The Condor96722-133 0 The Cooper Ornithological Society 1994

THE SPATIAL AND TEMPORAL DISTRIBUTION OF BREEDING FEMALE MONTEZUMA : EFFECTS ON MALE MATING STRATEGIES

MICHAEL S. WEBSTER~ Section of Neurobiology and Behavior, Cornell University, Seeley Mudd Hall, Ithaca, NY 14853

Abstract. In a Costa Rican population of Montezuma Oropendolas (Psarocolius mon- tezuma), females nested in the dry season(January through May) and males did not help build nests or behave parentally. The length of the nesting period, from nest initiation to fledging, was relatively long (ca. 2.5 months). Approximately one-third of all nests fledged young. Between 19 and 32% of nests fell out of colony trees before fledging, and most of thesefell during periods of heavy rain. Females bred colonially, with up to 130 nests(median = 24) in a single tree. In addition, nests were clustered within colony trees: several nests were often built togetheron a singlelimb. This behavior appearsto be costly, asthe clustering of nests caused limbs to break, contributing to the high rate of nest loss. The degree of nestingsynchrony was low among nest clusterswithin the colony, but relatively high within clusters.The overall degreeof breeding synchronywithin an oropendolacolony (SD of nest- lining date = 13-25 days) was lower than reported for most monogamous colonial , but similar to that reported for other polygynouscolonial .Breeding synchrony within but not among nesting clustersappears to increasethe ability of high-ranking male oropen- dolas to defend or court multiple females. Key words: Breeding synchrony; colonial nesting; female-defense polygyny; Montezuma ; Psarocoliusmontezuma.

INTRODUCTION lows), yet polygyny has evolved in very few (Lack Male mating strategies in birds are generally 1968). There are at least two hypothesized rea- thought to be determined, at least in part, by the sonsfor why this might be. First, if male parental spatial distribution of breeding females (Emlen care is essential to the survival of nestlings in and Oring 1977, Wittenberger 1980a, Davies colonial birds, such paternal care will limit a 1991). When sexually receptive females are male’s ability to attract or defend multiple mates clumped in space, such that they are defensible (Lack 1968, Emlen and Oring 1977, Davies 199 1). by a single male, males are expected to actively Although this is likely to be an important factor compete for and defend female groups, leading in some colonial species, particularly colonial to a polygynous . For the Amer- seabirds that feed on widely dispersed and un- ican blackbirds (Icterinae), comparative studies predictable resources,some recent studies have have demonstrated an association between po- suggestedthat the importance of male parental lygyny and the spatial dispersion of nesting fe- care is not the only factor determining male mat- males (Robinson 1986, Webster 1992), and in- ing strategiesin (Webster 199 1a, Dunn traspecificstudies have documented competition and Robertson 1992). among males of polygynous species for groups Alternatively, interspecific differencesin male of females or prime nesting sites (e.g., Lenington mating strategy might reflect differences in the 1980; Wittenberger 1980b; Robinson 1985a; Post opportunity to obtain or defend multiple mates 1992; Webster, in press). (Emlen and Oring 1977, Westneat et al. 1990, Factors other than the spatial distribution of Davies 1991, Webster 1991a). For example, if females must affect male mating strategies,how- females are highly synchronous, and males re- ever: many speciesof birds nest colonially (e.g., quire some time to court or guard each mate, many seabirds, shorebirds, herons and swal- then polygynous mating will not be possibleeven if females nest in denseaggregations and paternal care is unimportant. This hypothesispredicts that male mating behavior will be associated with 1Received 20 October 1993.Accepted 7 March 1994. patterns of female breeding synchrony within the * Present address:Department of Ecology and Evo- lution, University ofchicago, 1101 E. 57th Street,Chi- colony, and that female breeding synchrony will cago, IL 60637 be lower for colonial speciesin which males de-

