COMPARATIVE LIFE HISTORY OF IN THE NORTHERN PERUVIAN AMAZON

Daniel M. Brooks1'4, Lucio Pando-Vasquez2 & Angel Ocmin-Petit25

'Department of Wildlife and Fishery Science, Texas ASM University, College Station, Texas 77843, USA.

2Explorama, Box 446, iquitos, Peru.

3Deceased.

Abstract. We investigate sociobiology of sympatnc coringas by comparing variation of common characters. The species studied (smallest to largest) are: loilopienni isiiliel/ae, Porphyroliietittipiirphymhtetaa, Cotlnga mayiiMiti, C. ftryaiia, Upaitgm mciferans, Phoeiiicirctts nigfifollis, Quemla fmrpurafa, Gjmnodoenis fbetidus and Ceplj/iloptertts ornatus. The relationship between size and sexual dimorphism was correlated, with mass and tail length being significant, and all other characters (total length, wing chord, sexual dich.romatism and ornamenta- tion) being n on-significant. For mass and tail length, smaller species ate characterized by females being larger than males, whereas females are smaller in the larger species. Although nor significant with all spe- cies along the size gradient, sexual dichromarism is more extreme in certain smaller species (i.e., Porpbym- laema and Cotiugff), and sexual ornamentation is present primarily in larger species. Dietary specialization increases with size, and most species exhibit low food resource defense intra- and interspecifically. Smaller species use higher parts of habitat structural attributes, whereas larger species use lower parts, Most of the smaller species are solitary, whereas larger species tend to travel in small flocks. Regarding courtship, smaller species are characterized by solitary male systems (including polygamy), with lekking in the medium species and/or monogamous courtship in the larger species. \Ve offer three hypotheses (modified from Alcock's model) as they relate to courtship strategy: 1) Solitary, dichromatic males of smaller species are attracted above the canopy because it highlights their iridescence and lures in females (cost — increased predation risk above the canopy, benefit = lower energy expenditure). 2) Actively courting males of medium species are attracted below the canopy to more aggregated fruit clumps; since bright coloring cannot be detected as well under the canopy, the males compensate through active courtship such as lek- king and/or vocalizing to lure in females (benefit = decreased predation risk below the canopy, cost = increased energy expenditure during courtship). 3) Males of larger species are lured to their courting sites by females that are attracted to the habitat containing the preferred resource; males typically court a single female using subtle ornamentation, as well as calling in some species. Accepted24 June 1999.

INTRODUCTION This family is characterized by extensive mor- phological and chromatic variation at the The family Cotingidae contains 25 genera and generic level. For example this family contains approximately 66 species (Stotz et til. 1996). the most variable size range of all , with the largest species weighing 80 times that of the smallest (Snow 1982). Although the primary factors (i.e., habitat separation and size assortment) driving cotinga community tSKilllKSCT.-IL

