Wilson Bulletin 116(4):337±341, 2004

FIRST RECORD OF COOPERATIVE BREEDING IN A THRYOTHORUS

SHARON A. GILL1,2

ABSTRACT.ÐAlthough offspring delay dispersal past the age of independence in many Thryothorus species, cooperative breeding has not been recorded in this genus. Here, I present the ®rst observation of cooperative breeding in a Thryothorus wren (Buff-breasted Wren, T. leucotis). Of 41 offspring that delayed dispersal past the age of independence, 4 individuals stayed on their natal territory until their parent's next breeding attempt, indicating a low potential for cooperative breeding. Of these four individuals, one male provisioned 11- to 15- day-old nestlings and one female approached her parents' nest but was repeatedly driven away from it by her father. The retained female was apparently tolerated when in the vicinity of ¯edglings, but feeding was never positively con®rmed. Received 10 May 2004, accepted 15 October 2004.

In cooperatively breeding , more than spring is often associated with high adult sur- two adults participate in a single breeding vival, low reproductive rates, and deferred event, either by defending the nest, incubating maturation (Brown 1987), and may be a form eggs, or provisioning young (Brown 1987). of extended parental investment, in which par- This is the predominant social system in ents permit or encourage offspring (via access 18.5% of oscine species with bipa- to resources) to remain on natal territories rental care (383 of 2,067 species), and occurs (Ekman et al. 2001). occasionally in another 3.7% of oscines clas- (Family Troglodytidae) have been si®ed as pair breeders (Cockburn 2003). How- among the most studied New World avian ever, patterns of nestling care are unknown for taxa in terms of cooperative breeding, with most (n ϭ 2,385), particularly much of this research focusing on the genus those in the Neotropics (Cockburn 2003); , in which 9 of 13 species thus, cooperative breeding may occur in still are known to breed cooperatively (e.g., Aus- other species. Intensive studies of populations tad and Rabenold 1986, Rabenold 1990). Co- of banded birds are needed to determine the operative breeding occurs regularly in Black- extent of pair versus cooperative breeding in capped Donacobius (Donacobius atricapillus) these species. and has been reported in House Wrens (Trog- Cooperative breeding may arise in several lodytes aedon; reviewed in Cockburn 2003). different ways, including when (1) indepen- However, the breeding systems of most spe- dent offspring remain on natal territories until cies in this family, especially those in tropical their parents' next breeding period (Cockburn areas, have not been described and coopera- 1998), (2) individuals initially disperse and tive breeding may be more common than re- then preferentially immigrate into social ported. For example, many species in other groups composed of related individuals (Bag- genera, in particular Thyrothorus, are often lione et al. 2003), and (3) groups of unrelated found in family groups (e.g., Skutch 1960, individuals reproduce together (Davies 1992). 2001; Slud 1964; Hilty and Brown 1986; In the ®rst case, delayed dispersal by offspring Brewer 2001), suggesting that independent is a necessary prerequisite for cooperation. offspring may stay on natal territories for ex- However, in some species in which offspring tended periods, potentially setting the stage remain on their natal territories for extended for cooperative breeding. Here, I describe the periods, helping never occurs (reviewed in ®rst record of cooperative breeding in the ge- Ekman et al. 2001). Delayed dispersal by off- nus Thryothorus made during a study of pat- terns of nestling provisioning in Buff-breasted Wrens (T. leucotis; SAG unpubl. data). Pre- 1 Dept. of Biology, York Univ., Toronto, ON M3J vious studies have reported that Buff-breasted 1P3, Canada. 