I7221 FEMALE OROPENDOLA BREEDING BIOLOGY 723 fend multiple mates than for thosein which males place each dry seasonfrom 1986 through 1990, pair with a single female. for a total of 18 months. The Montezuma Oropendola (Psarocolius Birds were captured using elevated mist nets montezuma) is a colonial-nesting, neotropical (Mease and Mease 1980) placed near nesting col- blackbird (Icterinae, Emberizidae). Males of this onies and at foraging sites baited with . speciescompete for and defend groupsof females Captured birds were individually marked with at nesting colonies, and copulation successis leg bands made from colored PVC plastic. Sex strongly biased toward thosefew males most suc- was determined from body size, and male age cessful in this competition (Webster, in press). class (adult vs. juvenile) was determined using This male mating strategy is quite rare among the size of facial wattles (males with wattles less birds (Oring 1982, Post 1992) and has appar- than 8 mm in depth were considered to be ju- ently lead to the evolution of alternative male veniles, Webster 199 1b). mating strategiesand an extreme level of sexual For the purposesof this study, I defined a col- size dimorphism (Webster 199 1b, 1992). ony as a group of nests built in the same tree or Although nests and colonies have been de- adjacent trees during a breeding season (Fraga scribed for this (Skutch 1954, Fraga 1989) and 1989). In the majority of cases,trees containing other species of oropendola (Chapman 1929, nests were separatedby one or more kilometers, Schafer 1957, Tashian 1957, Drury 1962, Smith making colony delineation unambiguous (seebe- 1983) no published study has followed marked low). female oropendolas over the course of an entire Several assistantsand I observed oropendolas breeding season. In this paper I describe the at a focal nesting colony (“La Selva Colony”) breeding biology of female Montezuma Oropen- each season,for a total of more than 1,200 hr of dolas, including nesting phenology, estimates of observation. We conducted colony observations nesting success,and movements of marked in- from around dawn until late morning (ca. 10:30), dividuals between nesting attempts. In particu- and from mid-afternoon (ca. 14:30) until dusk. lar, I describe the spatial and temporal distri- Observations through the entire day indicated bution of sexually receptive females and the that very little activity occursat the colonies dur- factors likely to affect this distribution. I then ing the hot midday hours (Webster 1994). During compare the breeding synchrony of Montezuma colony observations, we noted individual arrival Oropendolas to that of other colonial birds, and and departure times, aggressiveinteractions and discuss the effects of coloniality and female attempted copulations. During the 1987, 1988 breeding synchrony on the evolution of male and 1989 seasons,we also recorded the nest to mating strategies. which an arriving female went and, if possible, whether or not she carried anything in her bill (e.g., nesting material, leaves, arthropod prey). MATERIALS AND METHODS These data were used to determine female nest- ing stage and sexual receptivity in those three Montezuma Oropendolas inhabit lowland for- years (because of the relative inaccessibility of ests ranging from southern to the Canal nests,it was impossible to examine nest contents Zone in (Skutch 1954). This speciesis directly). For most females, detailed records of highly sexually dimophoric in body size, with nesting activity were not made during the early adult males (mean = 52 1 g) weighing more than construction stageswhen the nest was difficult to twice as much as females (246 g, Webster 199 1b). discern. We monitored female nesting activity in There are no differences in plumage coloration 1986 as well, but did not keep detailed records betweenthe sexesor betweenjuveniles and adults. of daily activity in that year. In addition to the I studied a population of Montezuma Oro- focal observation colony, two other oropendola pendolas at the Estacibn Biologia La Selva near colonies in the La Selva area (“Crocodile Point” Puerto Viejo de Siripiqui, (10”25’N, and “Colony B”) also were observed, but on an 84”O1 ‘W). La Selva is a lowland rainforest site, irregular basis. All known colonies in the La Sel- with annual rainfall averaging 3.8 m/year va area were visited at the end of each breeding (McDade et al. 1993). A pronounced dry season seasonto determine colony size (total number of occursfrom early January into May, during which nests built during the course of the season). Montezuma oropendolas breed. My studiestook At the end of the 1989 field season, we at- 124 MICHAEL S. WEBSTER

tempted to find all oropendola colonies in an becausenest initiation was not observedfor most approximately 5 5 km2 area around La Selva. Be- nests. Females brought dead leaves, which they cause of their large size, open aspect, and the used to line the inside of the nest, for an average loudness of male displays, colonies were rela- of 5.8 +- 2.44 days (n = 35 nests). Incubation, tively easy to locate, and we believe we found defined as the period between last observations all colonies in the area. At each colony in the of leaf-lining and the first observed trips with study area, plus a small number incidentally en- food, lasted for 15.5 f 1.9 days (n = 14 nests). countered outside of it, we counted the number It was difficult to determine the length of time of nests at or near completion. Becausethis sur- that females fed nestlings, because females who vey was conducted relatively late in the breeding stopped bringing food to the nest may have season,this count gives an approximation of the fledged or lost their young. Skutch (1954) re- total number of nests that were built in each ported that feeding lasts for 30 or more days. In colony over the entire season(old nests fall out this study, one female known to have fledged of the colony tree between breeding seasons).We young brought foot to her nest for 32 days, and also recorded the spatial distribution of nests six other females thought to have fledged young within each colony by counting the number of brought food for 2942 days. I considered a fe- nestsbuilt togetherin a single cluster. Most nests male to have fledged young if she brought food in clusters were built on the same branch and to the nest for at least 29 days. A wide variety within a few centimeters of each other (see be- of food items were brought to the young, includ- low), though some were a meter or more away. ing large arthropods (mostly spidersand orthop- We arbitrarily defined two nests as being in the terans), prey (frogs and lizards), and same cluster when they were estimated to be some fruits. within 1.5 m of each other. Finally, we collected Work with closelyrelated species(Schafer 1957, fallen nests from beneath colonies for nest de- Robinson 1985a, Post 1992) suggeststhat a fe- scriptions. Unless otherwise indicated, all mea- male is sexually receptive at the end of nest con- sures are given below as mean f standard de- struction, when she is lining her nest with leaves. viation. Of 30 copulations observedduring this study that involved an identifiable female, 23 (77%) in- RESULTS volved females in the nest-lining stage,and five (17%) involved females within two days of nest- GENERAL BREEDING BIOLOGY lining. Therefore, a female was considered to be Nests. Female Montezuma Oropendolas built sexually-receptive during the period she was ob- long, pouch-like nests with no assistancefrom served bringing leaves to her nest. males. The nests were woven from a variety of Nestingsuccess.The clutch size for this species fibrous plant materials, including strips of palm has been reported to be two eggs(Skutch 1954), frond and epiphytal roots. The bottom of nests and in this study four of five neststhat fell during were lined with a thick layer of dry shreddedleaf incubation and were recovered on the day they material, possibly to cushion and insulate the fell contained two eggs(the fifth contained a sin- eggsand nestlings. Fallen nests(n = 28) averaged gle egg). However, females rarely fledged more 111.3 f 26.5 cm in length and 68.3 + 5.7 cm than a single young; of 22 nests that fell during in maximum circumference of the pouch. These the feeding stageor were accessedfor blood sam- nests weighed 4 14.4 f 101.4 g when dry, but ples (Webster 199 1b), all contained a single nest- were fully two times heavier (874.1 f 239.3 g) ling. after we soaked them in water. Of 45 nests started at the focal colony and not Nesting phenology and female receptivity. As abandoned within a day or two (31 in 1987 and with nest-building, females incubated and fed the 14 in 1988) 20% never reached the nest-lining young alone with no assistance.The frequency stage,36% were lined but never had food brought of a female’s arrivals and departures from her (eggsapparently lost or never laid), and 16% ap- nest and the identity of items she carried varied parently hatched but did not fledge any young acrossthe nesting cycle (Fig. 1). I used these ob- (female brought food to the nest for fewer than servations to determine each female’s stage in 29 days). Only 29% of these 45 nests (or 36% of the nesting cycle (see also Tashian 1957). Nest the nests were that were lined) were considered construction lasted 2-3 weeks, but the exact du- to have fledged any young (female brought food ration of this stage was difficult to determine to the nest for at least 29 days). FEMALE OROPENDOLA BREEDING BIOLOGY 725