from museum specimens and other dimor- 20 m apart. Primary, interior rainforest is high phism characters (sexual dichromatism and in plant diversity, is characterized by a dark ornamentation) were obtained from both understory due to few penetrable light gaps, field and museum data. Other data including and contains tall trees that form part of the habitat association, sociality and courtship canopy, buttresses, or canopy emergents (e.g., were obtained from the field. Field data on Cedrela sp., Ceiba pentandra, P/fi/s insipida and feeding (limited observations) and species Inga sp.), often exceeding 35 m. Other tall with limited sample size (i.e., Phoenicians, trees in the region include palms (e.g., Euter- lodopleara and Porphyrolaetna) were supple- pes prectoria, \lauritia flexttosa, Scheeka sp., mented with museum specimen tag data or Socratea sp.), often occupying a gradient of obtained from available literature (e.g., Wal- habitats, from river edge to interior forest. lace 1849, Von Hagen 1937, Niceforo 1947, Sampling was done during 11 weeks from Sick 1951, 1954, 1979, 1993; Kills 1952, Pinto November 1993 to October 1998 at the end 1953, Snow 1961, 1971, 1982, 1985; Dia- of high water (March-May) and low water mond & Terborgh 1967, Haverschmidt 1968, (October-November) seasons to account for Reraut 1970, Howe 1982, Hilty & Brown seasonal variation, that did not appear to vary 1986, Terborgh 1986, Terborgh <:t al. 1990, significantly (Brooks 1998). Sampling meth- Trail & Donahue 1991, Davis 1993, Ridgely ods were similar to those used previously by & Tudor 1994,Thiollay 1994, Brooks 1998). others. Habitat associations were logged tor all visually or auditorily recorded encounters Field collected data. Field data were collected in of Cotingids at each site visited per trip. Data the Napo River intersect region (2 45'S; were collected by walking slowly along 72 55'W, see Fig. 1) of the Peruvian Amazon. transects with frequent stops (Pearson 1975) Habitats have been overviewed elsewhere and boating along waterways (Diamond & (e.g., Brooks 1997, 1998), but will be Terborgh 1967). Additionally stipra-canopy described here briefly. The habitats sampled observations were facilitated by using the l-'Kl. 1. A map (if the study region. 1 = Iquitos (major city), 2 = Amazon River bank site, 3 — Vana Mono are primarily situated along water, within for- canopy walkway at ACKKR (Vmazonian Tributary/Yana Mono Lodge. 4 = Napo River bank site. 5 = Sucusari Tributary/Camp, 6 = ACEER, est, or a transition/edge situation between Center for Environmental Education and 7 = Yarina Island/Rosario Island, 8 = Chimiguy Lake, 9 = l.oren/o Lake, [0 = L'rco Tributary, these habitats. Regions along the immediate Research). Species were identified using Hilty 11 = Casares Lake, 12 = ManarJ Tributary. edges of water (e.g., marshes, lakes, streams & Brown (1986) and Parker et al. (n.d.). and rivers) are often dominated by sawgrass structure in the northern Peruvian Amazon throated Cotinga Porpfyro/tiema porphyrolaeaia, or cane, followed by thick undergrowth fur- Museum collected data. Mensural data were col- have been identified (Brooks 1998), relatively Bare-necked Fruirerow Gjmnodoerus foetidtis, ther inland from the water. Island edge vege- lected at the Museum of Natural Science, little is known about the amount of variation Purple-tilr( »l ted Fruitcrow Otteraia purpiirata, tation is similar, unless the island is young, in Louisiana State University on 3 October in sociobiology, particularly for cotingas living and Amazonian Umbrella Cepbaloptsnts which case it will be monoculrured with rap- 1996. Mass (indicative of size) was recorded in the same region. urnatus, Although former classification (e.g.. idly colonizing plant species that are primary (in g) from the tags when available; standard Herein we examine life history correlates Snow 1973a) included tinras and becards in successional island specialists (e.g., Cecropia measurements taken in cm using a metal ruler of 9 species of sympatric cotingas in the this family, we follow current classification sp., GyneriiiiH sp., and Helicottia sp.}. Aquatic included total length, wing chord and tail northern Peruvian Amazon comparing varia- (e.g., Stotzf/rt/. 1996). edge vegetation is separated from interior length (indicative of size and maneuverability tion among common characters. The species rainforest by dense forest with thick under- in habitat). Measurements taken in cm with studied include Black-necked Red Cotinga METHODS growth, with canopy height often not exceed- standard dial calipers include maximum man- Phoeaiarcas nigricoilis, White-browed Purple- ing 10-15 m. The exception includes dibular height (indicative of masticator mus- tuft lodop/ei/rn Isabellas, Screaming Piha LJpaa- Data collection can be divided into data col- floodplain, which is characterized by continu- culature - bite force) and maximum ffts vociferaiis. Plum-throated Cotinga Cotinga lected from the field and data collected from ous short stems and grasses (e.g., Tridescantia mandibular width (indicative of gape-size ot uiciyntina. Spangled Cotinga ('.. cayana. Purple- the museum. Mensural data were obtained sp.) and tail, leafy trees with trunks spaced 5- food taken) (Snow 1973b).

