behavior can provide insight(Loiselle and Blake 1991). Furthermore, into the ecology of overwintering migrantsbecause those movements often are not and might allow extension of conservationlimited to a single habitat patch, habitat type, efforts beyond simple habitat-based plans toor even watershed (Karr et al. 1982, Karr and dissection of the reasons why some species


or more vulnerable than others to conversion population dynamics. For instance, a reserveof native tropical landscapes. For example, system comprised of many small forestcertain forestry practices in North America patches connected by corridors may be morehave a differential effect on forest-dwelling suitable to frugivorous and leaf-gleaningbird Ispecies, and much of that variation has insectivorous migrants than to ground-beed associated with foraging behavior or foraging insectivores because a small forestother aspects of life history (Franzreb 1983, patch may not adequately buffer ground leafWebb et al. 1977). Similarly, modifications of litter from the desiccating effects of increasedhabitat structure in the Neotropics may wind and temperature; see results of Johnsresult in unique responses by foraging guilds (1989) and Thiollay (1992) above. Inter-of both migrants and residents. Thiollay estingly, of Neotropical migrants typically(1992) showed that selective logging in considered to be forest dwellers duringFrench Guiana resulted in a decrease in the winter, ground foragers tend to be morenumber of resident ground foragers and closely associated with undisturbed forestsflycatchers, but an increase in density of small and less with other habitats than species infrugivores. Johns (1989) found a similar effect other foraging guilds (e.g., Robbins et al.of logging on avian foraging guilds in a 1987, Greenberg 1992a, Lynch 1992, Petit etMalaysian rain forest (but see Wong 1986). al. 1993, Wunderle and Waide 1993).

These observations suggest that over- As stated above, the degree of territorialitywintering migratory birds could be indirectly exhibited by migratory species is often relatedaffected by forest disturbance through to their diet; insectivores tend to hold morealteration of food resources. Anthropogenic stable territories (Sliwa and Sherry 1992).disturbances may differentially affect abun- Most insectivorous migrants defend smalldances of insects and fruit in a habitat, and territories throughout the winter. But manythereby impact populations of insectivorous insectivores (and some frugivores/omnivores)birds differently than frugivorous birds. For also join mixed-species flocks as eitherexample, isolation of a moist tropical forest obligate or, especially, facultative membersmay lead to drier microclimates within the (join flocks only when they pass through anpatch (Kapos 1989, Williams-Linera 1990), individual's territory). Examples of migrantswhich, in turn, has been shown to be related that are often found in mixed-species flocksto lower abundances of arthropods in foliage include Solitary Vireo, Magnolia Warbler,(Janzen and Schoener 1968) and especially, Black-and-white Warbler, Worm-eatingleaf litter (Strickland 1947, Plowman 1979). Warbler, Northern Parula, Townsend'sAlternatively, fragmentation enhances pene- Warbler, Hermit Warbler, Olive Warbler,tration of sunlight (because of a larger edge: American Redstart, and Blue-gray Gnat-interior ratio), which has been related to catcher. Many migratory parulines inincreased fruit density (e.g., Blake and western Mexico appear to be obligate flockHoppes 1986, Levey 1988). Thus, some members (Hut to 1987, 1988b). Flocks in thefeeding guilds may be better able to withstand Neotropics typically have both "core" (ordisturbances than others. (Forest reserves "nuclear") species (Moynihan 1962) andcannot be so small, however, as to inhibit "attendant" (or "satellite") species (Randnormal production and dispersal of fruit.) 1954). Core species are required for a flockLikewise, foraging generalists (sensu Morse to form and contribute to its cohesiveness,1971) are thought to be more resilient to while attendant species tend to follow coredisturbances and environmental unpredict- individuals. Flocks occupying forests situatedability than are specialists (Glasser 1982, on the mainland almost always have resident1984). core species, whereas migrants are typically

These examples illustrate the complex attendants (Buskirk et al. 1972, Powell 1979,relationship among behavioral traits, habitat Tramer and Kemp 1980). However, inuse, and disturbance (Gordon 1991), and lowland deciduous forest of western Mexico,highlight the need to consider the behavioral migratory Nashville Warblers occupy theecology of migrants in projecting the effects role of nuclear species (R. L. Hut to,of land-use practices on migrant winter unpublished manuscript). In the Caribbean,


flocks comprised of both migrants and conservation designs are futile unless ampleresidents are uncommon (Eaton 1953, Lack suitable habitat exists or can be restoredand Lack 1972, Post 1978, Ewert and Askins within landscapes. In this chapter and1991). elsewhere (Petit et al. 1993), a general

Proposed advantages of flocking center overview of the types of land-use practicesaround foraging efficiency and predator that are compatible with migratory birdavoidance (Powell 1988). The presence of ecology in the Neotropics has been presented.flocks may be important for overwinter These arguments have stressed that, if globalsurvival of certain species of migrants in some populations of migratory birds are to beregions because of the benefits accrued by sustained, habitats in wintering (as well asflock members. Thus, understanding flocking breeding and migration) areas must bebehavior could provide supplementary protected.information for conservation of migratory Much of the focus of tropical conservationbirds. Participation by migratory birds in has been directed at two broad forestmixed-species flocks is dependent upon the associations: wet/moist and dry broadleavedpresence of core species, which may be more forests. Together, these two forest typessensitive to forest alteration and fragmenta- originally comprised more than two-thirds oftion than are migrants because core species all land area in the Neotropics (Murphy andare typically residents. For example, forest Lugo 1986, Myers 1991). Clearly, preserva-canopy flocks in Peru preferred areas offorest ti on of these forest types should remain atwith unbroken canopies (Munn 1985). In the forefront of conservation efforts, if onlyaddition, some resident core species have because of their extensive geographicrelatively large home ranges (often ~ 5-10 ha; coverage. However, within these major forestMunn 1985, Powell 1985, Terborgh et al. types, and in other forest and nonforest1990). Therefore, alteration and fragmenta- vegetative associations, certain habitats thattion of forest tracts may restrict occupancy are important to migratory and resident birdof those tracts by core species, .such that species are highly restricted in area ormigratory species also may not settle there geographically. Such habitats present specialbecause of absence of other flock members challenges for conservation biologists because(Rappole and Morton 1985). Experimental of the fragile existence of such localized plantfragmentation of an Amazonian forest communities (Goodman 1987, Reznicekresulted in the virtually complete disappear- 1987).ance of species that forage in mixed species Fragmentation and isolation of Neo-flocks (Bierregaard 1990). The flexibility of tropical forests is an ominous threat to themigrant winter social systems may be integrity of both resident and migratory birdimportant in determining population re- populations. Fragmentation of forests hassponses to habitat alteration and frag- forced ecologists to incorporate the issues ofmentation (Ewert and Askins 1991, Staicer reserve size and placement into conservation1992). plans. This leads to the concept of conserving

landscapes, the " greater ecosystem " (Grum-

bine 1990) approach to managing biologicalCONSERVATION OF HABITATS IN diversity. Below, three issues pertinent to theTHE NEOTROPICS conservation of habitats for migratory birds

in the Neotropics are discussed. First,Conserving species usually requires pro- migrant use of restricted natural habitats andtection of habitats (Grumbine 1990). Indeed, the unique problems faced by conservationby documenting bird-habitat associations, as biologists attempting to protect these areaswell as understanding the sources of are reviewed. Next, how forest-dwellingvariation, and the behavioral and ecological migratory birds might respond to severecorrelates of these relationships, sound reductions in size of tropical forest tracts areconservation plans can be developed for evaluated. Finally, the first two topics andmigratory birds in the Neotropics. However, others are incorporated into a discussion ondespite the best intentions of ecologists, landscape-Ievel approaches to conservation


of migratory bird habitats within the Thus, mangroves may provide an importantNeotropics. alternative forested habitat in areas where

deforestation has already claimed much of...the surrounding lowland forests.

Conservation of Naturally Restricted Habitats M th t ho hlangroves occupy areas a are 19 y

Most migratory species occupy a range of valued for development projects. Intensebroad leaved forest habitats during winter, pressure for continued destruction ofwhich generally are widespread. As a mangrove swamps exists because of theconsequence, many migratory species can be foreign revenue generated from coastalprotected on their wintering grounds by development (i.e., resort areas). However,preserving various seral stages of major economic incentives also exist to preservetropical vegetative associations, such as mangrove ecosystems; these systems arelowland or montane moist forest or dry critical for local economies because of thesemideciduous forest (see Petit et al. 1993). benefits they provide to commercial fisheriesHowever, there are other, less publicized, (Christensen 1983, Hartshorn et al. 1984).natural vegetation types that are potentially Mangroves also help control shorelineimportant for certain migratory (and erosion and are important in nutrient flowresident) bird species. These habitats are between marine and terrestrial ecosystemsoften locally restricted, such that wildlife (Kennish 1990). These observations highlightspecies that rely upon them are likely to be the need to base conservation plans largelyat greater risk of extinction (Terborgh and upon both the long-term economic andWinter 1980). Several examples are given ecological benefits derived from preservationbelow of migrant use of naturally restricted of mangroves; arguments based solely onvegetation types. These examples are not wildlife conservation may be insufficient toexhaustive, but are illustrative of the curtail conversion of important habitatspotentially tenuous existence of these types (Freese and Saavedra 1991). Protection ofof habitats, as well as the social, economic, mangrove swamps will require concertedand ecological factors that must be legislation and subsequent enforcement byincorporated into plans for their protection. national governments.

Mangrove swamps fringe many salt and Pine forests are widely dispersed, yetbrackish water ecosystems throughout the patchy in their distribution, throughoutworld, mostly in the tropics. In the New Mexico, northern Central America, and theWorld tropics, these forests are disappearing Caribbean. The importance of pine forests tomainly due to development along coast- migratory birds was discussed above and inlines (Christensen 1983, Leonard 1987, a companion paper (Petit et al. 1993). PineWorld Resources Institute 1992). Mangroves associations have received little conservationact as a principal habitat for at least attention, probably because of their lowthree long-distance migrants (Prothonotary species diversity compared to tropicalWarbler, Swainson's Warbler, and Northern broad leaved forests. However, because moreWaterthrush; possibly migratory Yellow than a dozen species of migratory landbirdsWarblers) throughout much of those species' rely on pine-dominated forests as a principalwintering ranges. Mangroves are not a habitat during winter, these habitats cannotprimary habitat for migrants in general, be considered trivial in attempts to sustainalthough in certain regions, such as the diverse migratory bird populations. More-Caribbean, mangroves provide important over, many migratory species using coniferouswintering areas for migratory birds (see forests are habitat specialists, being relativelyabove). For example, several parulines, such restricted to those habitats. For example,as American Redstart, Ovenbird, Black-and- Hut to (1992) calculated the habitat breadthwhite Warbler, Hooded Warbler, and (Levins 1968) of migratory species over-Magnolia Warbler, use mangroves exten- wintering in western Mexico. His data showsively on Caribbean islands, but are much that species that make extensive use ofless commonly found in those swamps in coniferous forests (detected more often inCentral and South America (see above). pine than in any other habitat type; only


species that were found on ~ 3 sites were need to be incorporated into the rotation andincluded) had a lower average habitat management schemes of tropical pine forests.breadth {mean = 1.22 :t 0.38 SD) than Here lies a situation in which internationalspecies that use deciduous forest {1.33 :t cooperation could substantially help ensure0.57), early second growth {1.49 :t 0.62), an ecologically and economically viablethorn forest {1.49 :t 0.52), or cloud forest solution to a relatively simple problem.{1.88 :t 0.67). Exchanges of technical information on

Thus, migrants that inhabit pine forests not management of forests for the dual purposeonly occupy geographically restricted hab- of wildlife habitat and timber production,itats, but they are also limited in their and initial training of Latin Americans atdistribution across habitat types in western foreign institutions, are thought to beMexico. Habitat specialization and habitat important requisites for the long-termrestriction are two ecological factors that preservation of tropical forests {Thelen 1990,have been closely linked to a species' Cornelius 1991).susceptibility to population declines {Ter-borgh and Winter 1980, Rabinowitz et al.