8 Nest Material 1 Leaves 1 Nothing IFood

Nest # 1

Nest Material “1 6 6 /& 4- i% ;f: 2,

$ & O-

Day Number

FIGURE 1. Female activity at two nests in the focal colony (1989). Bars at top show items the female was observed carrying in her bill on most visits during that period. Day 0 correspondsto January 15 for both nests.

Many oropendola nests fell out of colony trees holding all eight snapped under their combined during this study, and this appearedto be a major mass and fell. Similarly, six nests fell, one at a sourceof nestling mortality: seven of the 36 nests time, from a focal colony during 11 days of un- that were lined (19%) fell during incubation or seasonable rains in March 1990 (all but one of the feeding of the nestlings. If data from the two the other 16 females building nestsat that colony additional colonies that were studied less inten- abandoned their nestsduring this period of heavy sively also are included, 22 of 68 nests (32%) rains). were lost by falling. Some of these nests fell be- causethe limb to which they were attachedbroke, while others tore free. Severe weather appeared COLONY DESCRIPTIONS AND THE SPATIAL to be the immediate cause of falling for 19 of DISTRIBUTION OF FEMALES these 22 nests, because they fell during or im- During the 1989 breeding season, 27 colonies mediately after a rainstorm. For example, eight containing a total of 795 nests were found in the nests fell when, during a heavy rain, the limb 55 km2 region around La Selva. As each female 126 MICHAEL S. WEBSTER

All were located in cleared or abandoned agri- a 12 cultural land and were less than 200 m from a forest edge.