195 An effort was made to measure at least throughout the study)], courtship strategy total length (r = 0.502, ns, N = 9), M/F wing 110267) had arachnid remains in its stomach. four specimens of each sex per species. In (type of courtship and number of males/dis- chord (r = 0.4.31, ns, N = 9), sexual dichro- The most general trend for feeding strat- cases where specimens were represented by play group) and competition at feeding trees. matism (r = -0.554, ns, N — 9) or ornamenta- egy was that smaller species tend to be gorg- less than four individuals of a given sex from Information on feeding strategy was tion (r = 0.730, ns, N = 9). ers, medium and larger sized species shift Dpto. Lore to, additional specimens measured largely supplemented with the literature, and For mass and tail length there is a trend of their diet seasonally depending upon whether included those that were closest to the sam- species were identified as gorgers (species females being relatively larger in the smaller fruits or prey are more abundant, and pling region, that sit in low bushes ot mistletoe and gorge), species towards the reverse (i.e., males being the largest two species enjoy fruits snch as seasonal resource specialists (species that sea- larger) in the larger species (Table 1). The palm fruits that grow along the river (Table ••Analytical methods. Morphological life history sonally concentrates on insects or fruit) and strongest exception to this rule is wing chord, 1). In sum there appears to be a general trend parameters included size and sexual dimor- riverine specialists (species preferring fruits which is relatively longer in females in all spe- towards increased dietary specialization with phism. Mass was used to measure size, rank- growing along river, such as palm fruits). Spe- cies except for C. cayana and Phnenidra/s. It is larger size. ing the species from smallest to largest in cies habitat associations were identified based interesting to note that C. cayana is the only range. Sexual dimorphism was measured by upon strata of tropical forest occupied (i.e., smaller-sized species that shows reverse Habitat association. The general trend is for obtaining male/female (M/F) ratios for mass, mid-upper, upper or canopy) or type of trends in dimorphism; total and tail lengths in smaller species to be associated with higher total length, tail length, and wing chord. Val- aquatic habitat occupied (i.e., varzea/water C, cayana are relatively longer in males than in parts of the forest and larger species to use ues < 1 indicated females were larger than edge transition or island varzea). females and wing chord is shorter in males. lower parts of the forest (Table 1). Smaller males; values >1 indicated males were larger Competition at feeding trees was exam- The case of Pboetiicirats is equally interesting species are forest canopy specialists and/or than females. Sexual dimorphism was also ined at two levels. The first involved different because all measured sexual dimorphism are often observed perched along the upper measured by examining degree of sexual scales of taxonomic orientation: whether con- ratios break the continuous trend towards part of a tree within thick riverine vegetation. dichromatism and ornamentation. Sexual specifics of the same or different sexes, other increasingly larger males in larger species. In contrast, the medium to larger species uti- dichromatism was defined as extreme colora- cotingas, or other avian species were present The most extreme measures of sexual lize the middle and/or upper strata of interior tion differences where the female is more in the same tree at the same time. The second dichromatism characterizes species at the foresr. The two largest species are water edge drab and was ranked as: 0 (none, for no dif- examined mtraspecific competition between smaller end of the spectrum (i.e., Cotiaga and specialists, with the largest (Cephalnpterns) ferences between male and female), 1 (slight, sexes more closely by plotting M/F ratios of Porplyro/aema), with the most diminutive spe- being an island specialist. for subtle differences, with both sexes rela- maximum mandibular height against maxi- cies (lodop/eitra) and some of the larger species tively drab), 2 (moderate, for subtle differ- mum mandibular width (Selander 1966). The showing slight or moderate sexual dichroma- Sociti/ity. Most of the smaller species are soli- ences with both sexes relatively colorful) and closer the value is to 1, the more likely the tism (Table 1). The two species exhibiting no tary whereas larger species tend to travel in 3 (extreme, for male relatively colorful and chance of the sexes competing for the same sexual dichromatism whatsoever (Ltpcwgi/s small flocks (Table 1). It is possible that female drab). Sexual ornamentation was food resources. and Cephnlopter/is) are the two species where group living evolved to enhance food finding defined as modified external morphology males use their voice to lure females, in larger, sub-canopy dwelling species versus adapted for displaying and was ranked as: 0 RESULTS Cotingids show relatively little sexual smaller, supra-canopy dwelling species. Since (none, for no differences), 1 (slight, for slight ornamentation, with only the smallest all species are frugivorous and fruit is spa- differences between male and female) and 2 Si^e. The plotted trend in size from smallest (ledopktira) and largest two species exhibiting tially non-ptedictable, it should be beneficial (moderate, for both sexes sharing similar to largest is as follows (with letter codes used slight to moderate ornamentation (Table 1). to have multiple individuals search for a fruit ornamentation, but male having stronger in tabularized data following each species par- Both of the largest species exhibit ornamen- patch that provides plenty of food for all ornamentation), Pearson product-moment enthetically): 1. isabellae (li), P. porphyrolaema tation in the head and throat region; Gymnodo- group members. Group living may benefit correlations were used to assess the relation- (Pp), C. maynana (Cm), C. cayana (Cc), L. votif- ems males have fleshy facial folds whereas riverine species such as Gjmnodoerus'm. finding ship between size [ranked from smallest (1) to erans (Lv), P. nigricollis (Pn), O. purpiirata (Qp), Ci'phaloptetw males have larger crests and an preferred resources such as palm fruits. This largest (9)| and each sexual dimorphism char- G.foetidus (Gf) and C ornatits (Co). external air sac projecting from the throat. may also explain subtle differences in breed- acter, using the computer program SPSS ing strategy for the cooperative breeder, (1996). Sexual dimorphism. Sexual dimorphism closely Feeding strategy. All species are primary trugi- Oiientla (Snow 1971), In contrast solitary, Other life history parameters include matches the range of sizes. Significant corre- vores, consuming small invertebrates such as supra-canopy dwelling indi\s can per- feeding strategy, habitat association, sociality lations occur between size and M/F mass (r — insects to a lesser extent. Although all litera- haps locate fruit easier because they can see [mean flock size and relative abundance (1,976, P < 0,001, N = 9) and M/F tail length ture reports Phoenicians to be a strict frugi- farther distances unobstructed by forest (expressed by numbers of encounters/species (r = 11.728, P < 0.05, N = 9), but not for M/V vore, at least one specimen (LSL'MNS- growth. It should be more profirable for