2 Current address: Dept. of Ecology and Evolution- Wrens are often found in groups of 3±4 in- ary Biology, Guyot Hall, Princeton Univ., Princeton, dividuals (Skutch 1968, Farabaugh 1983, NJ 08544, USA; e-mail: [email protected] Ahumada 2001), but none has recorded dis- 337 338 THE WILSON BULLETIN • Vol. 116, No. 4, December 2004 persal delayed into the breeding season or co- gave their ®rst song of the morning (these are operative breeding. typically given when birds are close to their nests), during regular searches of territories, METHODS and by following individuals that were col- Buff-breasted Wrens are small (16±23 g), lecting or holding nesting materials. I did not monochromatic birds that inhabit secondary mark nest locations, as they were easy to re- forest throughout their range from central ®nd. Nests were checked every 2 days until Panama to northern South America (Ridgely clutch completion (modal clutch size ϭ 3, and Tudor 1989). They maintain territories range ϭ 2±3, n ϭ 42) and irregularly there- and partnerships throughout the year, and after until the young ¯edged or the nest failed. breed during the wet season (April to October; As part of a broader study examining male SAG unpubl. data); initiation of breeding is and female effort during nest construction and closely tied to the onset of rains (Ahumada nestling provisioning, I made 1-hr observa- 2001). Buff-breasted Wrens are socially and tions of nestling provisioning every 2 days genetically monogamous: extra-pair young from the day after hatching (day 0) until the were detected in only 1 of 31 broods (Gill et nestlings ¯edged (usually day 15). All obser- al. 2005). Throughout the year, pairs produce vations were performed between 06:00 and antiphonal duets in which males and females 14:00 EST, from a position at least 5 m from give alternating sex-speci®c songs (Farabaugh nests, to avoid in¯uencing the behavior of the 1983). parents as they returned to feed or brood (fe- I studied a color-banded population of Buff- males only). During these observations, I breasted Wrens in Gamboa, Republic of Pan- quanti®ed the number of visits per hour by ama (9Њ 7Ј N, 79Њ 42Ј W) during both dry and male and female parents, as well as visits by wet seasons (February±May 1997, February± the retained offspring. In addition, I noted the July 1998, October 1998, February±June singing behavior of offspring, as well as ag- 1999). The study site was a 22-ha secondary gressive interactions between them and their forest bordered on three sides by Gamboa, the parents. Rio Chagres, and the Panama Canal, and it was separated from the nearest forest by a RESULTS 100-m grassy marsh. In 1997, both individuals in 15 of 18 pairs were banded, whereas in Of 57 offspring banded as nestlings, ¯edg- 1998 (n ϭ 24) and 1999 (n ϭ 19), both mem- lings, or independent birds, 35 remained on bers of all pairs were banded. I banded a total their natal territories for Ͼ3 months after of 65 adults, 17 adult philopatric offspring, 27 ¯edging. Another six unbanded individuals ¯edglings, and 13 nestlings. I sexed individ- lived on territories with banded adults. I as- uals based on body size (females are smaller sumed that these unbanded individuals were than males; SAG unpubl. data), singing be- the offspring of the banded adults because (1) havior (Farabaugh 1983), and egg laying and their presence was consistent with the ¯edging incubation (Gill 2003). When more than two of offspring in those territories in the previous adult birds resided on a given territory, I dis- breeding season; (2) 17 individuals, banded as tinguished paired adult birds from their adult adults and residing on territories with 2 other philopatric offspring by behavior, as paired in- adult birds, were the genetic offspring of those dividuals duetted more frequently, and spent adults (SAG unpubl. data); and (3) I have no more time in close association (Ͻ5 m apart) evidence that groups formed other than by the than with their offspring. I de®ned a philo- retention of offspring. Thus, a total of 41 off- patric offspring as one that remained on its spring (65.1% of offspring ¯edged) remained natal territory for Ͼ3 months. on their natal territories for Ͼ3 months. Prior Buff-breasted Wrens typically construct to their parents' next breeding attempt, most separate dormitory nestsÐin which they roost philopatric offspring either dispersed to be- overnightÐand breeding nests, although ap- come territory holders in the study population proximately 10% of dormitory nests are used (n ϭ 13) or disappeared (n ϭ 24). Only four for breeding (SAG unpubl. data). I located individuals (9.8% of philopatric offspring) nests by searching areas from which pairs postponed dispersal for Ͼ1 year, staying on Gill • HELPING IN THRYOTHORUS LEUCOTIS 339