CA Median colony size was 24 nests (n = 41 col- .P8 8 onies), and ranged up to 130 completed or nearly completed nests (Fig. 2). The size of colonies 3 varied somewhat acrossyears (Fig. 3), and fluc- s‘ tuations in size were not consistent across col- . 4 onies. One cause of fluctuations in colony size z” was the movement of individual females be- tween colonies: marked females switched colo- 0 nies on 43% of all breeding attempts (n = 28 20 40 60 80 loo 120 casesin which a marked female was seen during Number of Nests two consecutive nesting attempts). Although oropendola colonies are typically thought to be built in traditional sites, the focal 20 b study colony (La Selva Colony) occupied five different trees within a 1.5 km2 area from 1985 16 r through 1990 (Fig. 4). Three colony site changes W occurred when females abandoned active nests 2 12 and moved to a new colony site. Human distur- u cr bance was responsible for one of these moves (in 0 8 . 1989 the birds were disturbed by out attempts to accessnests for blood samples), and severe i2 4 weather appearedto be the causeof colony aban- donment in early 1987 (move from “Site A” to 0 “Coop”) and in mid- 1990 (abandonment of “LS 10 20 30 Road” site; it is not known to where females Number Nests per Cluster moved). The other colony site changesoccurred FIGURE 2. Number of nestsin (a) coloniesand (b) between breeding seasonsand were due to un- nestclusters (nests 5 1.5m apart)within colonies.Based known causes. Counting all colonies in the La on 4 1 coloniessurveyed during the 1989breeding sea- Selva area, small colonies appeared to be less son. Only nestsbuilt in discreteclusters (694 nestsin stable and more likely to switch location between 67 clusters)are includedin b. years than large colonies (Fig. 3). Nests were not randomly distributed within builds her own nest, this gives a density ofap- colony trees: of 597 nests in colonies where nest proximately 14.5 nesting females per km*. We clustering data were collected, 530 (88.8%) were encountered three casesin which pairs of colo- built in distinct clusters (Fig. 5), and only 67 nies were located within 100-400 m of each oth- (11.2%) were built in relative isolation (> 1.5 m er. In these cases,marked individuals from each from nearest neighbor). Many of the “isolated” colony visited and interacted with individuals at nests appeared to be grouped in loose, poorly the other. Therefore, nests occupying two trees defined clusters, but did not meet our criteria to a few hundred meters apart should probably be be considered part of a nesting cluster. The size considered part of the same colony. Excluding of nesting clusters ranged from two to 32 nests these three casesand colonies near the edge of (Fig. 2), with a median of eight nests. the study area, nearest neighbor distance be- tween colonies was 1.33 f 0.45 km (n = 9 col- THE TEMPORAL DISTRIBUTION OF onies). BREEDING FEMALES Most colonies (39 of 41) were located in single, Nesting began in early January of each year. The isolated trees whose canopies did not touch that degree of breeding synchrony within a colony of any other. These colony trees were tall, con- was low in this study, and sexually receptive fe- tained few lower branches or epiphytes along the males (those lining nests with leaves) could be trunk, and had large, umbrella-shaped crowns. found at the La Selva Colony throughout much FEMALE OROPENDOLA BREEDING BIOLOGY 127

-0 CrocodilePt

1986 1987 1988 1989 1990 Year FIGURE 3. Colony size versus year of observation. Shown are the number of nests counted in each colony near the end of each breeding season. Only those colonies observed in three or more seasonsare included. Dashed lines indicate that the colony changedlocations between seasons.The focal observation colony is shown with unfilled squares. of each breeding season (Fig. 6). The standard cause the colony was abandoned mid-way deviation of date of first nest-lining was 24.88 through the breeding season.) days in 1987 and 13.24 days in 1988. (The 1989 In contrast to the over-all asynchrony of col- data were excluded from synchrony analyses be- onies, breeding synchrony within a nesting clus-

FIGURE 4. Location of Montezuma Oropendolacolonies in La Selva area, 1985-l 990. Colony Band Crocodile Point Colony were located in the same two trees throughout this period, whereas females nesting in the focal colony (La Selva Colony) changed locations frequently. The name and years of occupation are given for each colony site used by the focal colony. Solid arrows show changesin colony location between breeding seasons, and dashedarrows show changesthat occurredwithin a season.Females nesting at the “LS Road” site abandoned that site in the middle of the 1990 season;it is not known where they nested subsequently. I 728 MICHAEL S. WEBSTER

I FIGURE 5. A photoof La SelvaColony in 1986 showing distinct nesting clusters(right side of colony tree) , and two relatively isolated nests (left lower side of colony tree).

ter was relatively high. Analyses of variance ation for randomly selectednests (mean + stan- showed that mean date of first nest-lining dif- dard error: 23.3 + 0.7 and 12.0 + 0.3 days, fered significantly among clusterswithin a colony respectively). (1987: df = 2, 24, F = 26.51, P 5 0.001; 1988 df = 1, 8, F = 5.63, P = 0.045); variation among DISCUSSION nesting clusters accounted for 41% (1988) and 69% (1987) of the total variance in mean nesting COSTS AND BENEFITS OF CLUSTERING date (ratio of between group sum-of-squares to NESTS WITHIN COLONIES total sum-of-squares). To test further whether One striking feature of Montezuma Oropendola nests within a cluster were more synchronous nesting colonies is the degree to which females than the colony as a whole, I randomly selected cluster their nests within the colony tree (Fig. 5, n nests from the colony and calculated the stan- see also Fraga 1989). This behavior is likely to dard deviation of first nest-lining for these nests be costly, as nest clustering is likely to magnify (where n = the mean number of nests in a cluster the costs of coloniality. For example, ectopara- in each year). By repeating this 100 times, I could site transmission has been shown to increasewith calculate the mean standard deviation of first- nest density in several speciesof colonial swal- nest lining. In both 1987 and 1988, the mean lows (Hoogland and Sherman 1976, Brown and standard deviation of nest lining date within a Brown 1986, Moller 1987). Oropendola nestlings nesting cluster (14.3 and 10.3 days, respectively), are parasitized by avian lice and , and was significantly lower than the standard devi- nests in clusters might be more susceptiblethan FEMALE OROPENDOLA BREEDING BIOLOGY 729

