reproduction, habitat use, and natural history...

ORNITOLOGIA NEOTROPICAL 13: 121–136, 2002© The Neotropical Ornithological Society

REPRODUCTION, HABITAT USE, AND NATURAL HISTORY OF THE BLACK-AND-YELLOW SILKY-FLYCATCHER (PHAINOPTILA

MELANOXANTHA), AN ENDEMIC BIRD OF THE WESTERN PANAMA-COSTA RICAN HIGHLANDS

Gilbert Barrantes1 & Bette A. Loiselle2

1Escuela de BiologÍa, Universidad de Costa Rica, San José, Costa Rica, C. A. E-mail: [email protected]

2Bette A. Loiselle, Biology Department, University of Missouri-St. Louis, 8001 Natural Bridge Rd., St. Louis, MO 63121, U.S.A. E-mail: [email protected]

Resumen. – Reproducción, uso de habitat e historia natural del Capulinero negro y amarillo(Phainoptila melanoxantha), una especie endémica de las tierras altas de Costa Rica y oeste dePanamá. – Estudiamos la abundancia relativa, distribución geográfica, dieta y reproducción del Capuli-nero negro y amarillo (Phainoptila melanoxantha), una especie endémica a las tierras altas de Costa Rica yoeste de Panamá. A lo largo de su ámbito de distribución, el Capulinero negro y amarillo habita principal-mente los bosques montano alto, subalpino y enano. La abundancia de esta especie es mayor en los bos-ques subalpino y enano. En el bosque montano alto esta ave puede llegar a ser abundante en clarosnaturales y áreas de crecimiento secundario con abundantes frutos carnosos. La dieta del Capulinero negroy amarillo consiste exclusivamente de frutos. Esta especie se alimenta de una gran variedad de especies (36géneros de semillas fueron registrados en 58 contenidos estomacales). El número de géneros de semillasconsumidos es aparentemente mayor en la Cordillera de Talamanca (región sur de Costa Rica y oeste dePanamá). Este patrón es posiblemente causado por la mayor riqueza de especies de plantas en esta cordi-llera. La reproducción del Capulinero negro y amarillo ocurre principalmente de Marzo a finales de Mayo,antes de la estación lluviosa en las tierras altas de Costa Rica. El ajuste de la época reproductiva del Capuli-nero negro y amarillo con la estación lluviosa probablemente asegura un suministro de frutos suficientepara polluelos y volantones.

Abstract. – We studied relative abundance, geographical distribution, habitat use, diet, and reproductionof the Black-and-yellow Silky-flycatcher (Phainoptila melanoxantha), an endemic species of the highlands ofCosta Rica and western Panamá. Black-and-yellow Silky-flycatchers occupy mainly upper montane, subal-pine, and elfin forests throughout its geographical distribution. The abundance of this species is greater insubalpine and elfin forests; in upper montane forest Black-and-yellow Silky-flycatchers become abundantonly when fleshy fruits are abundant in natural gaps and second growth areas. The diet of this bird consistsexclusively of fruits. It feeds on a large variety of fruits (36 seed genera from 58 stomach contents) and thenumber of seed genera consumed is apparently greater in the Talamanca mountain range (southern regionof Costa Rica and western Panamá). This pattern is likely caused by the larger plant richness present in theTalamancas. Reproduction of the Black-and-yellow Silky-flycatcher occurs mainly from March throughMay, before the rainy season on the highlands of Costa Rica. Reproduction timing in Black-and-yellowSilky-flycatchers likely assures fruits available for nestlings and independent fledglings. Acceptd 10 October2001.

Key words: Phainoptila melanoxantha, Black-and-yellow Silky-flycatcher, local distribution, habitat use, dietcomposition, breeding biology, highlands, Costa Rica, Panamá.

121

BARRANTES & LOISELLE

INTRODUCTION

Knowledge on the biology and natural historyof tropical birds is essential for understandingrelationships among species, ecological pack-ing, niche partitioning, patterns of migration,and structure of communities (Diamond1975, Fjeldså 1985, Haffer & Fitzpatrick1985, Remsen 1985, Stiles 1985a). Unfortu-nately, very little is known about the naturalhistory of the great majority of tropical birdspecies. Despite the long legacy of ornitho-logical studies in southern Central America,birds of Costa Rican and western Panamáhighlands are no exception to this knowledgegap. The latter region contains the highestdegree of endemism in Central America (Slud1964, Stiles 1983, Hernández-Baños et al.1995, Watson & Peterson 1999). Of 72 high-land resident species (species present at 2600m or higher in the Talamanca mountains,sensu Wolf 1976) reported for this region inMiddle America, 50% (36 species) areendemic to these high elevations (Slud 1964,Wolf 1976, Stiles 1983).

The Black-and-yellow Silky-flycatcher isone of the 36 endemic species that inhabitsthe highlands in Costa Rica and western Pan-amá. Throughout its geographical distribu-tion, populations of this species show a cleardichotomy in plumage color; each morph isassociated to particular mountain ranges (Bar-rantes & Sánchez 2000). This geographicaldiscontinuity has led to the separation ofPhainoptila, a monotypic genus, into two sub-species: P. m. melanoxantha, which occurs fromthe Central Valley in Costa Rica to westernPanamá, and P. m. parkeri, which inhabits thenorthern mountains in Costa Rica (Barrantes& Sánchez 2000). The Black-and-yellowSilky-flycatcher differs markedly from othermembers of the family Bombycillidae (Slud1964, Stiles & Skutch 1986). For example,external appearance and behaviour of Phainop-tila are very atypical (Stiles & Skutch 1989).