196 197 Parameters Species1'

Cm Pn Gf Mean mass (g) 18 49 70 111 284 571 Sexual dimorphism

Male/lHemalc mass M/F total length M/F tail length M/F wing chord

Sexual dimorphism1

Sexual ornamentation'1 Feeding'

Habitat1 Soliality Mean flock size 1.0 1.5 2.9 2.4 1.5 2.1 Relative abundance 2 29 44 29 31 35 Courtship strategy Sexual dichomatism and sexual ornamentation For all species, see values above

Number of males/display group 1 1 Several'" Several"1 1 /fern. I/fern"

Solitary male or lek sol. lek1" Visual or vocal visual visual vocal Competition .it feeding trees' More than 1 sex present yes yes yes

Other cotingas present1

I'XBI.r. 1 Continuation.

Parameteri;

Pp Pn _k Other species present s,p mk,t,f k,m I'bod compel ion between sexes M/F maximum gape M/F mandible height

"Data not available. 'Sexual dichromarism: 0 — none, 1 = slight, 2 =• moderate, 3 - extreme. ''Sexual ornamentation: 0 = none, 1 = slight, 2 = moderate. 'I'eeding strategy: g -" gorger, s = seasonally resource specialist, r = riverine specialist. I labitat association: m ~ mid-upper strata, u = upper strata, c = canopy, v = varzca/water edge, t = island varzca. "A male may displace another male from a preferred calling site. ''Leks are often loosely associated. 'Quantified by whether other individuals were ever observed at tree or not. Lack of data should be interpreted as species possibly shared resource patches with other individuals, but it went undetected. 'c = (.. fuyaiiu, m = C. mayitiiiia, q = O. piirpurala, g = C. foelietm.