Eff t f F t F .

..ec s ° ores ragmentatlon on

1986). Most continental pme forests are O .. M ... d ...verwlnterlng Igrantsloun m mountamous hIghlands, and areoften exploited and managed for timber Extensive research in North America hasproduction {Hartshorn et al. 1984, Leonard shown that the presence of migratory birds1987). Although total area of pine forest has breeding in temperate forests is intricatelynot declined appreciably in recent years, related to tract size {Whitcomb et al. 1981,mature and old-growth stands are contract- Lynch and Whigham 1984, Blake and Karring at the expense of younger, more 1987, Robbins et al. 1989b). Similarintensively managed plantations {Leonard information on the effects of insularization1987). Implications of this shift in manage- in Latin America, however, is meager.ment practices for overwintering migratory Fragmentation and isolation of tropicalbirds are unknown. Given the patchy nature broadleaved forest may have pronouncedof pine forests and the potential value of the biological ramifications not only for plantspine resources to national economies, special and animals that are directly displaced, butattention needs to be devoted to this also for individuals that remain aftervegetation type to ensure that forests remain isolation. For example, the microclimate ofnot only sustainable, but also ecologically forest fragments can be substantially alteredviable for wildlife, including migratory birds. within several hundred meters of edges, suchFor example, loss of broadleaved understory that a concomitant change in plant speciesin pine forests causes a severe reduction in often results; this, in turn, is partlyquality of these sites for most migratory and responsible for the redistribution or localresident birds {see above). Relatively little extinction of the fauna inhabiting forestpressure has been exerted on Neotropical islands {e.g., Lovejoy et al. 1986, Williams-pine forests in the form of agriculture or cattle Linera 1990, Laurance 1991, Laurance andgrazing because pines frequently grow on Yensen 1991, Saunders et al. 1991). Tropicalsoils that are too poor to support those land- forest fragments generally do not contain asuse practices. At the same time, however, pine many resident bird species as contiguousforests do not provide major natural history tracts during either the breeding orattractions to foreign tourists because of the nonbreeding season {Terborgh 1974, Willisrelatively depauperate fauna and flora found 1979, Robbins et al. 1987, Karr 1990). Mostthere; consequently, one generally cannot cite long-distance migrants, however, do notforeign currency derived from ecotourism as appear to exhibit such pronounced relation-a benefit of preserving pine forests. Thus, the ships with forest area during the borealfate of these forests for wildlife may rest with winter {Robbins et al. 1987). In fact, manythe degree of responsibility exercised by migrants that are area sensitive on thenatural resource agencies and timber breeding grounds show no such relationship,harvesters. Wildlife management principles or are found more often in forest remnants

180 TEMPORAL P..SPECTIVES ON POPULATION UMITATION AND HABITAT USE

than ;n la,ge, !'acts, du,;ng w;nte, (Robb;m Indeed, consecvaL;on b;olog;sts have ,eal;,edet al 1987, 1992, Ask;ns et al (1992) that because hab;tat patches are not closed,eco,ded no ';gn;ficant ,elat;onsh;p between systems and that ;nd;v;duals of a populat;onfmest frngment s;,e and m;grntory b;,d can move fceely among hab;tat patches,abundance on St John and St Thomas landscape-level patterns and processesWand, ;n the Ca,;bbean In add;t;on, Ewert (Tmne, 1989) can g'eatly ;nfluence theand A,k;m (1991) found !hal frngmentat;on compos;t;on of looal faunal a,semblagesof fmest "acts on Caribbean ;,lands d;d not (Lynch and Wh;gham 1984, Saunde" et alalt" the wc;al sy,tem (fiock;ng behav;m) of 1991, The,efme, ma;ntenance of ecooystemm;grntmy wa'ble" The" few ,tud;es ;ntegrity and funct;on ;, fundamenlal to anyruppO" the p,ev;ou, concloo;ooo that 10ng-"'m e!Tmt to p,ese,ve biologicalm;grnnL' as a group may not be g'eaJly dive,s;tya!Tecled by sl;ghJly d;,tmbed and frngmented From a themet;cal standpoint, few canhab;tat, question the value of a landscape-le,el

D"p;te the apparent lack of de"imental approach to comecvation Prnct;cal appl;ca-e!T,cts of forest frngmentat;on on moot (;on of theory i, difficult to ;mplement,m;grntmy 'p'c;es ;n the Neo"opi", few data how"", because of the a",umulat;on ofha" be,n collected to e,am;ne fcagmenta- problem, aswc;ated w;th ;ncrease, ;n landt;on ;sru,s, and result, mu,' be conside,ed a,ea ,equ;,ed fm comecvation plans"ntat;', Indeed, se"rnl m;grntory 'p'c;es Imag;ne, fm "ample, the ;nc,eased numbe,ha" e,h;bited d;spropon;onate u" of IMge of political, wc;al and cultillal boundari"fmested ,tands (Robbim et al 1987) and that a land-oo, planne, would ha', to c'oo,some of the m;gcatory specie, idenL;fied as ;n establ;sh;ng a 100,000 ha b;oophe,e 'ese',eh;ghly ru,cept;ble to fore,! al"rnt;on (Table compMed to, say, a 200 ha ,esecve No easy6-3) may be equally ,elect;" ;n me of IMge wlut;om ',;,t fm prop" management of"acts Ask;m et al (1992) at,,;buted ent;re wa""heds m land,cap', A 'ys"m ofrelat;'cly low m;grnnt abundance ;n fmests IMge ,esecves i, often o!Te,ed as the best hop'on St Thomas to illban;,at;on and ofp'e,ecv;ng the naturnl d;",sity of "op;calfcagm,ntat;on Fillth"more, Rappole " al communities (Rubino!T 1983, Corncl;us(1989) ,howed that fcagm,"tat;on of forest 1991, In most s;tuations, human populat;oncan fmce ind;,iduals to wand" ;n 'eMch of presrure continue, to place "calatingsu;table hab;tat, beha,im that ;n turn wa, demands on "op;cal ,ewillce. such !halcmrelated w;th inc,eased mortal;ty Most landseap's managed fm both nati,e 'p'c;esresea,ch on fcagm,"tM;on, how"", has not and su,ta;ned econom;c de,elopment ;s thebeen based upon a ,;g;d "p',;mental de,;gn moot prnct;cal approach to ma;nta;ningThus, futu'e 'eseMch must add,es, both natillal d;",s;ty on a long-t"m basis (Say"d;,ect and ;ndi,ect consequence, of fcag- 1991, Woodwell1991, A p;,otal ;"ue ;, Ihementat;on (see T"bmgh 1992) on m;gmtory de,elopment of a protocol that ;ncmporntesb;,d populat;ons through Well-d,s;gn,d boLh w;ldl;fe and wcioeconom;c cons;d,,-,tud;es at;ons ;nto long-"'m ma;n"nance of

land,capec Many consecvat;onists bel;",Ihat ruch plans Me technically, economically,

L,"d",p~L.,.1 Appro"h to CO"se",UO" d It II ' bl th Cbbe df M " an po I .ca y ,,as. e m e an an ano 'grn" Lat;n America, albe;t w;th prod;gious ,!Tort

Wildl;re-hab;tat relat;omhip, hi,tm;cally (eg" F,arns;d, 1989, Vaughan 1990,ha" been exam;ned at the m;crohabitat m Se,e,al key componenl' of a landscap'-stand le,cl (eg" MacAnhill and MacAnhill le'el approach to the consecvation of196i, lame, 1971, Holm" et al 1979) migrntory bi,ds ;n the Neo"op;cs ha"Recently, how"", emphasi, ;n North already been "am;n,d ;n th;s chapterAmerica and EillOp' has been placed on fcagmentation e!Tecl' (patch ,i'., cmridills,;nco,pornt;ng h;gh,,-le,cl 'patial scale, ;nto and landscape compos;uon and s"ucturethe ,tudy of w;ldl;fe populat;ons (Fmman (Turn" 1989) While detallede'am;nat;on ofand Godron 1981, O'Neill et al 1986) the thoo,y and appl;cat;on of landscape


ecology is beyond the scope of this chapter, Within the winter geographic range ofbelow are several general considerations for migratory species, critical habitats must belandscape-level conservation of migratory present in proportions sufficient to maintainbird habitats in the Neotropics. overwinter survival. Above, an example of

Landscape-level designs for conservation "landscape complementation " was provided

of wildlife during nonbreeding periods must that highlighted the relevance of maintainingprovide the quantity, diversity, and quality an accurate representation of naturalof habitat to support local populations of landscapes. Native plant species in relativelynaturally occurring species and to maximize natural landscapes are the "building blocks"their survival rates. The goal of maintaining of any long-term conservation strategy.high integrity (sensu Karr and Dudley 1981) Research is needed critically not only onof wildlife communities in tropical landscapes the types of landscape elements (sensurequires that at least three ecological Forman and Godron 1986) or habitats thatconcepts be integrated into long-term are used by migratory birds, but also on theconservation blueprints: (1) native plant minimum sizes of those elements. Minimumspecies and landscape composition; (2) area requirements are know only for aminimum area requirements; and (3) handful of forest-dwelling migratory specieslandscape connectivity. (see above). However, available information

The highly coevolved interactions between on these requisites is limited and, as moretropical plants and animals is well docu- data are gathered (especially in Southmented (Gilbert 1980, Terborgh 1986). Those America), additional species undoubtedlytightly coupled systems make existence will be identified as requiring largeprecarious for species intricately dependent forested tracts. Nevertheless, sizable areas ofupon others, such that removal of a keystone unbroken forest are imperative for manyspecies (Paine 1969, Terborgh 1992) can lead resident species and for maintaining theto cascading losses of dozens of other species. integrity of tropical forest ecosystemsIn the context of this chapter, loss of keystone (Terborgh 1992); that integrity will ultimatelyplants (Mills et al. 1993) could cause determine the long-term persistence of manydegeneration of entire ecosystems upon migratory and resident species.which migratory birds are reliant (Terborgh Migratory birds must be able to travel1986). More generally, alteration of native among habitat patches during overwinteringplant species composition through an- and migration periods. Conservation bio-thropomorphic disturbance, such as frag- logists have recognized that, by connectingmentation (Laurance and Yensen 1991) or patches of similar habitat via corridors (Nossfire (Kellman and Tackaberry 1993), could and Harris 1986), animal movement ishave serious consequences for migratory facilitated, which may result in easier accessbirds and other animals dependent upon to critical food resources (Loiselle and Blakethem for overwinter survival (Terborgh 1991) and, ultimately, in reduced mortality1989). Furthermore, introduction of exotic rates (Rappole et al. 1989). Maximization oftree species of economic value can render landscape connectivity is believed to optimize"forested " areas virtually useless to migra- the ecological benefits accrued through

tory birds. For example, pine plantations in protection of critical habitat patches (NossPanama do not offer many migratory or 1987, Hansson 1991), although certain risksresident bird species suitable habitat (L. J. also are inherent with this scheme (SimberloffPetit, unpublished data). By maintaining and Cox 1987). An essential first step foraccurate representation of native plant development of landscape and regionalspecies, conservationists increase the prob- conservation plans is to quantify theability of maximizing diversity of native distribution and extent of different land useanimal species, including migratory birds and cover types within national boundaries.(Scott et al. 1987). Although lack of information on the extent

The composition of landscapes refers to of forest conversion and other land uses hasthe representation of different native and been a major problem plaguing Latinartificial cover types (Dunning et al. 1992). American governments, recent progress in


this area (mainly through application of that overwinter in North America, but dataremote sensing) has been encouraging (e.g., from south of the United States border areSader and Joyce 1988, Dirzo and Garcia limited. In addition, locations of individuals1992, Powell et al. 1992). Moreover recently captured in the Neotropics and banded withdeveloped conservation plans that emphasize US Fish and Wildlife Service (USFWS)landscape connectivity for migrating birds aluminum bands are stored on computer atare being implemented in the Neotropics the Bird Banding Laboratory in Laurel.(e.g., Powell and Bjork 1994). In general, the distributional data available

One application of a landscape or regional for migrants are imprecise because theapproach to migratory bird conservation fine-scale details from local accounts are lostcould be identification and conservation of when information is summarized at relativelythose areas used as major stopover sites by broad geographic scales (e.g., states ormigrating birds (Rappole 1991). Although countries). Furthermore, individuals are notbirds may not remain at stopover sites for distributed equally throughout a species'more than several days, these areas are range, and centers of abundance are notcritical for completion of the migratory trek revealed in range maps. Improvement of(Moore et al., chapter 5, this volume). winter range maps for migratory birds isProtection of large areas of native habitat dependent upon efforts aimed at: (1)along coastlines and on Caribbean islands collection of more information on the localseems especially significant for trans-Gulf distribution of species, and (2) a bettermigrants (Moore et al. 1993). method of information storage and retrieval.