Species of the other two related genera (Phain-opepla and Ptilogonys) are crested and possess auniform plumage, characteristics absent inBlack-and-yellow Silky-flycatchers. Also,Phainopepla and Ptilogonys species forage on thewing, capturing insects and plucking fruits,whereas Black-and-yellow Silky-flycatcherspick berries from a perch (cf. Stiles & Skutch1989). Other basic aspects of Black-and-yel-low Silky-flycatcher’s biology and natural his-tory are essentially unknown. This lack ofinformation has been attributed to the secre-tive habits and soft calls of this bird (Skutch1965). Here, we describe habitat use, diet, andreproductive behavior of this species in CostaRica.

METHODS

Abundance and habitat use. We visited 16 high-land sites throughout the four main mountainranges in Costa Rica (Talamanca, Central Vol-canic, Tilarán, and Guanacaste) between 2 to6 times each over a period of 26 months(December 1996–January 1999) (Fig. 1). Dur-ing each visit, we observed individuals ofBlack-and-yellow Silky-flycatchers and esti-mated habitat use and relative abundance inhighland forests above 2300 m in Talamanca,2000 m in Central Volcanic, and 1300 m inTilarán and Guanacaste mountains. In eachmountain range we covered areas withinlower montane, upper montane, and subal-pine (or elfin) forests (see Gómez 1986, Kap-pelle 1996). The lower limit of the study areasvaried between mountain ranges in order tocover appropriate vegetation types as theseshifted up and down in altitude with topogra-phy and size of the mountains. During eachvisit, we spent a similar proportion of time inlower montane forest, upper montane forest,and subalpine or elfin forest. In Tilarán, Gua-nacaste, and most part of the Central Volcanicmountain ranges, elfin forest occurs abovethe upper mountane forest zone replacing the

122

NATURAL HISTORY OF PHAINOPTILA MELANOXANTHA

subalpine forest. Considering the differencesin vegetation structure, composition, anddominance among these forests, we regardeach one as a different habitat for Black-and-yellow Silky-flycatchers.

Lower montane forest is a 2-strata forestwith an intermediate height canopy (25–30m), relatively dense understory, and lowspecies dominance (Hartshorn 1983). Treesare stout and often twisted with compactumbrella-like crowns; buttresses are uncom-mon. Upper montane forest has a similarcanopy height (except Quercus that may reach40 m), and is dominated by species in theMyrtaceae and Fagaceae families. The under-growth is dense with many individuals fromthe Rubiaceae, Melastomataceae, andChusquea (Poaceae) (Hartshorn 1983). Subal-

pine forests occur in the Talamanca moun-tains and on the Irazú volcano (CentralVolcanic mountain range) from 2900 to 3300m, between upper mountane forest andpáramo (Kappelle 1996). Plant families suchas Ericaceae, Winteraceae, Rosaceae, Arali-aceae, Clusiaceae, Caprifoliaceae, Myrsinaceaeand Cunoniaceae are common in the subal-pine forest zone (Gómez 1986). Vegetationcomposition of subalpine and elfin forests issimilar, although the structure, particularlythe height of the trees, differs between them.Elfin forests are found on exposed peaks andridges of all mountains but Talamanca andIrazú volcano. These forests are distinguishedby their low stature and the exuberant bryo-phytes that cloak the tree trunks and limbs(Lawton 1980). Elfin forests are composed of

FIG. 1. Map of Costa Rica showing main mountain ranges (Talamanca, Central Volcanic, Tilarán, andGuanacaste) and localities visited during this study.

123


dense growths of gnarled trees, generally 2- to10-m tall. Tree crowns in these forests aretightly packed into a single even canopy layer.The understory is dominated mainly by tree-lets and bushes of Melastomataceae (Miconiaspp), Rubiaceae (Psychotria spp), Araliaceae(Schefflera sp), Ericaceae, and Myrsinaceae(Myrsine spp). The ground is covered by a matof bryophytes from which Heliconia spp,Anthurium spp, bromeliads, palms, and otherherbs emerge (Stiles 1985b). The dominantcanopy species vary from one location toanother. For example, Shefflera sp is the domi-nant species in the Tilarán mountains (e.g.,Monteverde, Lawton 1980) and on CacaoVolcano in Guanacaste mountains (GB, pers.

observ.), whereas on the Miravalles massif,Guanacaste mountains, Clusia sp and Schefflerasp dominate the canopy (Stiles 1985b, GB,pers. observ.).

To estimate the relative abundance ofPhainoptila, we used abundance categoriesbased on the daily number of observations(Stiles 1980). Categories included: abundant(> 10 individuals recorded in 6 h), common(5–10 individuals recorded in 6 h), uncom-mon (1–4 individuals recorded in 6 h), andrare (1 individual recorded in two days). Thismethod was selected over transects or pointcounts, because this species has a very secre-tive behavior and soft calls that make it diffi-cult to detect while conducting transects or

TABLE 1. Relative abundance per forest type (habitat) of Black-and-yellow Silky-flycatchers in 16 sites inthe mountain ranges of Costa Rica. Localities are distributed throughout most of the geographical distri-bution of the species.