Incident invoking a female C. ciiyaiia and L typopyrrha simultaneously mobbing a female (iymnodoeras in primary rainforest canopy, at;rpical habitat for Gy/niiodoeras. The Gymnodoerus flew off shortly after being mobbed and was probably simply moving through the atypical habitai. k = Plumbeous Kites (itiit/iap/umbea); s = Short-tailed Parrot (Gmydidascalus brnchyunts); p = Cobalt-winged Parakeet (Bro/cgf/w cyaaopfera); m = Cinereous Mourner (Lamoarca bypnpyrrha}; f = Crowned-slary (Y^apidimumus attraiilioatroaistatus) and Dusky-chested Flycatchers (Tyraiinopsis lutciventris); c = Giant Cow-bird (Scaphiditra oryyvora); t = Blue-gray 1 anager (I'braupis efuscopiis). '"Data supplemented or obtained with one or more of the following: Wallace 1849, Von Hagen 1937, Niceforo 1947, Sick 1951, 1954, 1979, 1993; h'.llis 1952, Pin 10 1953, Snow- 1961, 1971, 1W2, 1985; Diamond & Terborgh 1967, I laverschmidt 1968, Beraut 1970, I iowe 1982, Hilty & Brown 1986, Tcr- borgh 1986, Terborgh eta/. 1990, Trail & Donahue 1991, Davis 1993, Ridgelv & Tudor 1994, Thiollay 1994, Brooks 1998. smaller species co forage alone if fruit clumps coloring cannot be detected as well under the length and wing chord) show a significant are "thinner" above than below the canopy. canopy the males compensate through active pattern ot females being larger than males in There is an interesting relationship Courtship strategy. Smaller, supra-canopy dwell- courtship such as lekking and/or vocalising smaller species towards males being larger between sexual dichromatism and group size. ing species tend to be more colorful (Table 1). to lure in females. The benefit is decreased than females in larger species (see VX'ebster The species with extreme or moderate sexual The males of primarily larger species (that predation risk below the canopy but the cost 1992). C. cayana and Pboenicimts are the two dichromatism tend to be solitary [Table 1). show the least amount or no sexual dichro- is increased energy expenditure during court- species that show exception to this rule. P/ioe- The brighter coloration of males of the matism) attract females through vocalizations ship. 3) Males of larger species are lured to nidrciis may be closer related to manakins smaller, solitary species serves to attract and/or subtle ornamentation (Table 1). Such their courting sites by females that are (Pipridae) than other cotingas, perhaps serv- females. However this does not mean that vocalizations are often more intensified in attracted to the habitat containing the pre- ing as the phylogenetic link between the two females of these solitary species are territorial dwelling in interior forest (Morton ferred resource. The male typically courts a groups (Trail & Donahue 1991). for mating rights. For example two female C. 1975) as seen in the case of Upturns for single female using subtle ornamentation, as The benefit of females generally being mayttana sharing the same tree showed no ago- example. The common underlying pattern well as calling in some species. larger in smaller species may be linked to nistic interactions despite .being courted by here is that males utilize different adaptations parental investment (see Payne 1984). For the same male. Females of the more social to lure females to their courting site. Competition at feeding trees. Data for seven of example, larger egg and offspring size would species are lured to a display site by male The number of males displaying to the nine species indicate that more than one have a reduced chance of predation and activity such as lekking and/or vocalizations, females can he divided into three general pat- sex, species of cotinga or species of bird may therefore favor mothers of smaller species or are present at the site as part of an inte- terns (Table 1). I) Smaller species tend to be present at the same feeding tree (Table 1), being larger (see Lack 1968, Andersson grated social unit (e.g., Otientla}. have only one male courting at a time, even if However, with the possible exception of 1994). Alternatively, smaller females in fluctu- Despite species packing rules (i.e., an area polygamy is involved. In one aforementioned Cepbaloptems, there appears to be a relatively ating environments may be able to breed at of the same size will contain more smaller