Additional data on the abundances of speciesin different areas could be gathered either

FUTURE RESEARCH ON HABITAT USE through a systematic survey of habitats andAND CONSERVATION regions (see below), or through compilation

of information from field notes of naturalists.Elsewhere, guidelines for the conservation of In this latter case, if bird lists were availablemigratory birds in the Neotropics have been from as few as ten people who visited a givenpresented (Petit et al. 1993). However, much area, a reasonable index of abundance couldadditional research needs to be completed on be estimated (because frequency of occur-the Neotropical wintering grounds before rence is related to abundance; e.g., Hutto etcomprehensive conservation plans that al. 1986, Wunderle and Waide 1993).cover all species and regions can be devised. Comparisons using these data would have toSeveral important directions for future be restricted by habitat, but it would beresearch on the distribution and habitat use reasonable to assume that a bird speciesby migratory birds are summarized below. detected by 80 of 100 people in one place is

more common there than in another locationG h " D " t "

b t " f M " t where the species was detected by only 30 ofeograp IC IS rl u Ion o Igran s .100 people. Many professional and nonpro-

Knowledge of geographic distribution of fessional field biologists have notebooks fullspecies is a basic element in development of of valuable information that may never beconservation plans (Terborgh and Winter used. Perhaps the National Biological1983, Margules et al. 1988). Historically, the Service or a nongovernmental organizationwintering ranges of migratory birds, as could serve as a repository for suchdepicted in field guides or by text in the information.American Ornithologists' Union (AOU) These data, along with those from museumcheck-list (AOU 1983), have been determined specimens, banding records, and systematicthrough a synthesis of information available surveys (see below), could be entered into afrom published regional accounts and from geographic information system (GIS), suchmuseum specimen tags (C. S. Robbins, that distributional maps for various types ofpersonal communication). The National data (sex or age distribution) or for differentBiological Service (Laurel, Maryland) has spatial scales (range-wide or regional) couldmuch of this information mapped for species be produced. Within a few years, generation


of winter distribution maps for tropical combinations of habitat groupings, whichlatitudes that are as accurate as those enhances the types of questions that can beproduced for temperate North America from addressed.Christmas Bird Counts (Bock and Root1981) may be possible. The importance of H b.t t A . t . f M . t S .

...a I a ssocla Ions o Igra ory peclesaccurate Information on the wInter geo-

graphic distribution of migrants cannot be A broad overview of winter habitat use byoverstated. Without such data, identification migratory birds has been presented in thisof areas most appropriate for conservation chapter. However, because of relatively scantof migratory birds will remain speculative. data and the great geographic variability in

habitat use exhibited by many species, thisD f .. H b .t t T U d b M . t B . d preliminary assessment must be interpreted

e Imng a I a ypes se y Igra ory Ir s cautiously. Indeed, the validity of applyIng

A major obstacle to summarizing informa- the general information presented here toti on on habitat use by migrants is the specific local field situations is known. Ainconsistent fashion in which avian ecologists comprehensive, quantitative survey ofidentify habitat types; there is a genuine need habitat use by resident and migratory birdsfor standardization. Also, many authors in each of the major physiographic regionsprovide vague descriptions of study sites. (or life zones) of each country is sorelyResearchers should include the following needed. These data are already available forinformation in site descriptions: (1) elevation; certain regions of some countries (e.g.,(2) latitude and longitude; (3) type of Yucatan Peninsula and high-elevation lifevegetative association, as defined by a zones of western Mexico). The product ofrecognized regional botanical guide; (4) these efforts would be an abundance-canopy height; (5) major plant taxa; and (6) weighted habitat and geographic analysis ofextent and types of disturbance. This the winter distributions of migratory (andinformation could be used to classify habitat resident) birds, a central component neces-types by means of a systematic, multivariable sary to build a conservation strategy forprocedure. For example, in western Mexico, migratory birds in the Neotropics (TerborghR. L. Hut to classifies habitats by two main 1974, Greenberg 1986).criteria: (1) broad habitat cover type [as To assess habitat use and changes indefined by Rzedowski (1983)], such as long-term abundances of migratory birds,marsh/wetland, streamside riparian, man- surveys based upon point counts aregrove, tropical evergreen forest, tropical recommended (Hut to et al. 1986). The exactdeciduous forest, cloud forest, oak woodland, method needs to be determined (see Ralphpine-oak woodland, and fir forest; and (2) et al. 1994), but would require distinguishingthe type and extent of disturbance, such as individuals that are both within and beyondundisturbed, selectively cut, understory some threshold distance (e.g., 25 m or 50 m)removal, severely cut and understory from the observer (see Hut to et al. 1986).replaced by crops (e.g., coffee), clear-cut Mist-netting could be used to supplementpresently in early stages of regrowth ( < 1 m), visual/auditory surveys and may provideclear-cut presently replaced by tall-stature important information about habitat-basedagriculture (e.g., banana, citrus), and survival rates. However, comprehensiveclear-cut largely replaced by man-made visual/auditory surveys should also bestructures (e.g., towns). Thus, if there are ten conducted in all regions because of the biaseshabitat types and eight kinds of disturbance, associated with mist-netting.there would be 80 different categories Density estimates of each species withinpossible. Some of these combinations habitats and regions would provide the mostprobably would not occur and some useful data for assessing habitat relation-categories could be lumped together to ships. Indices of abundance, as are typicallyreduce the number of categories. An generated through point-count techniques,advantage of this system is that it allows may be misleading in comparisons amonganalyses to be performed for many different habitats because of inconsistent probabilities


of detection of species in different vegetation ultimate correlates of habitat selection maytypes (Petit et al. 1995). Estimates of density at first seem to be a purely " academic "

based upon point counts or transects, exercise, answers to these inquiries canhowever, are subject to a host of similar provide important insight into conservationcriticisms (Ralph and Scott 1981) and, issues. For example, if habitat use bytherefore, may not provide information insectivorous migratory species is closelyany more accurate than indices of abun- linked to insect density, then development ofdance. Nevertheless, in theory, estimates thorough conservation plans for migrantsof density offer the best opportunity for requires an understanding of the effects ofsurvey data to identify habitats and areas of disturbance on insect populations. Futuregreatest importance to migratory birds. research in the Neotropics needs to beThose survey and analytical techniques need directed at determining the biotic factorsto be pursued. influencing habitat use by migrants, and how

Long-term monitoring of populations also anthropogenic disruption of these factorswould provide information on the temporal might affect overwintering populations ofsuitability of habitat associations during migratory birds.winter and the relative importance of certainhabitats as refugia for migrants during. ." bottle eck" ( W " 1977) a F Effects of Land-Use Practices on HabItat Use

n sensu lens ye rs. or d O . S . I1 d ..an verwlnter urvlvaexamp e, unng drought years mIgrants maybe unable to find sufficient food in some Economists, ecologists, and conservationistshabitats, such as dry forests, and be forced need to collaborate to evaluate the potentialto relocate to moister habitats (Faaborg importance and compatibility of certainet al. 1984). In this case, the benefit of access sustainable land uses and wildlife conserva-to moist forest in sustaining long-term ti on (Freese and Saavedra 1991). However,populations is clear, although in most years few detailed research projects have focusedmoist forests may contain relatively few on the relationship between sustainableindividuals. forestry or agroforestry and bird populations,

and few of those projects have actually beenBehavioral, Ecological, and Evolutionary implemente.d a.cross mul.tiple sites (CorneliusCorrelates of Habitat Use 1991). In pnnclple, sustaInable land uses offer

the most promising alternatives for long-termSeveral hypotheses have been used to explain stability of tropical economies and wildlifethe general pattern of habitat use by populations (Petit et al. 1993). An importantoverwintering migratory birds in the area of future tropical research for avianNeotropics: (1) migrants track sources of ecologists is to determine the relative benefitssuperabundant food (Willis 1966, Karr 1976); and costs of different modes of sustainable(2) migratory birds spend only 6 months on forestry (e.g., Brokaw 1995), and of differenttheir wintering grounds and, therefore, are land-use practices in general, to migratoryforced out of certain habitats by dominant and resident bird species. Furthermore, asresident species (Slud 1960, Leck 1972a); and agricultural lands expand throughout Latin(3) migrants may select wintering habitats America, traditional wildlife managementthat are similar to those used during the practices (e.g., maintenance of hedgerows andbreeding season (Morse 1971, Hut to 1980). woodlots; Gradwohl and Greenberg 1991)Conflicting evidence exists for all three of need to be incorporated into landscape plan-these nonmutually exclusive hypotheses ning. A valuable type of information is the sur-because of the complex environmental factors vival rates of species occupying natural andinfluencing migrant distributions. The food anthropogenically disturbed habitats (bothhypothesis seems to be favored by many survival rates and reproductive success areavian ecologists (e.g., Levey and Stiles 1992), necessary for resident tropical species). With-but definitive evidence has yet to be collected out those data, development of conservation(Petit 1991). While resolution of these and plans that incorporate sustainable land-useother questions relating to the proximate and practices will remain tenuous.