Localities Latitud-Longitud n1 Forest types

Lower montane Upper montane SubalpineTalamanca Cerro Echandi * Cerro Chirripó Cerro Cuericí Villa Mills * Reserva Tres de Junio *Central Volcanic Irazú Volcano Turrialba Volcano * Barva Volcano * Poás VolcanoTilarán Monteverde *Guanacaste Arenal Volcano Tenorio Volcano Miravalles Volcano * Rincón de la Vieja Volcano Cacao Volcano* Orosí Volcano

09°01'N, 82°49'W09°30'N, 83°30'W09°34'N, 83°38'W09°33'N, 83°43'W09°36'N, 83°49'W

09°59'N, 83°52'W10°01'N, 84°46'W10°08'N, 84°06'W10°11'N, 84°13'W

10°19'N, 84°47'W

10°27'N, 84°42'W10°40'N, 85°00'W 10°44'N, 85°09'W10°49'N, 85°19'W10°56'N, 85°27'W10°58'N, 85°28'W

56388

5463

6

228651

AbsentAbsentAbsentAbsentAbsent

AbsentAbsentAbsentAbsent

Absent

AbsentAbsentAbsentAbsentAbsentAbsent

AbundantRareRare

AbsentCommon

AbsentAbundant

UncommonUncommon

Rare

AbsentAbsentRareRare

AbsentAbsent

CommonRareRareRare

Abundant

RareAbundantAbundantCommon

Abundant

AbsentRare

AbundantRare

UncommonRare

1Number of visits to each locality.*Represent localities where specimens were collected.

124


point counts.We evaluated habitat use by Black-and-

yellow Silky-flycatchers within 16 study areasbased on observed abundance in each foresttype. During the initial two visits to each ofthe 16 study areas, we spent roughly equaltime in each forest type searching for Black-and-yellow Silky-flycatchers. We evaluatedhabitat use within each study area based onobserved abundance within each forest type.During later visits, we focused activities instudy areas and in forest types that werefound to contain more individuals of Black-and-yellow Silky-flycatchers (Table 1)

To estimate potential available habitatwithin each of the four mountain ranges, wefirst determined the elevation (i.e., foresttype) at which Phainoptila became rare. Forexample, in the Talamancas, the center ofabundance of Black-and-yellow Silky-fly-catchers is around 2700–2800 m and birdsbecome rare below 2400 m in elevation. Cor-responding levels where Black-and-yellowSilky-flycatchers become rare are 2400 m inTalamanca mountains, 2000 m in CentralVolcanic and 1400 m in Tilarán and Guana-caste mountains (Table 1). The area of poten-tial habitat in each mountain range was thencalculated using Geographic InformationSystem software (ArcView, ESRITM) on a dig-ital elevation map of Costa Rica (1:200,000;Lambert projection), and included areasabove the lower limit to the summit of thefour mountain ranges. Potential habitat avail-able is slightly overestimated in the Talaman-cas, because of the presence of páramo at thehighest elevations; Black-and-yellow Silky-fly-catchers are not found in páramo.

Diet. We quantified diet of Black-and-yellowSilky-flycatchers based on direct observationsof foraging birds and stomach contents frombirds collected for a genetic study. Fifty eightspecimens were collected from seven studyareas: Cerro Echandi, Villa Mills, and Reserva

Tres de Junio in the Talamanca mountains;Turrialba and Barva volcanos in the CentralVolcanic mountains; Monteverde in theTilarán mountains; and Miravalles and Cacaovolcanos in the Guanacaste mountains. Fromthese specimens, we identified the digestivetract contents (stomach contents hereafter).As these birds feed almost exclusively uponfruits, seeds contained in their stomachs wereidentified to the lowest level possible, mostlyto genus level. Seeds of different generafound in stomach contents were describedusing accumulation curves for each mountainrange, and slopes were compared using themethod described by Zar (1984: 292–302).We also calculated overlap in diet (seed gen-era) between mountain ranges using Pianka’sindex for niche overlap (Pianka 1973). Thesame index was used to calculate overlap offleshy-fruited plants (genera) between moun-tain ranges, using the plant data base of theInstituto Nacional de Biodiversidad (INBio)of Costa Rica. These data allowed us to evalu-ate whether diet overlap of Black-and-yellowSilky-flycatchers among mountain ranges iscorrelated with the overlap of plant speciespotentially eaten by this bird.

We also used information in the plantdata base at INBio to evaluate changes inspecies richness of plants potentially eaten byBlack-and-yellow Silky-flycatchers amongmountain ranges. Although this is themost complete database for plants in CostaRica, sampling efforts have not been homo-geneous across highland forests in all moun-tain ranges (J. Gonzales, pers. com.). Forexample, plants in the Tilarán mountainshave been the most intensively sampled,while the Central Volcanic range has been themost understudied. We analyzed the data atthe genus level, to reduce the errors due todifferential sampling effort. We compared,using a two-way ANOVA without repetition(model I), the number of plant genera foundwithin families across mountain ranges

125


for plants that are potentially eaten by Black-and-yellow Silky-flycatchers; main effects inthe model were mountain range and plantfamily.

Reproduction. We evaluated the reproductiveactivity of Black-and-yellow Silky-flycatchersusing gonad size. Reproductive males wereadult individuals having testes volume of 200mm3 or more, while reproductive females hadfollicles of 7 mm3 or more. Limits for repro-ductively active gonad size were establishedbased on comparisons with cloacal protuber-ance for males and brood patch in females forbirds of similar size (i.e., Turdus spp) usinginformation from museum specimens (Museode Zoología, Universidad de Costa Rica). Weselected this method because it likely is moreaccurate than using development of broodpatch or size of cloacal protuberance alone.Moreover, brood patch in female Black-and-yellow Silky-flycatchers are not as well devel-oped as in other birds (e.g., Turdus spp).