case a single male C. mayaatia displayed to two high amount of feeding morphology overlap an earlier age than larger females, therefore species and fewer larger species; see G>t- females about 25 m away in an adjacent tree. between sexes (Fig. 3). Nonetheless, several breeding at a younger age would favor smaller greave 1993) the trend of relative abundance 2) Medium sixed species (Ufmt/giis and Phoe- observations suggest relatively low food females of the larger species (see Downer- from smallest to largest species is somewhat tiicirciii) have communal display sites where resource defense in these species. hoxver 1976). Finally, males of larger species reversed (Table 1). This can perhaps be several males will display at once. These dis- For example, the same sexes of a given (e.g., Cephalopterus) may benefit through explained by variation in species biology. For play arenas often span areas of 2500 - 40,000 species may share feeding sites for at least enhanced defense of resources such as court- example, Porpbyrolaema, a smaller species, is m2. Finally, 3) the larger species (e.g., Qi/erti/a some species. VX'e observed three male Cotinga ing sites (see Andersson 1994). naturally rare and is encountered with far less and Cepbaloptems) are quite variable depending maynana foraging together on one occasion. The pattern of size dimorphism (mass frequency in the field than most of the other upon the situation, but a single male tends to On other occasions we observed groups of and tail length) increasing with larger males is species with the possible exceptions of court one female. In sum we see a trend from two male C.ntinga cayana foraging together concordant with Snow's hypothesis (1982). Ifidopleitra and Phoenicians. Therefore the low solitary male systems including polygamy in with or without females present. Although not significantly correlated with abundance (2 individuals) can be explained by the smallest species, to lekking in the medium Moreover, during more than 10 weeks of size, smaller species such as Porpisyralaema and naturally low numbers of this species in species and/or monogamous courtship in the field time we only observed a single direct Cotinga show increased sexual dichromatism nature. In contrast Ijpmtgits is a loosely-asso- larger species (Table I). agonistic interaction at a feeding tree. This whereas lodopkura and the largest species ciated lekking species. A solitary individual Figure 2 depicts three courtship models was an incident involving a female Cotinga cay- show slight or moderate ornamentation. was only encountered on one occasion; on all (modified from Alcock's model 1984), each ana and Lrtniocerca hypopjrrha simultaneously Thus what certain species lack in sexual other occasions they were in groups. Thus the describing hypotheses built upon life history mobbing a female Gymnodeertis in primary dichromatism they may partly compensate higher abundance of Upai/gi/s is perhaps due patterns described herein: I) Solitary, dichro- rainforest canopy, which is atypical habitat for with slight or moderate sexual ornamen- to social courtship constraints (i.e., where matic males of smaller species are attracted for Gymnodoems. The Gynmodoents flew off tation. Non-dichromatic species also com- there is one, there are many). For example, on above the canopy because it highlights their shortly after being mobbed and was probably pensate through intense vocalizations one occasion we observed a group of ~Ljpaitgn$ iridescence and lures in females. The cost is simply moving through atypical habitat. produced by the males to lure in the females. in a loosely associated lek, when suddenly increased predation risk above the canopy but It is important to note that different they formed a very tight lek, calling more rap- the benefit is lower energy expenditure. 2) DISCUSSION results might be obtained at other study sites, idly and louder. It is possible that a female Actively courting males of medium species even with slightly different communities (M. entered the arena at that time, though we are attracted below the canopy where fruit Thery pers. comm.). For example, of 11 for- were unable to verify this since the sexes are clumps may be more aggregated. Since bright est-dwelling cotingas in French tluiana, at BROOKS £TAL