SUMMARY AND CONCLUSIONS with higher altitudinal locations. Thiselevational relationship disappeared, how-

The downward trends in populations of ever, in disturbed forests. Neotropicalmany species of Neotropical migratory birds migrants also were underrepresented onhave been documented through long-term Caribbean islands and in habitats in lowermonitoring of populations in the United Mesoamerica and South America, which mayStates. Those alarming patterns, in light of be associated with the energetic andthe widespread alteration and loss of habitats evolutionary costs of travelling long dis-in the Western Hemisphere, suggest that tances or over inhospitable areas (water).humans have played a fundamental role in Patterns of habitat use varied geogra-the declines of these intercontinental phically for several species, and for themigrants. The first step in developing effective migratory bird community as a whole.conservation plans for these species is to Furthermore, different geographic regionsidentify habitats, areas, and species in need supported distinct assemblages of migratoryof greatest protection. This review provided birds. Thus, a regional approach based upon,a general synthesis of the mass of ecological for example, physiographic boundaries, mayinformation on Neotropical migratory birds provide the most effective means ofavailable to conservation planners, as well as identifying, implementing, and coordinatingpotential approaches that may be effective in conservation practices necessary to maintainconserving this natural resource in tropical winter populations of migratory birds. Inwintering areas. addition, ecologists must incorporate into

Migratory birds were found to occupy conservation directives the variation inmany different vegetation types during the habitat use associated with age, sex, and timeboreal winter and only highly managed of season.pasture and large monocultures of mono- Knowledge of the natural history andcotyledonous crops, such as sugar cane, behavioral ecology of species may assist inbanana, and rice, seemed to be overtly predicting the potential impacts oflocalland-incompatible with the goal of protecting use practices on migratory birds. Formigratory birds. Native broadleaved forests, example, forest fragmentation may havein all stages of succession, supported more greater repercussions for species that searchmigrants than highly managed and/or for insects in ground leaf litter or forageartificial habitats, such as pine plantations, within multispecies flocks because of theresidential areas, and pastures. Generally, relative influence of tract size on leaf litterslightly disturbed native forests, such as forest invertebrates and presence of core-flockingfragments and selectively logged tracts, species, respectively. Furthermore, thatharbored a greater number of migratory information can assist land-use planners inspecies than pristine forests. However, at least developing optimal strategies for landscape-one dozen migratory species, and many level conservation of tropical wildlife.resident tropical species, are highly de- Large-scale approaches to conservation alsopendent upon large tracts of undisturbed must incorporate state-of-the-art knowledgeforest. Because of this habitat specialization, of habitat and minimum area requirementsthese are the species most in danger of severe of target species, as well as principles ofpopulation declines. Thus, the current focus landscape ecology.among environmentalists on preservation of Further research on the ecology oftropical forests seems warranted for migra- Neotropical migrants needs to be conductedtory birds as well. before reasonable confidence can be placed

Contrary to conventional wisdom, migra- in conservation plans. That research musttory birds did not concentrate at middle focus on quantifying the geographic dis-elevations during winter. Rather, migrants tributions, habitat requirements, and be-represented a greater proportion of the entire havioral ecology of migrants, as well as theiravifauna at lower elevations, this being a survival rates in native habitats and underresult of a greater number of migratory various agricultural, agroforestry, andspecies occupying lowland sites compared wildlife management regimes. Those goals


further mandate that ecologists devise more populations in southern Wisconsin forests:effective ways to quantify, store, and 1954 and 1979.J.Field Ornithol.53:149-158.distribute information on migrant abund- American Ornitholo~ists' 1!nion. 1983. Check.-Iistance in different habitats and geographic of ~orth ~me.nca~ birds, 6th ed. AmerIcanregions Ornithologists Union, Lawrence, KS.

Th ~ t f N t . I . t b . d Arendt, W. J. 1992. Status of North Americane la e o eo roplca mlgra ory Ir s migrant landblrds In the CarIbbean region: a

?epends. upon the. ~ctlve partIclP.atIon of an summary. Pp. 143-174 in Ecology andm~ernatlo~al C?alltlon. ~f .ecologl~ts, econo- conservation of Neotropical migrant land-mIsts, socIologIsts, pOlitIcIans, pnvate landbirds (J. M. Hagan, III and D. W. Johnston,owners, national governments, nongovern- eds). Smithsonian Institution Press, Washing-mental organizations, academic institu- ton, DC.tions, business leaders and private citizens. Askins, R. A., J. F. Lynch, and R. Greenberg. 1990.Equally important to understanding the Population declines i~ migratory bir~s inscience behind sustaining native popula- eastern North AmerIca. Curr. Ornlthol.. f .Idl ." . h .d .f 7.1-57 tlons o WI lIe, IS t e urgency to I entl y A k .. R .A d D N E t 1991 I t.

b ll d . h II Slns,..,an ..wer. .mpacsust~ma e an. -u~e practIces t. at a ow of Hurricane Hugo on bird populations

coexIstence of wIldlIfe and expandmg human on St John's US Virgin Islands. Biotropicapopulations in Latin America. Neotropical 23:481-487.migratory birds appear to be more re- Askins, R. A., D. N. Ewert, and R. L. Norton. 1992.silient to small-scale alteration of native Abundance of wintering migrants in frag-vegetation than are tropical resident species. mented and continuous forests in the USThus, efforts aimed at improving the Virgin Islands: Pp. 197-206 .in Ec.ologywelfare of rural human populations, while and ~onservatlon of Neotroplcal migrantprotecting habitats of resident wildlife landblrds (J. M. .Haga?, III .an~ D. W.

. fli th t fti t .Johnston, eds). Smlthsonlan InstItution Press,

species, may o er e mos e ec Ive means W h .t DCf .. I . b .l . f II as Ing on, .

o .sustalm~g ?ng-te.rrn vIa Ilty. o a. Baltz, M. E. 1993. Abundance of NeotropicalnatIve sp.ecles, mcludmg Neotroplcal ml- migrant songbirds on North Andros Is-gratory bIrds. land, Bahamas. Florida Field Natur. 21 :

115-117.Bennett, S. E. 1980. Interspecific competition and

ACKNOWLEDGMENTS the niche .o~ the .Ame~can. Redstart (Seto-phaga rutlclila) In wInterIng and breed-

Bob Askins Christine Champe Russ i~g c~mmunities. Pp.. 319-335 in Migr~nt, ...' bIrds In the Neotroplcs: ecology, behavior,

Gr~enb~rg, To~ MartIn, Char~le Pame, Ken distribution, and conservation (A. Keast andPetIt, LIsa P.etlt, Chan R.ob?ms, and Tom E. S. Morton, eds). Smithsonian Institution

Sherry provided many mslghtful sugges- Press, Washington, DC.tions for the improvement of our work. Bierregaard, R. 0., Jr. 1990. Avian communitiesBob Askins, Christine Champe, Courtney in the understory of Amazonian forestConway, Lisa Petit, Tom Sherry, Jay fragments. Pp. 333-343 in Biogeography andVaninni, and Joe Wunderle generously ecology of forest b~rd com:m~nities (A. Keast,offered access to unpublished manuscripts ed.). SPB Academic PublishIng, The Hague,and data The Netherlands.

.Binford, L. C. 1989. A distributional survey of thebirds of Oaxaca. Ornithol. Monogr. no.43:1-418.

Blake, J. G., and W. G. Hoppes. 1986. InfluenceLITERATURE CITED of resource abundance on use of tree-fall gaps

by birds in an isolated woodlot. AukAlcorn, J. B. 1990. Indigenous agroforestry 103:328-340.

strategies meeting farmers' needs. Pp. 141- Blake, J. G., and J. R. Karr. 1987. Breeding birds151 in Alternatives to deforestation: steps of isolated woodlots: area and habitattoward sustainable use of the Amazon rain relationships. Ecology 68: 1724-1734.forest (A. B. Anderson, ed.). Columbia Blake, J. G., and B. A. Loiselle. 1992. Habitat useUniversity Press, New York. by Neotropical migrants at La Selva

Ambuel, B., and S. A. Temple. 1982. Songbird Biological Station and Braulio Carrillo

w,m" H""', US, 'ND CONS"',nON

N"""..P"kCo",'i,,pp25c2nm Ch,i",".,B'983M,",ro", wh"""h"E,",°","drorn,~"j,"ofNw,ropi"f woc'h'U".",,34"'Smj8""',""dbi,d.f!MH""m,odO Cod"ML(""198SH,bj",.,I~'jo"j"bj,d'W fuhrn'o, ,d.' Sm"hwoj," rn"i",jo" A"d,,", p".. S'O Oi"o CA

BI'k:'7'G~'~i":,";;j'~.C,"d B ALoj"n, CO"~:;;C6;,~;.~:,~:;;~~;~0~;i1;0~~:i~~;'990B"d.of",Scl,.h,b",,0., mi""oc,bj,d.j"""N-"jo,,,."d',"ph" ,ompo';'io, ,"d mi","" Pp m,,"" 'ropj", fore," Corn", Biof (j"16f"82j"Fo",Nw,ropjc"fo"",(' p,~.,G,"", ,d, y", U"i""j,, p".. N,w COC"clio, S E '99' Wj'dlj" W""",ti'" j"H"" CT C,""" Am'ri,. wjn i' ,""j" 'h, 90'?

Boc'C E ,"d T L '0", '98' Th,Ch,i"m,. Trn",No"hAm" WifdlN" 'rnoocC,"("j" Co""""d,,i," "0,"" Pp'723 m 56"'-49

~:\;:'~:rJ"~'~~;,\";:::;'"~,~:d" l;i,; c'"",~; '~::; ';lj::,"\::~",~ ";:;il~~ ' J:"~:Biof "0 .Coopo, Oc"j,hofo.ru' SOO,," 52"74279"'w,,"~ KS Di,mo"d A W '98S Th, ""ctio. of ,ri'jrul

Bo.q"CC","dMLo"'jw'98cTh",..", ',rn.,"d,o"W'W".'~"iO""OC"i"ofNo"h'm"i',"m"",0",,"dbi,d. ""pi,"(oc~,bj,d,P~33-48j"C"".",,'hro",h","ph"jch,bi"'.of'h,w~"ro ","ofUopi",fo""b"d,(AWDi,mo"doo"'of'w',"cl.Bjo"opi,,'9".c2n '"dTELo"jo""."CBPT"hP"b'

B"gg,SA","dlHC,i,wcl'1978G"d,,1 "c4."0"",of,p""8,"WMh,",,o.A"'",j, Dj,moodAW'99'A,...mw,of'h'ri,""omN",M32"926 "opi~'d,fo"",'ioo,oC."di..,o"gbi,d,

Bri"i",h'mMC,odSAT,mp"'983H", T"mNMhAm"Wj'd'N.'R,w",Comoowbj,d,~""'fore.wo8bj,d,,"d~li",? 56"7H94BjoS,iw,,33"C35 Din.R".odMAG."i.'992R,",of

B,"k,w N '995 M'ho,.", ,i"icwru" ,"d d"oc~",jO" '" Lm TM'", .Nw'ropj",

mig"",bj"ooo.""i,"j"Bcl,,"p,,m,,, ""i"w",h~"M"jooC,"."Bio'i" Fligh' N,w""",4"'26 68490

Brow.JL'969'"ri",i"b,h",joc,"d D,""j".JB"BJ?",clw.,"dHRPwli"mPOP"""'""g,"'jo"j.bj,d'WjbooB"" '992&01o"c"pro=",h".ff~'g,"93329 POpw,tiom j" oomp"' ""d=po, Oikm

B",""FWI9'8Mo"j",i",biol,8i," 65"69'75dj"",,yfOC""i"gpnocj,jMi"00""~"i"" E","SW'953Woodw"b""wj"",i"8i"Pp22c230mB,od","",(EOW",o. C""W"woBw'6"'6904;;'i N"i,"" Ac,d,m, p,~, W",hi"g,". Ew,,;;~;i~; ;r"~j:;,~0;'w':;:":;'j; ::;;,~i7:

B",kj"WH'976Soci""",m,jo",opi," 'h,Vi,gi°"'""d,Coodoc93864g68fo"" .""",, Am., N,,", UC293 "'0 F",ho,. J" ""d W J A""d' '992 L008""m

B"'k;;;\;,,"':;;,,~;R~B;;kj~,:dO;'~p:w,~ ~:'~;;:'"i%:;;;';0';:;";;:,hw,~f::,~~i:'9n,",,"p"'fi,bjm,oc"m,mpic" """b"?pp5763m&ol08'."doo"hj,h""d P"'m. A"k 896"624 ,."","" of Nw'mpj", mj,,"", l""dbi,d,

Ch,~::: ;~w',::;i B;;ti~::;~:";:'~,:~::~:~ ~;'[;'h,~;j::.,;:~,:~10~p';.;:0~::::;:,~~Flo!id" MS 'h"j, U"j"",j" F,",jd, OCG.""ru"F,"73pp F,"boc,J"WlA""d',"dMSK.~,'9g4

Chip""'M'980NO"b,~djo8~olo8'of'h, R,i""noo,""",ofbj,dPOpw"j,"'"",B",kb,rn,""W"b",PP309317jo 'iOMiO"P"",oRj""d"fo",'."i",,.",Mjg,"m bj,d, m 'h, Nw'mpj" ,oofo" ""d, Wj"," Bwl96575593boh',jocdi"'ibWjo.""dooo"""io"(A F",o,jdcPM'989E"""j",=~,,j"K,M' 'od E S Mo"o. ,d" Sm,'h,o"j," B"n,j,"AmMOoi.Bw5,iw~39"87393'm'i,"'i," p"" WMhj"g"" DC F,=" R T T" ""d M Godro. '98' P",h,.