RESULTS AND DISCUSSION

Altitudinal distribution. The area of forestpotentially occupied by Black-and-yellowSilky-flycatches includes primarily subalpine(or elfin) and upper montane forests and

decreases in area from the Talamancas toGuanacaste (Table 2). The Talamancasaccount for 73% of the total habitat available(153,750 ha), Central Volcanic 18%, Tilarán6%, and Guanacaste 3% (Table 2). Uppermontane and subalpine forests occupy a con-tinuous band throughout the whole summitof the Talamanca mountains. The CentralVolcanic mountains present a different sce-nario. Forests occupied by Black-and-yellowSilky-flycatchers in this mountain range aredissected by two mountain passes: La Palma(at 1400m a.s.l.) between the massifs Irazú-Turrialba and Barva, and El Desengaño (at1550 m a.s.l.) between the Barva and Poásvolcanos. These depressions likely restrict, oreven eliminate, movement of birds betweenvolcanos. In the Tilarán mountains, the distri-bution of potential habitat is continuousalong the highest peaks (e.g., La Ventana andCerro Amigos within Monteverde Cloud For-est Reserve). The distribution of upper mon-tane and elfin forests, and consequently thedistribution of Black-and-yellow Silky-fly-catchers on Guanacaste mountains, isrestricted to the summit of the volcanos, withthe exception of the Arenal volcano, which ismostly devoid of vegetation at the summitdue to its intense volcanic activity. The Gua-nacaste mountain range constitutes a chain of

TABLE 2. Estimated area (hectares) of habitat available for Black-and-yellow Silky-flycatchers in the fourmain Costa Rican mountain ranges.

Talamanca1 Central Volcanic Tilarán GunacasteEchandiChirripó

Total

5,380148,370

153,750

IrazúTurrialba

BarvaPoás

15,86010,5005,8005,545

37,705

Monteverde 11,970

11,970

ArenalTenorio

MiravallesR. ViejaCacaoOrosi

317502000290010040

5,820

1Cuerici, Villa Mills, and Tres de Junio habitat are included into totals for Chirripó as habitat wascontinuous.

126


six volcanoes separated by foothills at about500 m in altitude. Here, the distribution ofBlack-and-yellow Silky-flycatchers likely con-sists of discrete populations restricted to theisolated mountaintops.

Abundance and habitat use. The relative abun-dance of Black-and-yellow Silky-flycatchersvaries greatly throughout its geographical dis-tribution (Table 1). For example, Black-and-yellow Silky-flycatchers are rare or uncom-mon at all locations of Guanacaste moun-tains, with the exception of the population onMiravalles volcano. In contrast, Black-and-yellow Silky-flycatchers are abundant or com-mon at most localities of the Central Volcanicmountains (Table 1). The source of this varia-tion in abundance is not completely clear.Local factors, such as vegetation structure,volcanic activity, and area of available habitatmay explain, in part, patterns of abundancefor the Black-and-yellow Silky-flycatcher(Tables 1 and 2).

The abundance of Black-and-yellowSilky-flycatchers also showed large variationamong forest types (Table 1). This bird usessubalpine/elfin forests intensively, and uppermontane forest to a lesser degree. However,Black-and-yellow Silky-flycatchers are absentin lower montane forests (Table 1). Forestpreference is likely partially explained by dif-ferences in abundance of fleshy-fruiting

plants among high elevation forests (Kap-pelle 1996). Subalpine and elfin forests aredominated by bird-dispersed plants (e.g., Ara-liaceae, Melastomataceae, Rubiaceae, Eri-caeae), while montane forests are dominatedby non-fleshy fruit species (e.g., Chusquea spp,Poaceae; Quercus spp, Fagaceae; Kappelle1996). However, natural gaps and advancedsecond growth areas in upper montane for-ests also contain abundant fleshy-fruitedplant species and are likely important forag-ing areas for Black-and-yellow Silky-flycatch-ers. Such areas contain vegetation similar tosubalpine forest (Kappelle 1996). The highconcentration of fruits in natural gaps andsecond growth areas largely determine thepresence of Black-and-yellow Silky-flycatch-ers in upper montane forests (GB, pers.observ.), sometimes below its center of abun-dance.

The relative abundance of Black-and-yel-low Silky-flycatchers across study sites islikely related to three factors: area of potentialhabitat available, recent anthropogenic activ-ity, and volcanic activity. For example, in theTalamanca mountains, Black-and-yellowSilky-flycatchers are abundant at both CerroEchandi and Reserva Tres de Junio. Thesetwo sites contain large extensions of subal-pine forests and second-growth areas due topast windstorms (southern part of the Tala-mancas) and past deforestation (Reserva Tres

TABLE 3. Dates of major eruptive events of Costa Rican volcanos based on Alavarado-Induni (2000).