isolated nests to these parasites. In addition, fe- low-ranking males, as hasbeen suggestedfor some male oropendolas nesting in the same cluster polygynous mammals (Le Boeuf 1972, Trillmich fought frequently and stole nesting material from and Trillmich 1984, Wrangham and Rubenstein each other (see also Skutch 1954, Robinson 1986, Mesnick and Le Boeuf 1991). Although 1985a): Finally building several heavy oropen- male courtship behaviors are vigorous and ap- dola nests on a single limb appeared to increase pear to be disruptive to females (Webster 199 1b), the risk of that limb breaking and falling, par- the consequencesof courtship to females and the ticularly during heavy rains. Skutch (1954) also frequency ofcourtship in the absenceofnest clus- reported finding broken limbs with nests on the tering are unknown. ground below colonies, one of which contained 21 nests. FACTORS AFFECTING FEMALE SYNCHRONY The reasons for aggregated nesting within WITHIN COLONIES Montezuma Oropendola colonies are unclear. It Whether aggregatednesting reduces predation, is unlikely that nest clustering is the result of brood parasitism, or both, the benefit of such a females independently converging on certain nesting dispersion is likely to be enhanced by “safe” limbs within the colony (Lightbody and breeding synchrony among females. For exam- Weatherhead 1987; Post, unpubl. ms.), because ple, if females feeding nestlings are unlikely to the packing of nests together was extreme (Fig. chase cowbirds away from their nests, females 5), many seemingly suitable limbs were unoc- near egg-laying would not benefit from nesting cupied, and the limbs used for nesting sometimes near feeding females. Similar arguments have changedfrom one year to the next (pers. observ.). been made for protection from nest predators Although the nests of social hymenoptera were (Hoogland and Sherman 1976, Robinson 1985b, located in a few colony trees, in no casedid Mon- Wittenberger and Hunt 1985). tezuma Oropendolas cluster their nests near the Although selection should favor breeding syn- nests (see also Fraga 1989) as has been chrony among female oropendolas, synchrony reported for other speciesof oropendola and ca- was only evident within nest clusters,and nesting cique (Smith 1968, 1983; Robinson 1985b). It in each colony as a whole was relatively asyn- is also unlikely that female oropendolas derive chronous (Fig. 6). Synchrony within but not any foraging benefits (Horn 1968, Ward and Za- among nesting clusters was most likely due to havi 1973, Emlen and Demong 1975, Brown two proximate factors. First, becausethe rate of 1986, Greene 1987) by clustering their nests nest failure was high, many females who started within a colony, because it is easy to observe nests early in the seasonlost those nests and had females carrying food from a distance and fe- to renest later (at the same or different colony). males could easily monitor the successof others Young females might also start nests later than without entailing the costs of clustered nesting. females that have bred in previous seasons(e.g., Females might benefit from nesting in close Post, unpubl. ms.). Second, although females in proximity to each other through increased pro- established nesting clusters were extremely ag- tection from predators (e.g., Hoogland and Sher- gressivetoward females not nesting in that clus- man 1976, Picman et al. 1988) and/or brood ter (seealso Robinson 1985a), females in the very parasites(Clark and Robertson 1979, Wiley and earliest stagesof nest establishment were some- Wiley 1980). Nesting in synchronousclusters has what more tolerant of each other and would often been shown to reduce the risk of nest predation perch together for long periods of time (pers. in the closely related yellow-rumped cacique obs.). As a consequence, females starting nests (Cacicus cela, Robinson 1985b). Unfortunately, at approximately the same time built nests near I have never observed a potential nest predator each other rather than near females in later stages at or near an oropendola colony and this possi- of the nesting cycle, leading to local synchrony. bility remains untested. Some observations, though, suggestthat nests in clusters are better THE DISTRIBUTION OF FEMALES AND protected than isolated nests from interspecific MALE MATING STRATEGIES brood parasitism by giant cowbirds (Scaphidura Although the ultimate factors underlying the oryzivora, Webster 1994). Finally, females might highly clumped spatial distribution of female cluster their neststo allow high-ranking males to Montezuma Oropendolas are not entirely clear, defend them more easily against harassment by this nesting distribution appearsto have been an 730 MICHAEL S. WEBSTER

121 rl- 1987 I I 1

6 1

2

b

2 4 6 8 10 12 14 16 Time Period

FIGURE 6. Nesting synchrony at the La Selva Colony in 1987, 1988, and 1989. Bars show the number of females lining nestsduring five-day time periods in each year. Individual nestingclusters are shownwith different patterns. In 1989, nests at the focal colony were built in a single large, loosely organized group and could not be easilyassigned to distinctclusters. This colonysite wasdeserted on day 42 (arrow). important factor leading to the female-defense suggestedthat, before nest building begins, it mating strategy of males. If males defended ter- might be difficult for males to determine which ritories, rather than converging on nesting col- potential colony sites will be used. By waiting, onies, the vast majority would obtain no mates on the other hand, a high-ranking male can go at all. Furthermore, many colonies changed lo- to where females begin nesting, defend those fe- cation frequently and unpredictably (Figs. 3, 4), males, and thereby obtain a high number of cop- FEMALE OROPENDOLA BREEDING BIOLOGY 731

TABLE 1. Mating systemand degreeof breedingsynchrony in colonialbirds.