30 min of this interaction, while the female visited the nest four times within that period. Near the end of the 1-hr observation period, the unbanded male, with food in his bill, tried to approach the nest once more. The parental male was in the nest, exited it and then chased the unbanded male away from the nest. Soon after, the parental male and the unbanded male were observed foraging within2mofeach other. On day 13, the unbanded male visited the nest six times within an hour, equaling the contribution of the parents combined. His ®rst FIG. 1. Pattern of provisioning by a single pair of visit was apparently undetected by the parents Buff-breasted Wrens and a philopatric, unbanded male when feeding three nestlings. Eggs hatched on day 0 as they were not in the immediate nest area at and were observed every second day from days 1 to the time. While approaching the nest for his 15, at which time the young ¯edged. The philopatric second visit, the unbanded male was supplant- male was not observed within5mofthenest area ed by the male parent, who gave an aggressive prior to day 11. churr, but the unbanded male was not chased out of the area. Rather, he continued to move around the nest, with the male parent follow- their natal territories through their parents' ing 1±2 m behind. The unbanded male then subsequent breeding season. entered the nest and fed the young. On the Between 6 and 10 June 1999, I observed next four visits, the unbanded male was fol- helping at the nest by an unbanded male dur- lowed several times by the male parent, but ing the second breeding attempt of a banded no aggression was exhibited. On the day of pair that had been together for Ͼ2 years. This ¯edgling (day 15), the unbanded male made pair had ¯edged two males in the previous four visits to the nest, more than the individual breeding season, one of which I captured and effort of either parent (Fig. 1). No aggression banded. Both male offspring were present on between the unbanded male and the parental the territory at the start of my ®eld season in male was observed at this time, and no ag- February 1999, with the banded male dispers- gression between the unbanded male and the ing to another territory before breeding com- female was ever recorded. Three young menced. The unbanded male was last ob- ¯edged from the nest on 10 June. I made three served on his natal territory on 30 June, a date additional untimed observations of this pair that corresponded to the end of my ®eld sea- after the young ¯edged, during which the un- son. banded male perched within 2 m of the ¯edg- Visitations by the parents and the unbanded lings and their parents, and once fed the ¯edg- male while provisioning a brood of three nest- lings. lings are summarized in Figure 1. Between During observations at a second nest in days 1 and 9 after hatch, the unbanded male June 1999, I witnessed aggressive encounters was not observed within 10 m of the nest, but between a banded female offspring (hatch often sang within the territory. On day 11 near year 1998) and her banded father around the the start of observations, the unbanded male breeding nest, but never observed the female entered the nest and stayed in it for approxi- enter the nest to feed the young. On day 5, mately 10 sec. Shortly after the unbanded the philopatric female approached within 5 m male left the nest there was a chase involving of the nest and was chased by the male 3±4 two unidenti®ed wrens, and ``aggressive'' times. She moved to 2 m from the nest and churring (such calls are heard only during ag- was chased again by the male parent, who gressive encounters; SAG pers. obs.). Less gave several aggressive churrs. When the than 30 sec later, a third unidenti®ed wren ap- nestlings were 13 days old, the philopatric fe- proached this pair, and another chase ensued. male perched 4±5 m from the nest and was The parental male did not visit the nest within immediately chased by the male, who uttered 340 THE WILSON BULLETIN • Vol. 116, No. 4, December 2004 aggressive churrs (the philopatric female did care to prevent non-breeders from accessing not approach the nest during observations nests, possibly to decrease activity around the made before day 5, between days 