Volcanoes Dates of major activitiesTurrialbaIrazúBarvaPoásArenalTenorioMiravallesRincon de la ViejaOrosi

1450 bC, 800 bC, 50, 650, 18661723, 1775, 1847, 19638050 years ago1834, 1860, 1880, 1888, 1895, 1899, 1903, 1910, 1914, 1925, 1929, 1932, 1953Since 1968No informationc. 16001860, 1863, 1912, 1922, 1966, 1970, 1983, 1991, 1992, 1995, 1998c. 6000 years ago

127


de Junio). Yet, at Villa Mills and Cuerecí, areasof both regeneration and subalpine forest arecommon, but Black-and-yellow Silky-fly-catchers are scarce at these localities (Table 1).Therefore, the presence of potential habitatalone does not explain differences in abun-dance of Black-and-yellow Silky-flycatchersamong localities. We know no reason whyCuerici and Villa Mills sites with apparentlygood available habitat and food resources,contain so few Black-and-yellow Silky-fly-catcher individuals.

Volcanic activity is another factor thatlikely influences abundance of Black-and-yel-low Silky-flycatchers on some massifs (seeTable 3). This bird is totally absent from theArenal volcano (Guanacaste mountain range),which has been active since 1968. Such activ-ity has destroyed the vegetation on practicallythe whole massif and thus there is little or nohabitat available for this species. Black-and-yellow Silky-flycatchers are rare at Irazú vol-cano (Central Volcanic mountain range).Irazú erupted in 1963 destroying all high veg-etation on the volcano and likely reducinggreatly, or eliminating, the Irazú population ofBlack-and-yellow Silky-flycatchers (see Table3). The relatively high abundance of this spe-cies on the Turrialba volcano, which has beeninactive since 1866 (Alvarado-Induni 2000)and only separated from the Irazú massif by10 km, supports the notion that volcanicactivity can impact abundance of Black-and-yellow Silky-flycatchers. Black-and-yellowSilky-flycatcher populations on Rincón de laVieja volcano have also been affected by vol-canic catastrophes. The activity of this vol-cano has increased during the last fourdecades (Table 3), and the elfin and highmontane forests have been partially or com-pletely destroyed several times since then. Onthe other hand, Poás volcano, which has beenactive since at least 1834 (Table 3), holds a rel-atively large population of Black-and-yellowSilky-flycatchers. The relatively high abun-

dance of Black-and-yellow Silky-flycatchesr inthis massif is likely explained by the connec-tivity that exists between Poás and Barva pop-ulations and the pattern of forest destructionproduced by the activity of the Poás volcano.Despite the existence of a mountain pass (ElDesengaño) between Barva and Poás volca-nos, this depression is insufficient to preventbird movement between these populations(GB, pers. observ.). The architecture of thecrater of the Poás volcano, and the foreststructure around the crater, suggest that theintense activity of this volcano has mostlyaffected the forest on the Pacific slope, leav-ing almost intact elfin and upper montaneforests on the Atlantic slope. The recurrentdestruction of the west slope forest in thePoás volcano has been registered since 1925(Alavarado-Induni 2000). Additionally, therapid colonization of pioneer elfin forest veg-etation after eruptions (GB, pers. observ.)lessens the effect of volcanic catastrophes onpopulation size, as habitat rapidly becomesavailable through regeneration.

The area of habitat (upper montane andelfin forests) available is a factor that likelyaffects abundance of Black-and-yellow Silky-flycatchers. For example, the low abundanceof Black-and-yellow Silky-flycatchers on theinactive Cacao and Orosi volcanos (Table 1) islikely a consequence of the small area (100 ha)of elfin and upper montane forests on thesemassifs. Although no information on volcanicactivity exists for Tenorio volcano, indicatingno recent activity, the relatively small amountof habitat available also probably explains thelow population size of Black-and-yellowSilky-flycatchers at this site.

Overall, volcanic activity directly and indi-rectly affects Black-and-yellow Silky-fly-catcher populations, when intense volcanicactivity destroys part (or all) of the vegetationused by Black-and-yellow Silky-flycatchersand partly (or completely) eliminates localpopulations. Other factors such as fruit avail-

128


ability and area of subalpine, elfin, and uppermontane forests, also play a significant role indetermining local abundance of Black-and-yellow Silky-flycatchers throughout its geo-graphical distribution.

Diet composition. The Black-and-yellow Silky-flycatcher is an obligate frugivore that evenfeeds fruits to its fledglings (GB, pers.observ.). We identified 36 genera of fruitsfrom 58 stomach contents (Table 4); only twostomachs contained insect parts. Thus, Black-and-yellow Silky-flycatchers consume a rela-tively large number of fruit species. Black-and-yellow Silky-flycatchers were furtherobserved to feed on 18 fruit species, 16 ofwhich were also present in stomach samples(Table 4). The number of seed speciespresent in the stomach samples varied littleamong localities and at least half of thestomachs contained seeds of only one species

(Table 5).The number of seed genera consumed by

Black-and-yellow Silky-flycatchers overall wasgreater in the Talamancas than elsewhere(Fig. 2). The slopes of cumulative curves ofseed genera consumed by birds (Talamanca =1.02, Central Volacanic = 0.36, Tialrán =0.75, and Guanacaste = 0.55) differ signifi-cantly among sites (F = 8.55, df = 3, 32, P <0.001) (Zar 1984: 300–302). We also foundsignificant difference when the two most sim-ilar slopes (Central Volcanic vs Guanacaste)were compared (t = –2.18, df = 16, P < 0.05).The comparison between Central Volcanicand Guanacaste slopes indicated significantdifferences among slopes of all mountainranges (Fig. 2). However, there is no evidencethat the curves are leveling off on any moun-tain range, indicating that the number of fruitgenera consumed is underestimated at allsites.