SD of nest date Species Mating system* (no. &Ys)t SOurCe MontezumaOropendola Highly polygynous 13.3, 24.9 This study Yellow-rumpedCacique Highly polygynous 38.9 Robinson1985b Boat-tailedGrackle Highly polygynous 7.0, 20.6$ Post 1992, submitted Black-headedWeaver Polygynous 16.9 Hall 1970 Viellot’s BlackWeaver Polygynous 18.1 Hall 1970 Brewer’sBlackbird Slightlypolygynous 5.5 Horn 1970 TricoloredBlackbird Slightlypolygynous 1.4 Lack and Emlen 1939 Bank Swallow Monogamous 6.3 Emlen and Demong 1975 Herring Gull Monogamous 3.1-8.1 Gochfeld1980 BlackSkimmer Monogamous 6.1-6.5 Gochfeld1980

* Degreeof polygyny: “monogamous” = most or all males obtain a single mate; “slightly polyp ous” = some males obtain more than one mate, but few obtain more than 2-3 mates; “highly polygynous” = most matings obtamed by a smal number of males at each colony; “polygynous” = some males obtain multiple mates, but exact level of poly t Standard deviation of nest completion date or date of 8”rst’ eggunknom. m each nest, measured across entire colony. Figures obtained or calculated from data presented in indicated source. $ First figure, reported by Post (unpubl. ms), includes only first nesting attempts; second figure was calculated for one colony (unpubl. data) and includes all nesting attempts.