7 and 11, nest and thereby minimize the risk of nest pre- or on day 15). Two nestlings ¯edged on 6 dation (Strickland and Waite 2001). Consid- June, and I observed the philopatric female ering pair-breeding species in the Neotropics, within3mofthem on the day of ¯edging as Skutch (1949) proposed that nest predation in- well as during observations 2 weeks later. At creases with the rate at which parents feed the both times, the family was in thick vegetation nestlings (see Martin et al. 2000). Thus, male and I made few de®nitive observations of parents may have behaved aggressively to- feeding by any adult. No aggression was not- ward their offspring to stop them from visiting ed between the male and the philopatric fe- nests. Interestingly, in the one case of coop- male after the young ¯edged, nor was aggres- erative breeding I observed, parents appeared sion observed at any time between the paren- to visit their nest less once their male off- tal and philopatric females. spring started to provision the young (Fig. 1). Two other banded offspring were observed Alternatively, parental aggression toward phil- on their natal territories during subsequent opatric offspring may occur when resources breeding attempts of their banded parents, but become depleted and competition for the re- helping was not observed. I observed a band- maining resources intensi®es (Strickland and ed, second-year female offspring on her natal Waite 2001). This hypothesis seems unlikely territory 1 day before her parents began con- to apply to Buff-breasted Wrens, as breeding structing their second breeding nest, but did occurs during the wet season, when arthropod not see her during 4 hr of observation of nest- prey are abundant (e.g., Wolda 1996). Prey ling provisioning. However, this female ap- depletion seems more likely to occur during parently stayed on her natal territory while her the dry season, yet I witnessed little or no pa- parents bred, as she was observed with them rental aggression toward philopatric offspring in the post-breeding season. Finally, a banded at that time. male was observed repeatedly foraging and Most offspring that delayed dispersal did singing with his banded parents over the not stay on their natal territories long enough course of their four breeding attempts. The to help with reproductive activities of their philopatric male built a dormitory nest with parents. On average, offspring dispersed just his father (while the female incubated eggs), prior to the onset of their parents' reproduc- but did not participate in the construction of tive period (SAG unpubl. data). Prior to dis- two breeding nests during timed observations. Although I did not observe this pair during persing, however, some offspring participated nestling provisioning, the male offspring was in the construction of dormitory nests, and still present on the territory at that time. most sang either as part of territory defense or to advertise themselves to potential mates. DISCUSSION Further studies are needed to determine the The potential for observing cooperative role of offspring in these activities, as well as breeding in Buff-breasted Wrens was low, as the bene®ts and costs of philopatry for both only 4 of 41 independent offspring postponed parents and their offspring. The phenomenon dispersal for Ͼ1 year. All four offspring that of delayed dispersal in the absence of coop- delayed dispersal may have participated in ter- erative breeding is underappreciated, yet its ritorial defense by singing on their own or study is a logical step toward understanding with their parents. One male helped by feed- the evolution of cooperation (Brown 1987). ing young, and one female appeared to try to The monophyletic wren family (Barker 2004) provision nestlings and may have successfully offers an exceptional opportunity to study the fed ¯edglings. The latter cases were marked evolution of delayed dispersal and cooperative by aggression by male, but not female, parents breeding due to the diversity of breeding and toward their adult offspring during the nest- dispersal patterns, ranging from pair breeders ling period. Similar aggressive interactions with early dispersal by juveniles, to those with between parents and non-breeders occur in delayed juvenile dispersal without coopera- some small-bodied corvids during nestling tion, to truly cooperative breeders. Gill • HELPING IN THRYOTHORUS LEUCOTIS 341