f Model for smaller-sized species (Porphyrolaema and Cotinga) \d A. Smaller species Larger species

l^predation risk,)

tr electiv umers Sunlight attracts males of Males perch supra .ures Upper parts ol vegetation Sub-canopyfRi™ edge above d) species with high Cl) -canopy to show d) n Solitary Small (locks canopy sexual dichromatism irnctescenl colors :emales h FIG. 4. General lite history treads in nine species 0.4 of cotingas. /lower energ^, 0.6 i ^expenditure^ J M/F maximum gape reveal a different picture. Of those species in FIG. 3. Food overlap between sexes. the Cotinga clade that C. maynana is closest related to (e.g., C. amabilis, C.. ridgwayi and C. ^ -N. Model for medium-sized species ( and Querula) females of the smaller, solitary species appear nattererii), C. maynana resembles C. cayana the to be lured to the brighter coloration of males least because it has a yellow iris and predomi- .''Higher energy) (Fig. 4). Whereas females of the more social nantly blue wings rather than dark iris and V£xpenditur>/ species are lured to display sites by male wings like C. cayana and the other species. tr activity (e.g., lekking, vocalizations, etc.) (Fig. Despite lack of competition for food Increased , attracts males of Active male courtshi( Lures 2). resources, there is a relatively low amount of resources W species with low constrained by 3 in feeding morphology overlap between sexes sub-canopy sexual dichromatism * drab colors females Lack of competition Jor food resources. There (Fig. 3). Perhaps this merely suggests that appears to be relatively little competition for males and females of most species only utilize similar food resources. Moreover fruit grows the same food resources, rather than compete as a clumped resource that varies seasonally for them (see Glutton-Brock & Harvey 1977). in spatial distribution. It is not profitable for We observed different sexes of Cotinga associ- primary frugivores to defend a resource that ated at the same resource patch on more than may not occur with any predictability within one occasion. Model for larger-sized species (Gymnodoerus and Cephalopterus) that territory (see Brown 1964, Snow 1985). The fact that food resources are not Ala/e reproductive strategy. Fruit as the main Preferred attracts Lures in males of Males court females defended is strengthened even more by look- food resource appears to play an important foods along 3 females to 3 species with low 3 with subtle the river river habitat sexual dichromaiism ornamentation and ing at the lack of social mimicry between two role in the lives of most species. Fruit is calling in some cases congeners: C. cayana and C. maynana. The always present temporally but often unpre- benefits of social mimicry include: 1) escap- dictable spatially. A lack of territorially is ing attack from larger models, and 2) deriving beneficial in areas where food resources are higher status at a resource site - smaller spe- spatially unpredictable because the ftuit may cies may be deterred from occupying the not blossom within the defended area. The lease three larger species (Haematndmts milita- enjoving items such as palm fruits. The two same site where the seemingly larger model main benefit of a clumped, temporally pre- ris, Procnias alba and Ri//}U'ola nipicola} show largest species were both water edge special- (that is really the mimic) is present (Diamond dictable resource is decreased foraging time strong sexual dichromatism buc hccle size ists in terms of habitat preference. This is 1982). For social mimicry to be achieved, two (i.e., take/search time) that permits a male to dimorphism (Snow 1982), contrary to the probably a consequence of their preference more distantly related species must be more mate with more than one female, and allows findings observed herein. for riverine fruits (Sick 1993). Thus it is plau- chromatically similar to one another than the female to raise the offspring alone (Snow sible, at least for these larger species, that dis- they are to other more closely related species. 1985). The lack of territorially between tribution within a landscape is dictated by Although C. maynana and C. cayana look simi- males combined with the potential to mate their preferred foods (Fig. 4). In contrast lar from a distance, specimens in the hand with more than one female results in intense . Volume 2: The suboscine pas- sexual competition and the evolution of behavior may explain the increased male sex- Diamond,]. M., & J. W Terborgh. 1967. Observa- serines, Univ. Texas Press, Austin, Texas. extreme courtship displays (Snow 1985), This ual dimorphism in such larger species (see tions on bird distribution and feeding assem- Selander, R. K. 1966. Sexual dimorphism and dif- Andersson 1994). pattern is especially apparent in rhe smaller blages along the Rio Callaria, Departamento of ferential niche utilization in birds. Condor 68: species that show higher sexual dichromatism Lore to, Peru. Wilson Bull. 79: 273-282. 113-151. (see Bradbury & Vehrencamp 1998). ACKNOWLEDGMENTS Diamond,}. M. 1982. Mimicry of friarbirds by ori- Sick, H. 1951. An egg of the umbrellabird. Wilson Bradbury & Vehrencamp (1998) have oles. Auk 99: 187-196. Bull. 63: 338-339, overviewed the pattern of more brightly col- We thank Marc Then; David Snow and Ken Downhower, |. F. 1976. Darwin's finch and the Sick, H. 1954. Ziir Biologic des amazomschen ored species living above the canopy to Risenhoover for helpful comments on this evolution of sexual dimorphism in body size. Schirm vogels, CtphaSafteras ornatus J. Orn. 95: opportune sunlight in order to advertise their ins. We are indebted to CONRPAC and Nature 263: 558-56.3. 