ulations (see Webster, in press). Interspecific male mating strategies are associated with the comparisons support the hypothesisthat the spa- spatial and temporal distribution of female oro- tial distribution of females is an important cor- pendolas within a colony. Females tend to breed relate of polygynous mating systems in the in synchronous clusters within a colony (see American blackbird subfamily (Robinson 1986, above), and observations of color-marked indi- Webster 1992). viduals have shown that .males (1) focus their Interspecific comparisons also suggestthat the defensive efforts on those clusterscontaining the temporal distribution of sexually receptive fe- greatestnumber of receptive females at any given males affects male mating strategies.Females of time, and (2) shift the focus of their defensive polygynous colonial speciesshow relatively little efforts away from areas where females have laid breeding synchrony within a colony (Table l), eggs(Webster, in press). Observations of copu- and sexually-receptive females are available lations (Webster 199 1b, in press) suggestthat throughout most of the breeding season (see asynchrony between clustersallows high-ranking Robinson 1985a, 1985b; Post 1992, submitted; males to more effectively guard females and mo- this study). In contrast, females of most colonial nopolize matings at large colonies: although the speciesshow a high degreeof breeding synchrony top two ranking males obtained virtually all cop- within a colony (Gochfeld 1980) and are mo- ulations at most colonies, a minimum of four nogamous or only slightly polygynous (Table 1). different males obtained copulations at one large A high number of females simultaneously recep- colony with three simultaneously active nesting tive will likely reduce the opportunity for polyg- clusters(La Selva Colony, 1986, 38 nests).How- ynous matings (Emlen and Oring 1977): if a male ever, this may not hold for smaller colonies: at must spend some time courting and/or guarding the La Selva Colony in 1987 (26 nests),two clus- a female during her receptive period, and females ters were simultaneously active (see Fig. 6) yet breed synchronously, few mating opportunities all observed copulations (n = 11) were obtained will remain after a male has mated with one by the alpha male. These results suggestthat the female (seeGrant and Kramer 1992). Under these ability of alpha male oropendolas to monopolize conditions, a male should remain with his single matings may be affectedby both colony size and mate and assist her in raising the brood. This within-colony female synchrony. hypothesismight also apply to non-colonial spe- In sum, female oropendolas and other polyg- cies. For example, breeding synchrony is lower ynous colonial speciesshow less breeding syn- among females of polygynous waterfowl than chrony than females of monogamous colonial among females of monogamous species(McKin- species,supporting the hypothesisthat, in species ney 1985, Sorenson 1992). where male parental care is not essential, a high Finally, the results of this study indicate that level of female breeding synchrony might con- 732 MICHAEL S. WEBSTER strain the mating opportunities open to males. ecology: an evolutionary approach. Blackwell, Moreover, the mating tactics of male oropen- London. DRURY, W. H. 1962. Breeding activities, especially dolas appear to be related to the spatial and tem- nest building, of the yellowtail (Ostinops decu- poral pattern of females within the colony; be- mums) in Trinidad. Zoologica 47:39-58. causefemales nest in synchronousclusters within DUNN, P. O., ANDR. J. ROBERTSON.1992. Geograph- the colony, high-ranking males are able to focus ic variation in the importance of male parental their defensive effortsand simultaneously defend care and mating systemsin Tree Swallows.Behav. Ecol. 3:291-299. those few females that are sexually receptive at EMLEN,S. T., AND N. J. DEMONG. 1975. Adaptive any given time. Localized synchrony may not be significanceof synchronizedbreeding in a colonial necessaryfor polygyny to evolve in some species : a new hvoothesis. Science 188:1029-l 03 1. (e.g., Post, unpubl. ms.), and a moderate level of EMLEN,S. T., AND-L. W. OWING. 1977. Ecology, sex- ual selectionand the evolution of mating systems. polygyny may be possible even when all females Science 197:215-223. in a colony are highly synchronous (Table 1). FRAGA,R. M. 1989. Colony sizes and nest trees of Nevertheless, synchrony within a nesting cluster Montezuma Oropendolas in Costa Rica. J. Field and relative asynchrony between clusters ap- Omithol. 60:289-295. pearsto facilitate the ability of a very small num- GOLD, M. 1980. Mechanisms and adaptive val- ue of reproductive synchronyin colonial seabirds, ber of males to monopolize copulations at Mon- p. 244-281. In J. Burger, B. J. Olla, and H. E. tezuma Oropendola colonies. Winn [eds.], Behavior of marine , Vol. 4. Plenum, New York. ACKNOWLEDGMENTS GRANT,J.W.A.,AN~D. L. KRAMER. 1992. Temporal clumping of food arrival reduces its monopoli- I thank S. T. Emlen, P. W. Sherman and L. K. Sterk zation and defence by zebrafish, Brachydanio re- for advice and encouragement,and I. Ayub, J. Guild, rio. Anim. Behav. 44: 101-l 10. K. Merg, E. Phillips, and L. Sterk for unflaggingassis- GREENE,R. 1987. Individuals in an osprey colony tance and friendship in the field. I thank W. Post for discriminate between high and low quality infor- discussionand accessto unpublished material. Valu- mation. Nature 329:239-241. able comments on an earlier draft were provided by HALL, J. R. 1970. Synchrony and social stimulation C. Aquadro, S. Emlen, R. Hoy, H. Landel, F. Pot@, in colonies of the black-headed weaver Ploceus S. Pruett-Jones, P. Sherman, L. Sorenson and two cucullatusand Viellot’s black weaver Melanopte- anonymous referees. The Organization of Tropical ryx nigerrimus.Ibis 112:93-104. Studiesand JessieSmith Noyes Foundation provided HOOGLAND,J. L., AND P. W. SHERMAN. 1976. Ad- critical financial and logistic support.Further generous vantagesand disadvantagesof bank swallow (Ri- support was provided by the following organizations: pariu ripariu) coloniality. Ecol. Monogr. 46:33- American Museum of Natural History (Frank M. 58. Chapman Fund), Cornell University, National Insti- HORN, H. S. 1968. The adaptive significanceof co- tute of Mental Health, National Science Foundation lonial nestingin the Brewer’s blackbird (Euphagus (NSF Graduate Fellowship), Organization for Ameri- cyanocephalus).Ecology 49:682-694. can States, and Sigma Xi (Grant-in-Aid-of-Research). HORN,H. S. 1970. Socialbehavior of nestingbrewer ’s During preparation of the manuscript, I was supported blackbirds. Condor 72: 15-23. by NSF grant IBN-9207609 to S. G. Pruett-Jones.Fi- LACK, D. 1968. Ecologicaladaptations for breeding nally, I thank the La Selva staff and the many Costa in birds. Methuen and Co., iondon. Rican residentswho allowed me to work on their land. LACK. D.. AND J. T. EMLEN. 1939. Observations on breedingbehavior in tricolored red-wings.Condor LITERATURE CITED 41:225-230. LE Bonus, B. J. 1972. Sexualbehavior in the northern BROWN,C. R. 1986. Cliff Swallow colonies as infor- elephant seal Mirounga angustirostris.Behaviour mation centers. Science 234:83-85. 41:1-26. BROWN,C. R., AND M. B. BROWN. 1986. Ectopara- LEN~NGTON,S. 1980. Female choice and polygyny in sitism as a cost of coloniality in Cliff Swallows Red-wingedBlackbirds. Anim. Behav.281347-36 1. (Hirundo pyrrhonofa). Ecology 67: 1206-l 2 18. LIGHTBODY,J. P., AND P. J. WEATHERHEAD.1987. CHAPMAN,F. M. 1929. The nestinghabits of Wagler’s Interactions among females in polygynous Yel- Oropendola (2urhynchu.swugZeri> on Barro Col- low-headedBlackbirds. Behav. Ecol. Sociobiol.2 1: orado Island. Bull. Am. Mus. Nat. Hist. 58:123- 23-30. 166. MCDADE, L. A., K. S. BAWA,H. A. HESPENHEIDE,AND CLARK,K. L., AND R. J. ROBERTSON.1979. Spatial G. S. HARTSHORNE.1993. La Selva: ecologyand and temporal multi-species nesting aggregations natural history of a Neotropical rainforest. Univ. in birds as anti-parasite and anti-predator defens- of Chicago Press, Chicago. es. Behav. Ecol. Sociobiol. 5:359-37 1. MC-, F. 1985. Primary and secondary male DAVIES,N. B. 1991. Mating systems,p. 263-300. In reproductive strategiesof dabbling ducks, p. 68- J. R. Krebs and N. B. Davies [eds.], Behavioural 82. In P. A. Gowaty and D. W. Mock [eds.],Avian FEMALE OROPENDOLA BREEDING BIOLOGY 733