FIG. 2. Cumulative curves of seed morpho-species eaten by Black-and-yellow Silky-flycatchers in fourmountain ranges (Talamanca, Central Volcanic, Tilarán, and Guanacaste).

129


TABLE 4. Fruit genera or species identified from stomach contents of 58 specimens of the Black-and-yel-low Silky-flycatcher. Number of specimens in which a given plant species was found is included.

Fruit genera or species Talamanca Central Volcanic Tilarán Guanacaste

(25)1 (11) (10) (11)Alstroemeriaceae

Bomarea spAquifoliaceae

Ilex spAraceae

sp 1sp 2

AraliaceaeDendropanax spOreopanax spSchefflera robusta

ArecaceaeGeonoma sp

AsteraceaeClibadium sp

BoraginaceaeTournefortia sp

CaprifoliaceaeViburnum sp

ChloranthaceaeHedyosmum sp

ClusiaceaeClusia estenofila

EricaceaeVaccinium consanguineum

ErythroxylaceaeErythroxylum sp

EuphorbiaceaeHyeronima sp

Fabaceaesp 1

LoranthaceaeGaiadendron punctatum Psittacanthus sp

MyricaceaeMyrica sp

MyrsinaceaeArdisia sp Myrsine pittieri

PolygonaceaeMuehlenbeckia sp

RosaceaeHesperomeles sp

11

1

4*

1*

1

1*

1

2*

1*

1* 7*

2*

1

1

61+

+

1*

1

++

1

2 3*

1*

5*

6

1

6

1*

1*

2

130


Composition of the fruit genera eaten byBlack-and-yellow Silky-flycatchers also var-ied among mountain ranges. In general, over-lap in diet was highest between birds of theTalamanca and Central Volcanic ranges, andlowest between birds of the Talamanca andGuanacaste mountains Table 6). Differencesin diet overlap across mountain ranges maybe explained by differences in plant composi-tion at those sites. A significant positive cor-relation was found between diet overlapvalues (Pianka’s index) and overlap of fleshy-fruiting genera occurring among mountainranges (r = 0.62, n = 6, P < 0.05).

Our estimation of the number of fruits

consumed by the Black-and-yellow Silky-fly-catcher throughout its geographical range inthis study, however, is biased by samplingprocedures and sample sizes. Due to logisticalconstraints, sites were visited at differenttimes of the year (although repeated visits didoccur). Consequently, apparent differences indiet among sites are likely influenced by tem-poral and spatial variations in phenologicalpatterns, as well as fruit availability (Frankie etal. 1974, Opler et al. 1980, Loiselle & Blake1991), and thus diet overlap values must cer-tainly be underestimated. In addition, oursample sizes are small for all mountain ranges(Fig. 2). Continued sampling would probably

TABLE 4. Continuation.

Fruit genera or species Talamanca Central Volcanic Tilarán Guanacaste

(25)1 (11) (10) (11)Rubiaceae

Guettarda poasana Psychotria sp

SapindaceaeCupania sp

SmilacaceaeSmilax sp

SolanaceaeCestrum sp Solanum sp sp 1

SymplocaceaeSymplocos sp

ThymelaeaceaeDapnopsis sp

WinteraceaeDrimys winteri

Undeterminedsp 1sp 2sp 3sp 4

7

1

1*

1

1

4*

1

11

9 1* 1*

1

+

1

3*

1

1Number of stomachs analysed.*indicates plants on which birds were observed consuming fruits but not recorded in stomach contents.+indicates plants on which birds were observed consuming fruits and also present in stomach contents.

131


result in more fruit species being recorded atall sites (Fig. 2).

Another factor that may also influencefruit selection by Black-and-yellow Silky-fly-catchers is the variation in plant species rich-ness among mountain ranges (Lawton &Dryer 1980, Kappelle 1996, INBio data base).Mountain ranges differ significantly in rich-ness of highland plant genera (Table 7). Inaddition, ANOVA results indicated that plantfamilies differed in number of genera. This isnot a surprising result and family wasincluded in the ANOVA as a second sourceof variation. In general, results show that theTalamanca and Tilarán mountains containmore plant genera per family than either theCentral Volcanic or Guanacaste mountains(Fig. 3). The Tilarán mountains have beenintensively sampled by botanists, while theCentral Volcanic is relatively undersampled (J.Gonzáles, pers. com.). Consequently, differ-ences in sampling effort likely confound

the geographical pattern of highland plantspecies richness. We expect that a gradient ofdecreasing plant species richness occurs fromTalamanca to Guanacaste mountain, mirror-ing the reduction in area of highland forests(Fig 2). Nonetheless, Talamanca and Tilaránbirds did contain more seed species thanthose birds in Central Volcanic and Guana-caste, thus matching results of plant richness.

Reproduction. Breeding activity of Black-and-yellow Silky-flycatchers is poorly known(Skutch 1965). The secretive behavior of thisspecies makes courtship behavior difficult tostudy and nests difficult to find. Only twocup-shaped nests were found in this study;one in April and one in May (nest structurewas described by Stiles & Skutch 1989). Thefirst one was under construction in Miravallesvolcano (Guanacaste) and the second wasfound in Villa Mills (Talamanca) with twoeggs. Nests were found in elfin and subalpine

TABLE 5. Presence and abundance of seed genera or species obtained from stomach contents of Black-and-yellow Silky-flycatchers from Talamanca, Central Volcanic, Tilarán, and Guanacaste mountains. Stan-dard deviations are given in parentheses.