233-244, Hberhardt, L. S. 1994. Oxygen consumption dur- Sick, H. 1979. Ziir l\istweise der Contingiden iridescent coloration to females. Moreover, ir INKL'.N'A for allowing us to work in the ing singing by male Carolina Wrens (Tlnyotliuriis lodopkura and X.ipliolena. J. Orn. 120: 73—77. is more difficult for smaller species to adver- region. Logistical support was provided by hidomdamts}. Auk 111: 124-130. Sick, H. 1993. Birds in : a natural history. tise auditorily (see Bradbury & Vehrencamp Bxpl orations, Inc. and Peter |enson of F'^llis, H. R. 1952. Nesting behavior of u purple- Princeton Univ. Press, Princeton, New Jersey. 1998), which may explain the bright iridescent F.xplorama. throated fruit-crow. Wilson Bull. 64: 98-100. Snow, B. K. 1961. Notes on the behavior of three coloration in species such as Cotutga. Alcock Kndler, J. A. 1991. Variation in the appearance of Coringidae. Auk 78: 150-161. (1984) predicts that species advertising with REFERENCES guppy color patterns to guppies and their pred- Snow, D. W 1971. Observations on the Purple- visual cues spend a low amount of energy ators under different visual conditions. Vision throated fruitcrow in Guyana. Living Bird 10: doing so. The trade-off is that such bright col- Res. 31: 587-608. 5-17. Alcock, \ 1984. Animal behaviour: an evolutionary ors may also advertise the males presence to Haverschmidt, F. 1968. Birds of Surinam. Oliver & approach. Sinauer, Sunderland, Massachusetts. Snow, D. W. 1973a, The classification of the Cot- Boyd, London. predators (I7,ndler 1991). In contrast, the pat- Andersson, M. B. 1994. Sexual selection. Princeton ingidae (Aves). Breviora 409: 1-27. tern of more drab species living below the Univ. Press, Princeton, Ne\ Jersey. Kilty, S. !,., & W. L Brown. 1986. A guide to the Snow, D. W. I97.ib. Distribution, ecology and evo- birds of Colombia. Princeton Univ. Press, Prin- canopy perhaps favors decreased predation Berauc, E. 1970. The nesting of Gywnockras f'ottidits. lution of the bellbirds (Proa/ins, Cotingidae). ceton, New jersey, (Bradbury & Vehrencamp 1998). However ibis 112:256. Bull. Br. Mus. Nat. Hisr. 25; 369-391. Howe, H. F. 1982. Fruit production and animal Snow, D. W 1982. The cotingas. Cornell Univ. some of these species with decreased sexual Bradbury, J. W, & S. 1.. Vehrencamp. 19%. Princi- activity in two tropical trees. Pp. 189-199 /// Press, Ithaca, New York. dichromatism (e.g., Upaiigas) advertise to ples of animal communication. Sinauer, Sun- Leigh, F.G., Rand, A.S., & DM. Windsor Snow, D. W 1985. The web of adaptation: Bird females using auditory cues that Alcock derland, Massachusetts. (eds.). The ecology of a tropical forest. Smith- studies in the American tropics. Quadrangle, (1984) predicts are energetically high, as con- Brooks, D. M. 199". jSon la competencia, <_•! tamario ) la superposicioo de dietas pronostica- sonian Inst. Press, Washington, DC, Times Book Co., New York. firmed in multiple subsequent studies (e.g., dores de la composicion de Ramphasiidae? Pp. Lack, D. 1968. Kcological adaptations for breeding SPSS. 1996. SPSS 7,5 for Windows. SPSS Inc., Rberhardt 1994). Nonetheless, auditory sig- in birds. Methuen, London. 283-288 in Fang, T. G., R. K. Bndiner, R. Chicago. nals are less constrained by environmental Aquino, & M. VakjLii (eds.). Manejo de fauna Morton, [•',, S. 1975. Ficological sources of selec- Stotz, D. I-'., J. W: Hizpairick, T. A. Parker III, & D. factors (e.g., filtering by sub-canopy vegeta- silvestre en la Amazonia. OFAVIM, La Pa*. tion on avian sounds. Amer. Nat. 109: 17-34. K. Moskovits. 1996. Neotropical birds: ecology Niceforo, M. H. 1947. Noras sobrc aves de tion) than visual signals (Bradbury & Vehren- Brooks, D. M. 1998. Competition and coexistence and conservation. Univ. Chicago Press, Chi- camp 1998). It appears that the courting site in Neotropical birds: a latitudinal comparison. Colombia II. Caldasia 4: 317-377. cago. of a male(s) is established to attract a Ph.D. diss., Texas A&M Univ., College Station, Parker, T. A., S. W. Cardiff, & T. S. Schulenberg. Terborgh, J. W 1986. Population densiritcs of n.d. Checklist of the birds of the Iquitos area. Amazonian birds; implications for conserva- temale(s) in the smaller and mediiim-st^ed Texas. Unpublished. tion. Ibis 128: 165-166. species (Fig. 2). In this case the females pat- Brown, J. 1.. 1964. The evolution of diversity in Payne, R. B. 1984. Sexual selecrion, lek and arena Terborgh. J. W., S. K. Robinson, T. \ Parker II I, C. tern of spatial distribution may be a conse- avian territorial systems. Wilson Bull. 76: 160— behavior, and sexual size dimorphism in birds. A. Munn, & N. Picrpont. 1990. Structure and quence of male distribution rather than 169. Glutton-Brock, T. H., & P. H. Harvey. 1977. Pri- Ornithol. Monogr. 33: 1-53. organization of an Amazonian forest bird presence of food resources. mate ecology and social organization. [. Z

205 BROOKS ETAl.

Von Hagen, W 1937. On the capture of the 207. umbrella bird (Cepbalopterus penduliger Sclater). Webster, M. S. 1992. Sexual dimorphism, mating Proc. Zool. Soc. London 27-30, system and body size in New Wodd blackbirds Wallace, A. R. 1849. On the umbrellabird (Cspha- (Icterinae). Evolution 46: 1621-1641. lopierus ornatiis). Proc. Zool. Soc. London. 206-

206