C",.&",d. R" A E""d, O P",h"y ,"d "od "'""",,, oompo"w" fo, , ""d,~poWB,"ow'985Lj"of'h'b"d'of'h'Lm '00,""Biosqw,,3'733740T""". B",""", S"'io. 'm'j",," d, Focm,. R T T" ,"d M God,". '98.Bjol°8j. U"i,,"id,d NMw,,1 A"m"o= ","d,~po ,00,"gy Joh" Wi", ""d S,",d,M"iooM"iooDF N,wyock


Franzreb, K. E. 1983. A comparison of avian behavior, distribution, and conservationforaging behavior in unlogged and logged (A. Keast and E. S. Morton, eds). Smith-mixed-coniferous forest. Wilson Bull. sonian Institution Press, Washington,95:60-76. DC.

Freese, C. H., and C. J. Saavedra. 1991. Prospects Greenberg, R. 1981. Frugivory in some migrantfor wildlife management in Latin America and tropical forest wood warblers. Biotropicathe Caribbean. Pp. 430-444 in Neotropical 13:215-223.wildlife use and conservation (1. G. Robinson Greenberg, R. 1986. Competition in migrant birdsand K. H. Redford, eds). University of in the nonbreeding season. Curr. Ornithol.Chicago Press, Chicago, IL. 3:281-307.

Fretwell, S. D., and H. L. Lucas. 1970. On Greenberg, R.1992a. Forest migrants in non-forestterritorial behavior and other factors affecting habitats on the Yucatan Peninsula. Pp. 13-22habitat distribution in birds. I. Theoretical in Ecology and conservation of Neotropicaldevelopment. Acta Biotheoretica 19: 16-36. migrant landbirds (1. M. Hagan, III and D.

Gates, J. E., and L. W. Gysel. 1978. Avian nest W. Johnston, eds). Smithsonian Institutiondispersion and fledgling success in field-forest Press, Washington, DC.ecotones. Ecology 59:871-883. Greenberg, R. 1992b. The nonbreeding season:

Gauthreaux, S. A., Jr. 1992. The use of weather introduction. Pp. 115-177 in Migrant birdsradar to monitor long-term patterns of in the Neotropics: ecology, behavior, dis-trans-Gulfmigration in spring. Pp. 96-100 in tribution, and conservation (A. Keast and E.Ecology and conservation of Neotropical S. Morton, eds). Smithsonian Institutionmigrant landbirds (1. M. Hagan, III and D. Press, Washington, DC.W. Johnston, eds). Smithsonian Institution Griffin, D. R. 1964. Bird migration. NaturalPress, Washington, DC. History Press, Garden City, New York.

Geerling, C., H. Breman, and E. T. Berczy. 1986. Grumbine, E. 1990. Protecting biological diversityEcology and development: an attempt to through the greater ecosystem concept. Nat.synthesize. Environ. Conserv. 13:211-214. Areas J. 10:114-120.

Gibbs, J. P., and J. Faaborg. 1990. Estimating the Hamel P. B. 1986. Bachman's Warbler: a speciesviability of Ovenbird and Kentucky Warbler in peril. Smithsonian Institution Press,populations in forest fragments. Conserv. Washington, DC.Bioi. 4: 193-196. Hansson, L. 1991. Dispersal and connectivity in

Gilbert, L. E. 1980. Food web organization and metapopulations. Bioi. J. Linnean Soc.the conservation of Neotropical diversity. 42:89-103.Pp. 11-33 in Conservation biology: an Harris, L. D., and P. Kangas. 1988. Recon-evolutionary-ecological perspective (M. E. sideration of the habitat concept. Trans.Soule and B. A. Wilcox, eds). Sinauer North Amer. Wildl. Nat. Resources Conf.Associates, Sunderland, MA. 53:137-144.

Glasser, J. W. 1982. A theory of trophic strategies: Hartshorn, G. S. 1983. Plants: introduction. Pp.the evolution of facultative specialists. Amer. 118-157 in Costa Rican natural history (D.Natur. 119:250-262. H. Janzen, ed.). University of Chicago

Glasser, J. W. 1984. Evolution of efficiencies and Press, Chicago, IL.strategies of resource exploitation. Ecology Hartshorn, G. S. 1992. Forest loss and future65:1570-1578. options in Central America. Pp. 13-22 in

Goodman, D. 1987. The demography of chance Ecology and conservation of Neotropicalextinction. Pp. 11-34 in Viable populations migrant landbirds (1. M. Hagan, III and D.for conservation (M. E. Soule, ed.). W. Johnston, eds). Smithsonian InstitutionCambridge University Press, Cambridge, Press, Washington, DC.England. Hartshorn, G. S., L. Nicolait, L. Hartshorn, G.

Gordon, D. M. 1991. Variation and change in Bevier, R. Brightman, J. Cal, A. Cawich, W.behavioral ecology. Ecology 72: 1196-1203. Davidson, R. Dubois, C. Dyer, G. Gibson, W.

Gow, D. D. 1992. Forestry for sustainable Hawley, J. Leonard, R. Nicolait, D. Weyer,development: the social dimension. Unasylva H. White, and C. Wright. 1984. Belize country43:41-45. environmental profile. Trejos. Hnos Sucs.

Gradwohl, J., and R. Greenberg. 1991. Small forest SA, San Jose, Costa Rica.reserves: making the best of a bad situation. Holdridge, L. R. 1947. Determination of worldClim. Change 18:253-256. plant formations from simple climatic data.

Greenberg, R. 1980. Demographic aspects of Science 105:367-368.long-distance migration. Pp. 493-504 in Holdridge, L. R. 1970. Natural vegetation andMigrant birds in the Neotropics: ecology, reservation prospects in northern Latin


America. Pp. 27-33 in The avifauna of 455-476 in Habitat selection in birds (M. L.northern Latin America (H. K. Buechner and Cody, ed.). Academic Press, New York.J. H. Buechner, eds). Smithsonian Contr. Hut to, R. L. 1987. A description of mixed-speciesZool. no.26. Smithsonian Institution Press, insectivorous bird flocks in western Mexico.Washington, DC. Condor 89:282-292.

Holdridge, L. R., W. Grenke, W. Hatheway, T. Hut to, R. L. 1988a. Is tropical deforestationLiang, and J. Tosi. 1971. Forest environments responsible for the reported declines inin tropical life zones: a pilot study. Pergamon, Neotropical migrant populations? Amer.Elmsford, NY. Birds 42:375-379.

Holmes, R. T., R. E. Bonney, Jr., and S. W. Pacala. Hut to, R. L. 1988b. Foraging behavior patterns1979. Guild structure of the Hubbard Brook suggest a possible cost associated withbird community: a multivariate approach. participation in mixed-species bird flocks.Ecology 60:512-520. Oikos 51:79-83.

Holmes, R. T., and T. W. Sherry. 1992. Site fidelity Hut to, R. L. 1989. The effect of habitat alterationof migratory warblers in temperate breeding on migratory land birds in a west Mexicanand Neotropical wintering areas: implications tropical deciduous forest: a conservationfor population dynamics, habitat selection, perspective. Conserv. BioI. 3:138-148.and conservation. Pp. 563-575 in Ecology Hut to, R. L. 1990. Measuring the availability ofand conservation of Neotropical migrant food resources. Pp. 20-28 in Avian foraging:landbirds (J. M. Hagan, III and D. W. theory, methodology, and applications (M. L.Johnston, eds). Smithsonian Institution Press, Morrison, C. J. Ralph, J. Verner, and J. R.Washington, DC. Jehl, Jr, eds). Stud. Avian BioI. no.13. Cooper

Holmes, R. T., T. W. Sherry, and F. W. Sturges. Ornithological Society, Lawrence, KS.1986. Bird community dynamics in a Hut to, R. L. 1992. Habitat distributions oftemperate deciduous forest: long-term trends migratory landbird species in western Mexico.at Hubbard Brook. Ecol. Monogr. 50: Pp. 211-239 in Ecology and conservation of201-220. Neotropical migrant landbirds (J. M. Hagan,

Holmes, R. T., T. W. Sherry, and L. Reitsma. 1989. III and D. W. Johnston, eds). SmithsonianPopulation structure, territoriality and over- Institution Press, Washington, DC.winter survival of two migrant warbler species Hut to, R. L., S. M. Pletschet, and P. Hendricks.in Jamaica. Condor 91:545-561. 1986. A fixed-radius point count method for

Howell, T. R. 1970. Avifauna in Nicaragua. Pp. nonbreeding and breeding season use. Auk58-62 in The avifauna of northern Latin 103:593-602.America (H. K. Buechner and J. H. Buechner, Hut to, R. L., S. Reel, and P. B. Landres. 1987. Aeds). Smithsonian Contr. Zool. no.26. critical evaluation of the species approach toSmithsonian Institution Press, Washington, biological conservation. Endangered SpeciesDC. Update 4:1-4.

Howell, T. R. 1971. An ecological study of the )ames, F. C. 1971. Ordinations of habitatbirds of the lowland pine savanna and relationships among breeding birds. Wilsonadjacent rainforest in northeastern Nicaragua. Bull. 83:215-236.Living Bird 10: 185-242. Janzen, D. H. 1973. Sweep samples of tropical

Hussell, D. J. T., M. H. Mather, and P. M. Sinclair. foliage insects: effects of seasons, vegetation1992. Trends in numbers of tropical- and types, elevation, time of day, and insularity.temperate-wintering migrant landbirds in Ecology 54:687- 708.migration at Long Point, Ontario, 1961-1988. Janzen, D. H. 1988. Tropical dry forests: the mostPp. 101-114 in Ecology and conservation of endangered major tropical ecosystem. Pp.Neotropical migrant landbirds (J. M. Hagan, 130-137 in Biodiversity (E. 0. Wilson, ed.).III and D. W. Johnston, eds). Smithsonian National Academy Press, Washington,Institution Press, Washington, DC. DC.

Hut to, R. L. 1980. Winter habitat distribution of Janzen, D. H., and T. W. Schoener. 1968.migratory land birds in western Mexico with Differences in insect abundance and diversityspecial reference to small, foliage-gleaning between wetter and drier sites during ainsectivores. Pp. 181-204 in Migrant birds in tropical dry season. Ecology 49:96-110.the Neotropics: ecology, behavior, dis- Johns, A. D. 1989. Recovery of a peninsulartribution, and conservation (A. Keast and E. Malaysian rain forest avifauna followingS. Morton, eds). Smithsonian Institution selective timber logging: the first twelve years.Press, Washington, DC. ForktaiI4:89-105.

Hut to, R. L. 1985. Habitat selection by Johnston, D. W., and J. M. Hagan, III. 1992. Annonbreeding, migratory land birds. Pp. analysis of long-term breeding bird censuses~


from eastern deciduous forests. Pp. 75-84 in Kricher, J. C., and W. E. Davis, Jr. 1992. PatternsEcology and conservation of Neotropical of avian species richness in disturbed andmigrant landbirds (J. M. Hagan, III and D. undisturbed habitats in Belize. Pp. 240-246W. Johnston, eds). Smithsonian Institution in Ecology and conservation of NeotropicalPress, Washington, DC. migrant landbirds (J. M. Hagan, III and D.

Kapos, V. 1989. Effects of isolation on the water W. Johnston, eds). Smithsonian Institutionstatus of forest patches in the Brazilian Press, Washington, DC.Amazon. J. Tropical Ecol. 5:173-185. Lack, D., and P. Lack. 1972. Wintering warblers

Karr, J. R. 1976. On the relative abundance of in Jamaica. Living Bird 11:129-153.migrants from the north temperate zone in Land, H. C. 1970. Birds of Guatemala. Livingston,tropical habitats. Wilson Bull. 88:433-458. Wynnewood, PA.