Cordillera Sample size Range of seed species/indiv.

Average seed species/indiv.

Average seed number/indiv.

Percentage of samples with one

seed speciesTalamancaCentralTilaránGuanacaste

25111012

1-41-31-41-2

1.8 (1.0)1.8 (0.78)2.1 (1.1)1.2 (0.45)

37 (36)36 (23)55 (27)15 (11)

44364075

TABLE 6. Overlap in plant (genera) composition in diet of Black-and-yellow Silky-flycatchers (above diag-onal) and fleshy-fruiting plants (below diagonal) between mountain ranges. Overlap values were calculatedusing Pianka’s index for niche overlap (Pianka 1973).

Talamanca Central Volcanic Tilarán GuanacasteTalamancaCentral VolcanicTilaránGuanacaste

0.810.760.61

0.56

0.780.68

0.250.11

0.66

0.040.000.28

132


forests respectively, within 2 m above theground. Considering the scarce informationavailable on the reproduction of Black-and-yellow Silky-flycatchers, we evaluated itsreproductive activity using gonad size, testesvolume for males and largest follicle volumefor females; data were available for January,

March, April, May, June, July, September, andDecember. The data for all mountain rangesshow that breeding activity in males was con-centrated between March and July (maleswith testes volume of 200 mm3 were consid-ered reproductive) (Fig. 4). The timing ofreproductive activity in males was similar

FIG. 3. Average and standard error of genera per family of fleshy-fruiting plants, potentially eaten byBlack-and-yellow Silky-flycatchers in Talamanca, Central Volcanic, Tilarán, and Guanacaste mountains.

FIG. 4. Testes volume (mm3) of Black-and-yellow Silky-flycatchers for Talamanca, Central Volcanic,Tilarán, and Guanacaste mountain ranges in Costa Rica. Reproductive individuals are distributed abovedashed line.

133


throughout all cordilleras (from March toJuly; Fig. 4). Some outliers in July and Septem-ber may represent a second attempt at repro-duction, which is frequent in tropical birdswhen the first brood has failed due to preda-tion or storms (Skutch 1976, Janzen 1978,Oniki 1979).

Reproduction in females presents a simi-lar scenario (females with follicle volume of 7mm3 were considered reproductive). Likemales, the reproductive period in females var-ied little among mountain ranges (Fig. 5). Ingeneral, the peak of reproduction in Black-and-yellow Silky-flycatchers occurs betweenMarch and May (fledglings were observed in

June), a period which coincides with the onsetof the rainy season in the highlands (Stiles1983). Adjustment of reproduction with therainy season is a common pattern in tropicalbirds, especially in relatively seasonal environ-ments such as highland forests (Marchant1960, Stiles 1983). This fine-tuning betweenbreeding and rainfall may assure Black-and-yellow Silky-flycatchers a sufficient quantityof resources, particularly fruits, for raisingoffspring, or for independent fledglings(Fogden 1972, Wheelwright 1983, Stiles 1983,Loiselle & Blake 1991). Black-and-yellowSilky-flycatchers feed very young fledgelingswith fruits and nestlings probably are also fed

TABLE 7. Summary of a two-way analysis of variance (without repetition) comparing number of generaper family among mountain ranges.

Effect Variance df F PMountain ranges FamilyError

36.9414.131.63

347141

22.658.66

< 0.001< 0.001

FIG. 5. Volume (mm3) of largest folicule of Black-and-yellow Silky-flycatchers from Talamanca, CentralVolcanic, Tilarán, and Guanacaste mountains. Reproductive individuals are distributed above dashed line.

134


fruits by adults. Iinsect parts were not foundin stomach contents of reproductive femalesor males.

CONCLUSION

Black-and-yellow Silky-flycatchers arerestricted to mainly subalpine/elfin andupper montane forests from western Panamáto the Guanacaste mountain range in CostaRica. The abundance of this highland birdvaries throughout its geographical distribu-tion. Local abundance of Black-and-yellowSilky-flycatchers is apparently affected byarea available of potential habitat (subalpine/elfin and upper montane forests) and volca-nic activity. Particularly, intensity andfrequency of volcanic catastrophes have likelyhad a significant impact on the relative abun-dance of Black-and-yellow Silky-flycatchersin the Central Volcanic and Guanacastemountains, since abundance is low (or speciesis absent) on those massifs that have hadfrequent catastrophes. This highland bird isan obligate frugivore that reproducesbetween March and May, with little variationamong mountain ranges. The peak ofreproduction coincides with the onset of therainy season. This pattern of reproductiontiming may assure Black-and-yellow Silky-fly-catchers a sufficient quantity of fruits for rais-ing the offspring or for independentfledglings.

ACKNOWLEDGMENTS

We thank John Blake, Robert Ricklefs, A. C.Vallely, R. McNeil, and an anonymousreviewer for their valuable comments on themanuscript. Also we want to acknowledgeJulio Sánchez, Bernal Rodríguez, and IsidroChacón for field assistance. We received sup-port for this study from Parker-Gentry Trop-ical Research Fellowship (ICTE, University ofMissouri-St. Louis).

REFERENCES

Alvarado-Induni, G. E. 2000. Volacanes de CostaRica. Universidad Estatal a Distancia, San José,Costa Rica.