ACKNOWLEDGMENTS 2001. Delayed dispersal: living under the reign of nepotistic parents. Auk 118:1±10. I thank the Autoridad Ambiente del Nacional for FARABAUGH, S. M. 1983. A comparative study of duet permission to work in Panama and the Smithsonian song in tropical Thyrothorus wrens. Ph.D. disser- Tropical Research Institute for logistical support. M. tation, University of Maryland, College Park. Griesser and two anonymous referees made construc- GILL, S. A. 2003. Timing and duration of egg-laying tive comments on an earlier version of this manuscript. in duetting Buff-breasted Wrens. Journal of Field This study was supported by grants and scholarships Ornithology 74:31±36. from the American Ornithologists' Union, American GILL, S. A., M. J. VONHOF,B.J.M.STUTCHBURY,E. Museum of Natural History, Behavior Society, S. MORTON, AND J. S. QUINN. 2005. No evidence Ontario Graduate Scholarships, Sigma Xi, and York for acoustic mate-guarding in duetting Buff- University to SAG; Natural Science and Engineering breasted Wrens (Thryothorus leucotis). Behavioral Research Council of Canada to B. J. M. Stutchbury, Ecology and Sociobiology. In press. and Smithsonian Institution to E. S. Morton and B. J. HILTY,S.L.AND W. L. BROWN. 1986. A guide to the M. Stutchbury. birds of . Princeton University Press, Princeton, New Jersey. LITERATURE CITED MARTIN, T. E., J. SCOTT, AND C. MENGE. 2000. Nest predation increases with parental activity: sepa- AHUMADA, J. A. 2001. Comparison of the reproductive rating nest site and parental activity effects. Pro- biology of two Neotropical wrens in an unpre- ceedings of the Royal Society of London, Series dictable environment in northeastern Colombia. B 267:2287±2293. Auk 118:191±210. RABENOLD, K. N. 1990. Campylorhynchus wrens: the AUSTAD,S.N.AND K. N. RABENOLD. 1986. Demog- ecology of delayed dispersal and cooperation in raphy and the evolution of cooperative breeding the Venezuelan savanna. Pages 157±196 in Co- in the Bicolored Wren, Campylorhynchus griseus. operative breeding in birds: long term studies of Behaviour 97:308±324. ecology and behavior (P. B. Stacey and W. D. BAGLIONE, V., D. CANESTRARI,J.M.MARCOS, AND J. Koenig, Eds.). Cambridge University Press, New EKMAN. 2003. Kin selection in cooperative alli- York. ances of Carrion Crows. Science 300:1947±1949. RIDGLEY,R.S.AND G. TUDOR. 1989. The birds of BARKER, F. K. 2004. Monophyly and relationships of South America. University of Texas Press, Austin. wrens (Aves: Troglodytidae): a congruence anal- SKUTCH, A. F. 1949. Do tropical birds rear as many ysis of heterogeneous mitochondrial and nuclear young as they can nourish? Ibis 91:430±455. DNA sequence data. Molecular Phylogenetics and SKUTCH, A. F. 1960. Life histories of Central American Evolution 31:486±504. birds. Paci®c Coast Avifauna, no. 34. BREWER, D. 2001. Wrens, dippers and thrashers. Yale SKUTCH, A. F. 1968. The nesting of some Venezuelan University Press, New Haven, Connecticut. birds. Condor 70:66±82. BROWN, J. L. 1987. Helping and communal breeding SKUTCH, A. F. 2001. Life history of the Riverside in birds: ecology and evolution. Princeton Uni- Wren. Journal of Field Ornithology 72:1±11. versity Press, Princeton, New Jersey. SLUD, P. 1964. The birds of Costa Rica: distribution COCKBURN, A. 1998. Evolution of helping behavior in and ecology. Bulletin of the American Museum of cooperatively breeding birds. Annual Review of Natural History, no. 128. Ecology and Systematics 29:141±177. STRICKLAND,D.AND T. A. WAITE. 2001. Does initial COCKBURN, A. 2003. Cooperative breeding in oscine suppression of allofeeding in small jays help to passerines: does sociality inhibit speciation? Pro- conceal their nests? Canadian Journal of Zoology ceedings of the Royal Society of London, Series 79:2128±2146. B 270:2207±2214. WOLDA, H. 1996. Seasonality of Homoptera of Barro DAVIES, N. B. 1992. Dunnock behaviour and social Colorado Island. Pages 319±330 in The ecology evolution. Oxford University Press, Oxford, Unit- of a tropical forest, 2nd ed. (E. G. Leigh, Jr., A. ed Kingdom. S. Rand, and D. M. Windsor, Eds.). Smithsonian EKMAN, J., V. BAGLIONE,S.EGGERS, AND M. GRIESSER. Institution Press, Washington, D.C.