Karr, J. R. 1981. Surveying birds in the tropics. Laurance, W. F. 1991. Edge effects in tropicalPp. 548-553 in Estimating numbers of forest fragments: application of a model forterrestrial birds (C. J. Ralph and J. M. Scott, the design of nature reserves. Bioi. Conserv.eds). Stud. Avian Bioi. no.6. Cooper 57:205-219.Ornithological Society, Lawrence, KS. Laurance, W. F., and E. Yensen. 1991. Predicting

Karr, J. R. 1990. Avian extinction rates and the the impacts of edge effects in fragmentedextinction process on Barro Colorado Island, habitats. Bioi. Conserv. 55:77-92.Panama. Conserv. Bioi. 4:391-397. Leck, C. F. 1972a. The impact of some North

Karr, J. R., and D. R. Dudley. 1981. Ecological American migrants at fruiting trees inperspective on water quality goals. Environ. Panama. Auk 89:842-850.Manag.5:55-68. Leck, C. F. 1972b. Observations of birds at

Karr, J. R., and K. E. Freemark. 1983. Habitat Cecropia trees in Puerto Rico. Wilson Bull.selection and environmental gradients: 84:498-500.dynamics in the "stable" tropics. Ecology Leonard, H. J. 1987. Natural resources and64: 1481-1494. economic development in Central America: a

Karr, J. R., D. W. Schemske, and N. Brokaw. 1982. regional environmental profile. TransactionTemporal variation in the undergrowth bird Books, New Brunswick, NJ.community of a tropical forest. Pp. 441-453 Levey, D. J. 1988. Spatial and temporal variationin Ecology of a tropical forest: seasonal in Costa Rican fruit and fruit-eating birdrhythms and long-term changes (E. G. Leigh, abundance. Ecol. Monogr. 58:251-269.A. S. Rand, and D. Windsor, eds). Levey, D. J., and F. G. Stiles. 1992. EvolutionarySmithsonian Institution Press, Washington, precursors of long-distance migration: re-DC. source availability and movement patterns

Keast, A. 1980. Spatial relationships between mi- in Neotropical landbirds. Amer. Natur.gratory parulid warblers and their ecological 140:447-476.counterparts in the Neotropics. Pp. 109-130 Levins, R. 1968. Evolution in changing environ-in Migrant birds in the Neotropics: ecology, ments. Princeton University Press, Princeton,behavior, distribution, and conservation (A. NJ.Keast and E. S. Morton, eds). Smithsonian Loiselle, B. A., and J. G. Blake. 1991. TemporalInstitution Press, Washington, DC. variation in birds and fruits along an

Keast, A., and E. S. Morton (eds). 1980. Migrant elevational gradient in Costa Rica. Ecologybirds in the Neotropics: ecology, behavior, 72:180-193.distribution, and conservation. Smith- Lopez Ornat, A. 1990. Ecologia de las passeri-sonian Institution Press, Washington, formes en la reserva de la biosfera de SianDC. Kaan, Mexico. PhD thesis, University

Kellman, M., and R. Tackaberry. 1993. Dis- Complutense de Madrid, Facultad deturbance and tree species coexistence in Ciencias Biologicas, Madrid.tropical riparian forest fragments. Global Lopez Ornat, A., and R. Greenberg. 1990.Ecol. Biogeog. Lett. 3: 1-9. Sexual segregation by habitat in migratory

Kennish, M. J. 1990. Ecology of estuaries. warblers in Quintana Roo, Mexico. AukVol. II, Biological aspects. CRC Press, Boca 107:539-543.Raton, FL. Lopez Ornat, A., and J. F. Lynch. 1991.

Ketterson, E. D., and V. Nolan, Jr. 1983. The Landbird communities in the coastal duneevolution of differential bird migration. Curr. scrub of the Yucatan Peninsula: speciesOrnithol. 1: 357-402. composition, ecology, and zoogeographic

Koplin, J. R. 1973. Differential habitat use by sexes affinities. Fauna Silv. Neotrop. 2:21-31.of American Kestrels wintering in northern Lovejoy, T. E., R. 0. Bierregaard, Jr, A. B. Rylands,California. J. Raptor Res. 7:39-42. J. R. Malcolm, C. E. Quintela, L. H. Harper,


K.S.Brown,Jr,A.H.Powell,G.V.N.Powell, Marra, P. P., R. T. Holmes, and T. W. Sherry.H. 0. R. Schubart, and M. B. Hayes. 1986. 1993. Territorial exclusion by a temperate-Edge and other effects of isolation on Amazon tropical migrant warbler in Jamaica: aforest fragments. Pp. 257-285 in Conser- removal experiment with American Redstartsvation biology: the science of scarcity and (Setophaga ruticilla). Auk 110: 565-572.diversity (M. E. Soule, ed.). Sinauer Martin, T. E. 1985. Selection of second-growthAssociates, Sunderland, MA. woodlands by frugivorous migrating birds in

Lugo, A. E., R. Schmidt, and S. Brown. 1981. Panama: an effect of fruit size and plantTropical forests in the Caribbean. Ambio density? J. Tropical Ecol. 1:157-170.10:318-324. Martin, T. E. 1992. Breeding productivity

Lynch, J. F. 1989. Distribution of overwintering considerations: what are the appropriatenearctic migrants in the Yucatan Peninsula, habitat features for management? Pp.I: general patterns of occurrence. Condor 455-473 in Ecology and conservation91 :515-544. of Neotropical migrant landbirds (J. M.

Lynch, J. F. 1991. Effects of Hurricane Gilbert on Hagan, III and D. W. Johnston, eds).birds in a dry tropical forest in the Yucatan Smithsonian Institution Press, Washington,Peninsula. Biotropica 23:488-496. DC.

Lynch, J. F. 1992. Distribution of overwintering Martin, T. E., and J. R. Karr. 1986. Temporalnearctic migrants in the Yucatan Peninsula, dynamics of Neotropical birds with specialII: use of native and human-modified reference to frugivores in second-growthvegetation. Pp. 178-196 in Ecology and woods. Wilson Bull. 98:38-60.conservation of Neotropical migrant land- McElroy, J. L., B. Potter, and E. Towle. 1990.birds (J. M. Hagan, III and D. W. John- Challenges for sustainable development inston, eds). Smithsonian Institution Press, small Caribbean islands. Pp. 299-316 inWashington, DC. Sustainable development and environmental

Lynch, J. F., and D. F. Whigham. 1984. Effects of management of small islands (W. Beller, P.forest fragmentation on breeding bird d' Ayala, and P. Hein, eds). Man andcommunities in Maryland, USA. Bioi. the Biosphere Series, Vol. 5. UNESCO,Conserv. 28:287-324. Paris.

Lynch, J. F., E. S. Morton, and M. E. van der Miles, D. B., and R. E. Ricklefs. 1984. TheVoort. 1985. Habitat segregation between the correlation between ecology and morphologysexes of wintering Hooded Warblers ( Wilsonia in deciduous forest passerine birds. Ecologycitrina). Auk 102:714-721. 65:1629-1640.

Mabey, S. E., and E. S. Morton. 1992. Mills, E. D., and D. T. Rogers, Jr. 1992. Ratios ofDemography and territQrial behavior of Neotropical migrants and Neotropicalwintering Kentucky Warblers in Panama. Pp. resident birds in winter in a citrus planta-329-336 in Ecology and conservation of tion in central Belize. J. Field Ornithol.Neotropical migrant landbirds (J. M. Hagan, 63:109-116.III and D. W. Johnston, eds). Smithsonian Mills, L. S., M. E. Soule, and D. F. Doak. 1993.Institution Press, Washington, DC. The keystone-species concept in ecology and

MacArthur, R. H., and J. MacArthur. 1961. on conservation. BioScience 43:219-224.bird species diversity. Ecology 42:594-598. Monroe, B. L. 1968. A distributional survey of the

MacAthur, R. H., and E. 0. Wilson. 1967. The birds of Honduras. Ornithol. Monogr. no.theory of island biogeography. Princeton 7:1-458.University Press, Princeton, NJ. Monroe, B. L. 1970. Effects of habitat change on

MacClintock, L., R. F. Whitcomb, and B. L. population levels of the avifauna inWhitcomb. 1977. Island biogeography and Honduras. Pp. 38-44 in The avifauna of"habitat islands" of eastern forest. II. northern Latin America (H. K. Buechner andEvidence of the value of corridors and J. H. Buechner, eds). Smithsonian Contr.minimization of isolation in preservation Zool. no.26. Smithsonian Institution Press,of biotic diversity. Amer. Birds Washington, DC.31:6-16. Moore, F. R., and T. R. Simons. 1992. Habitat

Margules, C., and M. B. Usher. 1981. Criteria used suitability and stopover ecology of Neotro-in assessing wildlife conservation potential: a pical landbird migrants. Pp. 345-355 inreview. Bioi. Conserv. 21:79-109. Migrant birds in the Neotropics: ecology,

Margules, C., A. 0. Nicholls, and R. L. Pressey. behavior, distribution, and conservation1988. Selecting networks of reserves to (A. Keast and E. S. Morton, eds). Smith-maximise biological diversity. Bioi. Conserv. sonian Institution Press, Washington,43:63-76. DC.


Moore, F. R., S. A. Gauthreaux, Jr, P. Kerlinger, Myers, N. 1980. Conversion of tropical moistand T. R. Simons. 1993. Stopover habitat: forests. National Academy of Sciences,management implications and guidelines. Pp. Washington, DC.58-69 in Status and management of Myers, N. 1991. Tropical forests: presentNeotropical migratory birds (D. M. Finch status and future outlook. Clim. Changeand P. W. Stangel, eds). Gen. Tech. Rep. 19:3-32.RM-229. USDA Forest Serv., Rocky Mt. Noss, R. F. 1987. From plant communities toForest Range Exp. Sta., Fort Collins, CO. landscapes in conservation inventories: a look

Moreau, R. E. 1972. The Palearctic-African bird at the Nature Conservancy (USA). BioI.migration systems. Academic Press, New Conserv.41:11-37.York. Noss, R. F., and L. D. Harris. 1986. Nodes,

Morse, D. H. 1971. The insectivorous bird as an networks and MUMs: preserving diver-adaptive strategy. Annu. Rev. Ecol. Syst. sity at all scales. Environ. Manag. 10:2: 177-200. 299-309.

Morse, D. H. 1980. Population limitation: Odum, E. P. 1971. Fundamentals of ecology.breeding or wintering grounds? Pp. 505-516 Saunders, Philadelphia, PA.in Migrant birds in the Neotropics: ecology, O'Neill, R. V., D. L. DeAngelis, J. B. Wade, andbehavior, distribution, and conservation (A. T. F. H. Allen. 1986. A hierarchy concept ofKeast and E. S. Morton, eds). Smithsonian ecosystems. Princeton University Press,Institution Press, Washington, DC. Princeton, NJ.

Morton, E. S. 1980. Adaptations to seasonal Orejuela, J. E., R. J. Raitt, and H. Alvarez. 1980.changes by migrant land birds in the Panama Differential use by North American migrantsCanal Zone. Pp. 437-453 in Migrant birds in of three types of Colombian forests. Pp.the Neotropics: ecology, behavior, distribu- 253-264 in Migrant birds in the Neotropics:tion, and conservation (A. Keast and E. S. ecology, behavior, distribution, and con-Morton, eds). Smithsonian Institution Press, servation (A. Keast and E. S. Morton, eds).Washington, DC. Smithsonian Institution Press, Washington,

Morton, E. S. 1990. Habitat segregation by sex in DC.the Hooded Warbler: experiments on Orians, G. H. 1969. The number of bird species inproximate causation and discussion of its some tropical forests. Ecology 50:783-801.evolution. Amer. Natur. 135:319-333. Orians, G. H., and J. F. Wittenberger. 1991. Spatial

Morton, E. S. 1992. What do we know about the and temporal scales in habitat selection.future of migrant landbirds? Pp. 579-589 in Amer. Natur. 137:S29-S49.Ecology and conservation of Neotropical Paine, R. T. 1969. The Pisaster- Tegula interaction:migrant landbirds (J. M. Hagan, III and D. prey patches, predator food perference, andW. Johnston, eds). Smithsonian Institution intertidal community structure. EcologyPress, Washington, DC. 50:950-961.