Barrantes, G., & J. E. Sánchez. 2000. A new sub-species of Black and Yellow Silky Flycatcher,Phainoptila melanoxantha, for Costa Rica. Bull.Br. Ornithol. Club 120: 40–46.

Diamond, J. M. 1975. Assembly of species com-munities. Pp. 342–444 in Cody M. L., & J. M.Diamond (eds.). Ecology and evolution ofcommunities. Harvard Univ. Press, Cambridge,Massachusetts.

Fjielså, J. 1985. Origin, evolution, and status of theavifauna of Andean wetlands. Ornithol.Monogr. 36: 85–112.

Fogden, M. P. L. 1972. The seasonality and popu-lation dynamics of equatorial forest birds inSarawak. Ibis 114: 307–343.

Frankie, G. W., H. G. Baker, & P. A. Opler. 1974.Comparative phenological studies on trees intropical wet and dry forests in the lowlands ofCosta Rica. J. Ecology 62: 881–919.

Gómez, L. D. 1986. Vegetación de Costa Rica.Apuntes para una biogeografía costarricense.Editorial Universidad de Costa Rica, San José,Costa Rica.

Haffer, J., & J. W. Fitzpatrick. 1985. Geographicvariation in some Amazonian forest birds.Ornithol. Monogr. 36: 147–168.

Hartshorn, G. S. 1983. Plants. Pp. 118–157 in Jan-zen D. H. (ed.). Costa Rican natural history.Univ. of Chicago Press, Chicago, Illinois.

Hernández-Baños, B. E., A. T. Paterson, A. G.Navarro-Sigüenza, & B. P. Escalante-Pliego.1995. Bird faunas of the humid montane for-ests of Mesoamerica: biogeographic patternsand priorities for conservation. Bird Conserv.Int. 5: 251–277.

Janzen, D. H. 1978. Predation intensity on eggs onthe ground in two Costa Rican forests. Am.Midl. Nat. 100: 467–470.

Kappele, M. 1996. Los bosques de roble (Quercus)de la Cordillera de Talamanca, Costa Rica.Biodiversidad, ecología, conservasión y desar-rollo. Grafisch Bredrijf and Looijen bv,Wageningen, Holanda.

Lawton, R. O. 1980. Wind and the ontogeny of

135


elfin stature in a Costa Rican lower montanerainforest. Ph.D. diss., Univ. of Chicago, Illi-nois.

Lawton, R. & V. Dryer. 1980. The vegetation ofthe Monterverde cloud forest reserve. Brenesia18:101–116.

Loiselle, B. A., & J. G. Blake. 1991. Temporal varia-tion in birds and fruits along an elevational gra-dient in Costa Rica. Ecology 72: 180–193.

Marchant, S. 1960. The breeding of some S. W.Ecuadorian birds. Ibis 102: 349–382.

Oniki, Y. 1979. Is nesting success of birds low inthe tropics? Biotropica 11: 60–69.

Opler, P. A. 1980. Plant reproductive characteris-tics during secondary succession in Neotropi-cal lowland forest ecosystems. Biotropica 12:40–46.

Pianka, E. R. 1973. The structure of lizard commu-nities. Annu. Rev. Ecol. Syst. 4: 53–74.

Remsen, J. V., Jr. 1985. Community organizationand ecology of birds of high elevation humidforest of the Bolivian Andes. Ornithol.Monogr. 36: 733–756.

Skutch, A. F. 1965. Life history of the Long-tailedSilky-flycatcher, with notes of related species.Auk 82: 375–426.

Skutch, A. F. 1976. Parent birds and their young.Univ. of Texas Press, Austin, Texas.

Slud, P. 1964. The birds of Costa Rica: distributionand ecology. Bull. Am. Mus. Nat. Hist. 128: 1–430.

Stiles, F. G. 1980. Evolutionary implications ofhabitat relations between permanent and win-ter resident landbirds in Costa Rica. Pp. 421–435 in Keast A., & E. Morton (eds.). Migrantbirds in the Neotropics: Ecology, behavior, dis-tribution, and conservation. Smithsonian Insti-tution Press, Washington, D. C.

Stiles, F. G. 1983. Birds. Pp. 502–530 in Janzen D.H. (ed.). Costa Rican natural history. Univ. ofChicago Press, Chicago, Illinois.

Stiles, F. G. 1985a. Seasonal pattern and coevolu-tion in the hummingbird-flower community ofa Costa Rican subtropical forest. Ornithol.Monogr. 36: 757–787.

Stiles, F. G. 1985b. Geographic variation in theFiery-throated Hummingbird, Panterpe insignis.Ornithol. Monogr. 36: 23–30.

Stiles, F. G., & A. F. Skutch. 1989. A guide to thebirds of Costa Rica. Cornell Univ. Press, Ithaca,New York.

Watson, D. M., & T. Peterson. 1999. Determinantsof diversity in naturally fragmented landscape:humid montane forest avifauna of Mesoamer-ica. Ecography 22: 582–589.

Wheelwright, N. T. 1983. Fruits and ecology ofResplendent Quetzals. Auk 100: 286–301.

Wolf, L. L. 1976. Birds of the Cerro de la Muerteregion, Costa Rica. Am. Mus. Novit. No. 2606.

Zar, J. H. 1984. Biostatistical analysis. Printice Hall,Englewood Cliffs, New Jersey.

136

reproduction, habitat use, and natural history...

Documents