monogramy, Ornithological Monographs #37. TRILLMICH,F., ANDK. TRILLMICH. 1984. The mating American Ornithologists’Union, Washington,DC. system of pinnipeds and marine iguanas:conver- MEASE,D., ANDE. MEASE. 1980. Aerial net assembly. gent evolution of polygyny. Biol. J. Linn. Sot. 2 1: N. Am. Bird Bander 5: 138-l 39. 209-216. MESNICK,S. L., AND B. J. LE BOEUF. 1991. Sexual WARD, P., ANDA. ZAHA~I. 1973. The importance of behavior of male northern elephant seals:II. Fe- certain assemblagesof birds as ‘information cen- male responseto potentially injurious encounters. tres’ for food-finding. Ibis 115:517-534. Behaviour 117:262-280. WEBSTER,M. S. 1991a. Male parental care and po- MATER, A. P. 1987. Advantages and disadvantages lygyny in birds. Amer. Nat. 137:274-280. of coloniality in the swallow, Hirundo rustica. WEBSTER,M. S. 1991b. The dynamics and conse- Anim. Behav. 35:819-832. quences of intrasexual competition in the Mon- ORING, L. W. 1982. Avian mating systems. Avian tezuma Oropendola: haremlpolygyny in a Neo- Bio. 6: l-92. tropical bird. Ph.D.diss.. Cornell Univeristv.. PICMAN,J., M. LEONARD, AND A. HORN. 1988. An- Ithaca, NY. tipredation role ofclumped nestingby marsh-nest- WEBSTER,M. S. 1992. Sexual dimophism, mating ing Red-winged Blackbirds. Behav. Ecol. Socio- system and body size in New World blackbirds biol. 22:9-15. (Icterinae). Evolution 46: 162l-l 64 1. POST,W. 1992, Dominance and mating successin WEBSTER,M. S. 1994. Interspecific brood parasitism male Boat-tailed Grackles. Anim. Behav. 44:9 17- of Montezuma Oropendolas by giant cowbirds: 929. narasitism or mutualism? Condor. 96:794-798. ROBINSON,S. K. 1985a. Benefits, costs, and deter- WEB&R, M. S. In press. Female-defence polygyny minants of dominance in a polygygnousoriole. in a neotropicalbird, the Montezuma Oropendola. Anim. Behav. 34:241-255. Anim. Behav. 48. ROBINSON,S. K. 1985b. Coloniality in the Yellow- WESTNEAT,D. F., P. W. SHERMAN,AND M. L. MORTON. rumped Cacique as a defense against nest preda- 1990. The ecology and evolution of extra-pair tors. Auk 102:506-519. copulationsin birds. Current Ornithology 7:331- ROBINSON, S. K. 1986. The evolution of social be- 369. havior and mating systemsin the blackbirds (Ic- WILEY, R. H., AND M. S. WILEY. 1980. Spacingand terinae), p. 175-200. In D. I. Rubenstein and R. timing in the nesting ecology of a tropical black- W. Wrangham [eds.], Ecologicalaspects of social bird: comparison of populations in different en- evolution. Princeton Univ. Press, Princeton, NJ. vironments. Ecol. Monogr. 50: 153-l 78. SCHXFER,E. 1957. Les conotos: etude comparative W~I-~ENBERGER,J. F. 1980a. Group size and polyg- de Psarocoliusangustifons et Psarocoliusdecu- amy in social mammals. Am. Nat. 115:197-222. manus. Bonner ZoologischeBietrage 1957:l-l 5 1. W~~-~-ENEIERGER,J. F. 1980b. Vegetation structure, SKUTCH, A. F. 1954. Life histories of central Amer- food supply, and polygyny in bobolinks (Dolicho- ican birds. Pacific Coast Avifauna #3 1. nyx oryzivorzu).Ecology 6 1:140- 150. SMITH,N. G. 1968. The advantageof being parasit- WITTF~VBERGER,J. F., AND G. L. HUNT. 1985. The ized. Nature 219:69&694. adaptive significanceof coloniality in birds. Avian SMITH,N. G. 1983. Zarhynchuswagleri, p. 614-616. Biol. 8: I-78. In D. H. Janzen [ed.], Costa Rican natural history. WRANGHAM,R. W., AND D. I. RUBENSTEIN.1986. Univ. of Chicago Press, Chicago. Social evolution in birds and mammals, p. 452- SORENSON,L. G. 1992. Variable mating system of a 470. In D. I. Rubenstein and R. W. Wrangham sedentarytropical duck:the White-cheeked Pintail [eds.], Ecological apsects of social evolution. (Anasbahamensis bahamensis). Auk 109~277-292. Princeton Univ. Press, Princeton, NJ. TASHIAN,R. E. 1957. Nesting behavior of the (Psarocoliusdecumanus) in northern Trinidad, B.W.I. Zoologica 42:87-98.