Morton, E. S., and R. Greenberg. 1989. The Parrish, J. D., and T. W. Sherry. 1994. Sexualoutlook for migratory songbirds: "Future habitat segregation by American Redstartsshock" for birders. Amer. Birds 43: 178-183. wintering in Jamaica: importance of resource

Morton, E. S., J. F. Lynch, K. Young, and P. seasonality. Auk 111:38-49.Mehlhop. 1987. Do male Hooded Warblers Pashley, D. N. 1988. Warblers of the West Indies.exclude females from nonbreeding territories II. The Lesser Antilles. Carib. J. Sci.in tropical forest? Auk 104:133-135. 24:112-126.

Moynihan, M. 1962. The organization and Paynter, R. A. 1955. The ornithogeography of theprobable evolution of some mixed species Yucatan Peninsula. Peabody Mus. Nat. Hist.,flocks of Neotropical birds. Smithsonian Yale Univ. Bull. 9:1-347.Misc. ColI. 143:1-140. Pearson, D. L. 1980. Bird migration in Amazon-

Munn, C. A. 1985. Permanent canopy and ian Ecuador, Peru, and Bolivia. Pp. 273-understory flocks in Amazonia: species 283 in Migrant birds in the Neotropics:composition and population density. Pp. ecology, behavior, distribution, and conser-683-712 in Neotropical ornithology (P. A. vation (A. Keast and E. S. Morton, eds).Buckley, M. S. Foster, E. S. Morton, R. S. Smithsonian Institution Press, Washington,Ridgely, and F. G. Buckley, eds). Ornithol. DC.Monogr. no.36. American Ornithologists' Petit, D. R. 1991. Habitat associations ofUnion, Washington, DC. migratory birds wintering in Belize, Central

Murphy, P. G., and A. E. Lugo. 1986. Ecology of America: implications for theory andtropical dry forest. Annu. Rev. Ecol. Syst. conservation. PhD dissertation, Univeristy of17:67-88. Arkansas, Fayetteville, AR.


Petit, D. R., L. J. Petit, and K. G. Smith. 1992. Neotropical migrant land birds (J. M. Hagan,Habitat associations of migratory birds III and D. W. Johnston, eds). Smithsonianoverwintering in Belize, Central America. Pp. Institution Press, Washington, DC.246-256 in Ecology and conservation of Probst, J. R., and T. R. Crow. 1991. IntegratingNeotropical migrant land birds (J. M. Hagan, biological diversity and resource management.III and D. W. Johnston, eds). Smithsonian J. Forestry 89:12-17.Institution Press, Washington, DC. Rabinowitz, D., S. Cairns, and T. Dillon. 1986.

Petit, D. R., J. F. Lynch, R. L. Hutto, J. G. Blake, Seven forms of rarity and their frequency inand R. B. Waide. 1993. Management and the flora of the British Isles. Pp. 182-204 inconservation of migratory landbirds over- Conservation biology: the science of scarcitywintering in the Neotropics. Pp. 70-92 in and diversity (M. E. Soule, ed.). SinauerStatus and management of Neotropical Associates, Sunderland, MA.migratory birds (D. M. Finch and P. W. Radabaugh, B. E. 1974. Kirtland's Warbler andStangel, eds). Gen. Tech. Rep. RM-229. its Bahama wintering grounds. Wilson Bull.USDA Forest Serv., Rocky Mt, Forest 86:374-383.Range Exp. Sta., Fort Collins, CO. Ralph, C. J., and J. M. Scott (eds). 1981.

Petit, D. R., L. J. Petit, V. A. Sa ab, and T. E. Estimating numbers of terrestrial birds. Stud.Martin. 1995. Fixed-radius point counts in Avian Bioi. no.6, Cooper Ornithologicalforests: factors influencing effectiveness and Society, Lawrence, KS.efficiency. Pp. 54-61 in Proceedings of the Ralph, C. J., J. R. Sauer, and S. Droege (eds). 1995.symposium on monitoring bird population Proceedings of the symposium on monitoringtrends by point counts (C. J. Ralph, J. R. bird population trends by point counts.Sauer, and S. Droege, eds). Gen. Tech. Rep. Gen. Tech. Rep. PSW-GTR. USDA ForestPSW-GTR. USDA Forest Serv., Pacific Serv., Pacific Southwest Res. Sta., Albany,Southwest Res. Sta., Albany, CA. CA.

Pielou, E. C. 1975. Ecological diversity. John Wiley Ramos, M. A., and D. W. Warner. 1980. Analysisand Sons, New York. of North American subspecies of migrant

Plowman, K. P. 1979. Litter and soil fauna of two birds wintering in Los Tuxtlas, southernAustralian subtropical forests. Austral. J. Veracruz, Mexico. Pp. 173-180 in MigrantEcol.4:87-104. birds in the Neotropics: ecology, behavior,

Post, W. 1978. Social and foraging behavior of distribution, and conservation (A. Keast andwarblers wintering in Puerto Rican coastal E. S. Morton, eds). Smithsonian Institutionscrub. Wilson Bull. 90: 197-214. Press, Washington, DC.

Powell, G. V. N. 1979. Structure and dynamics of Rand, A. L. 1954. Social feeding behavior of birds.interspecific flocks in a Neotropical mid- Fieldiana Zool. 36:5-71.elevation forest. Auk 96:375-390. Rappole, J. H. 1991. Migrant birds in Neotropical

Powell, G. V. N. 1985. Sociobiology and adaptive forest: a review from a conservationsignificance of interspecific foraging flocks in perspective. Pp. 259-277 in Conservingthe Neotropics. Pp. 713- 732 in Neotropical migratory birds (T. Salathe, ed.). ICBP Tech.ornithology (P. A. Buckley, M. S. Foster, E. Publ. no.12, Cambridge, England.S. Morton, R. S. Ridgely, and F. G. Buckley, Rappole, J. H., and M. V. McDonald 1994. Causeeds). Ornithological Monogr. no.36. American and effect in population declines of migratoryOrnithologists' Union, Washington, DC. birds. Auk 111:652-660.

Powell, G. V. N. 1988. Mixed species flocking as Rappole, J. H., and D. W. Warner. 1980. Ecologicala strategy for Neotropical residents. Pp. aspects of avian migrant behavior in813-819 in Acta XIX Congressus Inter- Veracruz, Mexico. Pp. 353-393 in Migrantnationalis Ornithologici (H. Ouellet, ed.). birds in the Neotropics: ecology, behavior,University of Ottawa Press, Ottawa, Canada. distribution, and conservation (A. Keast and

Powell, G. V. N., and R. D. Bjork. 1994. E. S. Morton, eds). Smithsonian InstitutionImplications of altitudinal migration for Press, Washington, DC.conservation strategies to protect tropical Rappole, J. H., and E. S. Morton. 1985. Effects ofbiodiversity: a case study of the Resplendent habitat alteration on a tropical avian forestQuetzal Pharomacrus mocinno at Monteverde, community. Pp. 1013-1021 in NeotropicalCosta Rica. Bird Conserv. Int. 4:161-174. ornithology (P. A. Buckley, M. S. Foster, E.

Powell, G. V. N., J. H. Rappole, and S. A. Sader. S. Morton, R. S. Ridgely, and F. G. Buckley,1992. Neotropical migrant landbird use of eds). Ornithol. Monogr. no.36. Americanlowland Atlantic habitats in Costa Rica: a test Ornithologists' Union, Washington, DC.of remote sensing for identification of habitat. Rappole, J. H., Morton, E. S., T. E. Lovejoy, III,Pp. 287-298 in Ecology and conservation of and J. L. Ruos. 1983. Nearctic avian migrants


in the Neotropics. US Fish Wildl. Service, ecology and management (S. L. Sutton, T. C.Washington, DC. Whitmore, and A. C. Chadwick, eds).

Rappole, J. H., M. A. Ramos, and K. Winker. Blackwell Scientific Publications, Oxford,1989. Wintering Wood Thrush movements England.and mortality in southern Veracruz. Auk Russell, S. M. 1964. A distributional study of the106:402-410. birds of British Honduras. Ornithological

Reed, J. M. 1992. A system for ranking Monogr. no.1. American Ornithologists'conservation priorities for Neotropical Union, Washington, DC.migrant birds based on relative susceptibility Russell, S. M. 1970. Avifauna in British Honduras.to extinction. Pp. 524-536 in Ecology and Pp. 45-49 in The avifauna of northern Latinconservation of Neotropical migrant land- America (H. K. Buechner and J. H. Buechner,birds (J. M. Hagan, III and D. W. Johnston, eds). Smithsonian Contr. Zool. no.26. Smith-eds). Smithsonian Institution Press, Wash- sonian Institution Press, Washington, DC.ington, DC. Rzedowski, J. 1983. Vegetacion de Mexico.

Reznicek, A. A. 1987. Are small reserves Editorial Limusa, Mexico, DF.worthwhile for plants? Endangered Species Sa ab, V. A., and D. R. Petit. 1992. Impact ofUpdate 5:1-3. pasture development on winter bird com-

Ridgely, R. S., and J. A. Gwynne. 1989. A guide munities in Belize, Central America. Condorto the birds of Panama. Princeton University 94:66-71.Press, Princeton, NJ. Sader, S. A., and A. T. Joyce. 1988. Deforestation

Robbins, C. S., D. Bystrak, and P. H. Geissler. rates and trends in Costa Rica. Biotropica1986. The breeding bird survey: its first fifteen 20: 11-19.years, 1965-1979. Resource Publ. no.157. US Salwasser, H., C. K. Hamilton, W. B. Krohn,Fish Wildl. Service, Washington, DC. J. F. Lipscomb, and C. H. Thomas.

Robbins, C. S., B. A. Dowell, D. K. Dawson, J. 1983. Monitoring wildlife and fish: mandatesColon, F. Espinoza, J. Rodriguez, R. Sutton, and their implications. Trans. North Amer.and T. Vargas. 1987. Comparison of Wildl. Nat. Resour. Conf. 48:297-307.Neotropical winter bird populations in Sauer, J. R., and S. Droege. 1992. Geographicisolated patches versus extensive forest. Acta patterns in population trends of NeotropicalOecol./Oecol. Gen. 8:282-292. migrants in North America. Pp. 26-42 in

Robbins, C. S., J. R. Sauer, R. S. Greenberg, and Ecology and conservation of NeotropicalS. Droege. 1989a. Popoulation declines migrant landbirds (J. M. Hagan, III andin North American birds that migrate to D. W. Johnston, eds). Smithsonian Institutionthe Neotropics. Proc. Natl. Acad. Sci. Press, Washington, DC.86:7658-7662. Saunders, D. A., R. J. Hobbs, and C. R. Margules.

Robbins, C. S., D. K. Dawson, and B. A. Dowell. 1991. Biological consequences of ecosystem1989b. Habitat area requirements of breeding fragmentation: a review. Conserv. BioI.forest birds of the middle Atlantic States. 5: 18-32.Wildl. Monogr. no.103. The Wildlife Society, Sayer, J. 1991. Conservation and protectionWashington, DC. of tropical rain forests: the perspectives

Robbins, C. S., B. A. Dowell, D. K. Dawson, J. A. of the World Conservation Union. UnasylvaColon, R. Estrada, A. Sutton, R. Sutton, and 42:40-45.D. Weyer. 1992. Comparison of Neotropical Schwartz, P. 1964. The Northern Waterthrush inmigrant landbird populations wintering in Venezuela. Living Bird 3:169-184.tropical forest, isolated forest fragments, and Scott, J. M., B. Csuti, J. D. Jacobi, and J. E. Estes.agricultural habitats. Pp. 207-220 in Ecology 1987. Species richness: a geographic approachand conservation of Neotropical migrant to protecting future biological diversity.landbirds (J. M. Hagan, III and D. W. BioScience 37:782-788.Johnston, eds). Smithsonian Institution Press, Severinghaus, W. D. 1981. Guild theoryWashington, DC. development as a mechanism for assessing

Robinson, S. K., J. Terborgh, and J. W. Fitzpatrick. environmental impact. Environ. Manag.1988. Habitat selection and relative abun- 5: 187-190.dance of migrants in southeastern Peru. Shaffer, M. 1981. Minimum population sizes forPp. 2298-2307 in Acta XIX Congressus species conservation. BioScience 31: 131-134.Internationalis Ornithologici (H. Ouellet, Sherry, T. W., and R. T. Holmes. 1989. Age-specificed.). University of Ottawa Press, Ottawa, social dominance affects habitat use byCanada. breeding American Redstarts (Setophaga

Rubinoff, 1.1983. A strategy for preserving tropical rutici/1a): a removal experiment. Behav. Ecol.forests. Pp. 465-476 in Tropical rain forest: Sociobiol.25:327-333.


Sherry, T. W., and R. T. Holmes. 1991. servation (A. Keast and E. S. Morton, eds).Population age structure of long-distance Smithsonian Institution Press, Washington,migratory passerine birds: variation in space DC.and time. Pp. 1542-1556 in Acta xx Terborgh,J. 1986. Community aspects offrugivoryCongressus Intemationalis Omithologici. in tropical forests. Pp. 371-384 in Frugi-New Zealand Ornithological Congress Trust vores and seed dispersal (A. Estrada andBoard, Wellington. T. H. Flemming, eds). Dr. W. Junk, The

Simberloff, D., and J. Cox. 1987. Consequences Hague, The Netherlands.and costs of conservation corridors. Conserv. Terborgh, J. 1989. Where have all the birdsBioI. 1 :63- 71. gone? Princeton University Press, Princeton,

Sliwa, A., and T. W. Sherry. 1992. Surveying winter NJ.Warbler populations in Jamaica: point counts Terborgh, J. 1992. Maintenance of diversity inwith and without broadcast vocalizations. tropical forests. Biotropica 24:283-292.Condor 94:924-936. Terborgh, J., and J. R. Faaborg. 1980. Factors

Slud, P. 1960. The birds of Finca "La Selva ", Costa affecting the distribution and abundance ofRica: a tropical wet forest locality. Bull. Amer. North American migrants in the easternMus. Natural Hist. 121: 1-148. Caribbean region. Pp. 145-156 in Migrant

Staicer, C. A. 1992. Social behavior of the Northern birds in the Neotropics: ecology, behavior,Parula, Cape May Warbler, and Prairie distribution, and conservation (A. Keast andWarbler wintering in second-growth forest in E. S. Morton, eds). Smithsonian Institutionsouthwestern Puerto Rico. Pp. 308-320 in Press, Washington, DC.Ecology and conservation of Neotropical Terborgh, J., and B. Winter. 1980. Some causes ofmigrant landbirds (J. M. Hagan, III and D. extinction. Pp. 119-134 in ConservationW. Johnston, eds). Smithsonian Institution biology: an evolutionary-ecological perspec-Press, Washington, DC. tive (M. E. Soule, ed.). Sinauer Associates,

Stewart, P. A. 1987. Decline in numbers of wood Sunderland, MA.warblers in spring and autumn migrations Terborgh, J., and B. Winter. 1983. A method forthrough Ohio. North Amer. Bird Bander siting parks and reserves with special12:58-60. reference to Colombia and Ecuador. BioI.

Stiles, F. a. 1980. Evolutionary implications of Conserv. 27:45-58.habitat relations between permanent and Terborgh, J., S. K. Robinson, T. A. Parker, Ill, C.winter resident landbirds in Costa Rica. Pp. A. Munn, and N. Pierpont. 1990. Structure421-435 in Migrant birds in the Neotropics: and organization ofan Amazonian forest birdecology, behavior, distribution, and con- community. Ecol. Monogr. 60:213-238.servation (A. Keast and E. S. Morton, eds). Thelen, K. D. 1990. Networking to share technicalSmithsonian Institution Press, Washington, knowledge on wildlife management in LatinDC. America. Unasylva 41:33-38.

Stiles, F. a., 1985. Conservation of forest birds in Thiollay, J.-M. 1992. Influence of selective loggingCosta Rica: problems and perspectives. Pp. on bird species diversity in a auianan rain141-168 in Conservation of tropical forest forest. Conserv. BioI. 6:47-63.birds (A. W. Diamond and T. E. Lovejoy, Tramer,E.J.1974.ProportionsofwinteringNortheds). ICBP Tech. Publ. no.4. American birds in disturbed and undisturbed

Stiles, F. a., and A. F. Skutch. 1989. A guide to dry tropical habitats. Condor 76:460-464.the birds of Costa Rica. Comell University Tramer, E. J., and R. Kemp. 1980. ForagingPress, Ithaca, NY. ecology of migrant and resident warblers and

Strickland, A. H. 1947. The soil fauna of two vireos in the highlands of Costa Rica. Pp.contrasted plots of land in Trinidad, British 285-296 in Migrant birds in the Neotropics:West Indies. J. Anim. Ecol. 16:1-10. ecology, behavior, distribution, and con-

Sutton, S. L., T. C. Whitmore, and A. C. Chadwick servation (A. Keast and E. S. Morton, eds).(eds). 1983. Tropical rain forest ecology and Smithsonian Institution Press, Washington,management. Blackwell Scientific Publica- DC.tions, Oxford, England. Turner, M. a. 1989. Landscape ecology: the effect

Terborgh, J. 1974. Preservation of natural of pattern on process. Ann. Rev. Ecol. Syst.diversity: the problem of extinction prone 20:171-197.species. BioScience 24:715-722. Van Home, B. 1983. Density as a misleading

Terborgh, J. 1980. The conservation status of indicator of habitat quality. J. Wildl. Manag.Neotropical migrants: present and future. Pp. 47:893-901.21-30 in Migrant birds in the Neotropics: Van Home, B., and J. A. Wiens. 1991. Forest birdecology, behavior, distribution, and con- habitat suitability models and the develop-~


ment of general habitat models. Fish Wildl. Wilcove, D. S. 1988. Changes in the avifauna ofRes. no.8. US Fish Wildl. Service, the Great Smokey Mountains: 1947-1983.Washington, DC. Wilson Bull. 100:256-271.

Vannini, J. P. 1994. Nearctic avian migrants in Wilcove, D. S., and S. K. Robinson. 1990. Thecoffee plantations and forest fragments of impact of forest fragmentation on bird com-south-western Guatemala. Bird Conserv. Int. munities in eastern North America. Pp.4:209-232. 319-331 in Biogeography and ecology of

Vaughan, C. 1990. Patterns in natural resource forest bird communities (A. Keast, ed.). SPBdestruction and conservation in Central Academic Publishing, The Hague, TheAmerica: a case for optimism? Trans. North Netherlands.Amer. Wildl. Nat. Resour. Conf. 55:407-422. Williams-Linera, G. 1990. Vegetation structure

Waide, R. B. 1980. Resource partitioning between and environmental conditions of forest edgesmigrant and resident birds: the use of irregular in Panama. J. Ecol. 78:356-373.resources. Pp. 337-352 in Migrant birds in Willis, E. 0. 1966. The role of migrant birdsthe Neotropics: ecology, behavior, distribu- at swarms of army ants. Living Birdtion, and conservation (A. Keast and E. S. 5: 187-231.Morton, eds). Smithsonian Institution Press, Willis, E. 0. 1979. The composition of avianWashington, DC. communities in remanescent woodlots in

Waide, R. B., J. T. Emlen, and E. J. Tramer. 1980. southern Brazil. Papeis Avulsos Zool.Distribution of migrant birds in the Yucatan 33:1-25.Peninsula: a survey. Pp. 165-171 in Migrant Willis, E. 0. 1980. Ecological roles of migratorybirds in the Neotropics: ecology, behavior, and resident birds on Barro Colorado Island,distribution, and conservation (A. Keast and Panama. Pp. 205-225 in Migrant birds in theE. S. Morton, eds). Smithsonian Institution Neotropics: ecology, behavior, distribution,Press, Washington, DC. and conservation (A. Keast and E. S. Morton,

Webb, W. L., D. F. Behrend, and B. Saisorn. 1977. eds). Smithsonian Institution Press,Effects of logging on songbird populations in Washington, DC.a northern hardwood forest. Wildl. Monogr. Wilson, E. 0.1988. The current state of biologicalno.55. The Wildlife Society, Washington, DC. diversity. Pp. 3-18 in Biodiversity (E. 0.

Whitacre, D. F., J. M. Madrid, C. Marroquin, Wilson, ed.). National Academy Press,M. Schulze, L. Jones, J. Sutter, and A. J. Washington, DC.Baker. 1993. Migrant songbirds, habitat Winker, K., J. H. Rappole, and M. A. Ramos. 1990.change, and conservation prospects in Population dynamics of the Wood Thrushnorthern Peten, Guatemala: some initial in southern Veracruz, Mexico. Condorresults. Pp. 339-345 in Status and manage- 92:444-460.ment of Neotropical migratory birds (D. M. Wong, M. 1986. Trophic organization ofFinch and P. W. Stangel, eds). Gen. understory birds in a Malaysian dipterocarpTech. Rep. RM-229. USDA Forest Serv., forest. Auk 103: 100-116.Rocky Mt. Forest Range Exp. Sta., Fort Woodwell, G. M. 1991. Forests in a warmiI\gCollins, CO. world: a time for new policies. Clim. Change

Whitcomb, R. F., J. F. Lynch, P. A. Opler, and C. 19:245-251.S. Robbins. 1976. Island biogeography and World Resources Institute. 1992. World resourcesconservation: strategy and limitations. 1992-93. Oxford University Press, Oxford,Science 193:1030-1032. England.

Whitcomb, R. F., C. S. Robbins, J. F. Lynch, B. Wright, D. H. 1991. Correlations betweenL. Whitcomb, M. K. Klimkiewicz, and D. incidence and abundance are expected byBystrak. 1981. Effects of forest fragmentation chance. J. Biogeog. 18:463-466.on avifauna of eastern deciduous forest. Pp. Wunderle, J. M., Jr. 1992. Sexual habitat125-205 in Forest island dynamics in man- segregation in wintering Black-throated Bluedominated landscapes (R. L. Burgess and D. Warblers in Puerto Rico. Pp. 299-307 inM. Sharpe, eds). Springer- Verlag, New York. Ecology and conservation of Neotropical

Whitmore, T. C. 1991. An introduction to tropical migrant landbirds (J. M. Hagan, III and D.rain forests. Clarendon Press, Oxford, W. Johnston, eds). Smithsonian InstitutionEngland. Press, Washington, DC.

Wiens, J. A. 1977. On competition and variable Wunderle, J. M., Jr., and R. B. Waide. 1993.environments. Amer. Sci. 65:590-597. Distribution of overwintering Nearctic

Wilcove, D. S. 1985. Nest predation in forest tracts migrants in the Bahamas and Greaterand the decline of migratory songbirds. Antilles. Condor 95:904-933.Ecology 66: 1211-1214. Wunderle, J. M., Jr. and R. B. Waide. 1994. Future


prospects for Nearctic migrants wintering in fragmentation on depredation of artificialCaribbean forests. Bird Conserv. Int. nests. J. Wildl. Manag. 52:158-161.4: 191-207. Zar, J. H. 1984. Biostatistical analysis. Prentice

Yahner, R. H., and D. Scott. 1988. Effects offorest Hall, Englewood Cliffs, NJ.

habitat use and conservation of migratory landbirds overwintering in the neotropics

Documents