SHORT COMMUNICATIONS 523 ing Research Council of Canada. Jeffrey’s multilocus tensity by male Purple Martins. Behaviour 101: probes were provided by the Cellmark Division of ICI. 21 l-224. We especially thank Patricia Parker and Paul Fuerst MORTONE. S., L. FORMAN,AND M. BRAUN. 1990. Ex- for generouslyproviding laboratory resourcesand ex- trapair fertilizations and the evolution of colonial pert advice on DNA fingerprinting. breeding in Purple Martins. Auk 107:275-283. MORTON,E. S., AND R. L. PATERSON.1983. Kin as- LITERATURE CITED sociation, spacing and composition of a post- breeding roost of Purole Martins. J. Field Orni- ALLEN,R. W., ANDM. M. NICE. 1952. A study of the thol. 54736-41. I breeding biology of the Purple Martin (Pt-ogne STUTCHBLJRY,B. J. 1991. The adaptive significanceof subis). Am. Midl. Nat. 47:606-665. male subadultplumage in Purple Martins: plum- DAVIDAR,P,’ AND E. S. MORTON. 1992. Living with age dyeing experiments. Behav. Ecol. Sociobiol. parasites: prevalence of a blood parasite and its 291297-306. effects on survivorshipin the Purple Martin. Auk WAGNER,R. H., M. D. SCHUG, AND E. S. MORTON. 110:109-l 16. 1996a. Condition dependent control of paternity HAMILTON,W. D. 1990. Mate choice near or far. Am. by female Purple Martins: implications for colon- Zool. 30:341-352. iality. Behav. Ecol. Sociobiol. 38:379-389. HAMILTON,W. D., AND M. ZLJK. 1982. Heritable true WAGNER,R. H., M. D. SCHUG,AND E. S. MORTON. fitness and bright birds: a role for parasites?Sci- 1996b. Confidence of paternity, actual paternity ence 218:384-386. and parental effort in Purple Martins. Anim. Be- MORTON,E. S. 1987. Variation in mate-guardingin- hav. 52:123-132.

The Condor 99~523-525 0 The CooperOm~tholog~cal Society 1997

COOPERATIVE BREEDING IN GRAY JAYS: PHILOPATRIC OFFSPRING PROVISION JUVENILE SIBLINGS’

THOMAS A. WAIT@ Department of Zoology, Ohio State University, Columbus, OH 43210

DAN STRICKLAND RR 1, Dwight, Ontario POA lH0, Canada

Abstract. We present evidence for cooperative The Gray Jay (Perisoreus canadensis) is conspicuous breeding in Gray Jays (Perisoreus canadensis), a spe- among the Corvidae for its apparentlack of coopera- ties characterizedby delayeddispersal but long thought tive breeding (Strickland 1991, Strickland and Ouellet not to exhibit the life-history tacticof helping. Our find- 1993). The failure to detect cooperative breeding in ings to date suggestthat the alloparentalcare of younger this species,and in its Eurasiancongener (Siberian Jay, siblingsby philopatricyearlings is confinedto the post- P. infaustus; Blomgren 1971, Lindgren 1975, Ekman fledgiig pehod of nut&ionaldependence..We e&our- et al. 1994), is unexpectedbecause these jays are char- age other workers to investigatewhether helping may acterized by delayed dispersal(Strickland and Ouellet be confined to this period in other semi-social species 1993), the necessarybut not sufficient precursorof co- as well. Ongoing work is aimed at evaluatingthe fitness operative breeding (Koenig et al. 1992), and because consequencesof helping in our study population. cooperative breeding is thought to be a phylogeneti- tally conserved trait in corvids (Edwards and Naeem Key words: cooperative breeding, delayed dispers- 1993, Cockburn 1996). al, helping, Gray Jay, Perisoreuscanadensis. We were prompted to renew our search for coop- erative breeding in Gray Jays by the apparently skep- tical suggestionthat alloparental care of younger sib- lings may occur during the post-fledgingperiod of nu- ’ Received 21 October 1996. Accepted 28 January tritional dependence (Ekman et al. 1994). Here, we 1997. documentthe discovery of helping during that period z Current address: School of Forestry, Michigan and encourageworkers to searchfor, and examine the TechnologicalUniversity, Houghton, MI 4993 l-1295, fitness consequencesof, this form of cooperation in e-mail: [email protected] other semi-social species. 524 SHORT COMMUNICATIONS

METHODS young with both retrieved caches and regenerating We studied Gray Jays in Algonquin Provincial Park, food. In 1994, the son, GOS, made 77 of 344 (22%) Ontario, Canada. Data collection took place during total feeds distributed among his younger siblings. Of April-May 1994, March-June 1995, and April-May the remainder,the mother, WOL, made 123 (36%) and 1996. Air temperatureranged from -6 to 30°C snow the father, ROS, made 144 (42%) feeds. This extensive and rain occurred on several days, and snow cover provisioning by the helper occurred despite parental persistedinto April each year aggressiontowards him (52 recorded supplantingat- Social groups of Gray Jays occupy year-round, all- tacks and chases),which deceleratedacross days (rF = purpose territories, where they store surplus food -0.79, P = 0.04, two-tailed; overall rate: 0.9 interac- (Strickland and Ouellet 1993, Waite and Ydenberg tions hrr). No significant trend acrossdays emerged 1996). These caches are exploited extensively during in the proportion of feeds performed by GOS (I; = winter and sometimesare used by adults to provision 0.33, P = 0.42). Provisioning by all three adults was nestlings and fledglings. The typical social group in still occurring on 13 May (19 days post-fledge), but our study area is a mated pair, which may be accom- had stoppedby 6 June. panied by a nonbreederwho is either an offspring or In 1995, the philopatric offspring (ROG) was con- an unrelatedimmigrant. This group compositionarises sistently supplantedand chased upon approachingthe several weeks post-fledging. The dominant brood- nest, and was never observed to provision the sole member evicts its subordinatesiblings and remains on nestling, his putative half-sibling, LOS. The father the natal territory, where parental facilitation appar- (ROS) made 38 feeds and the mother (WOB) made 8 ently enhancessurvival (Ekman and Rosander 1992). feeds. However, in the 25-day period beginning one Surviving siblings settle singly with unrelated pairs. day after LOS fledged, the older sibling, ROG, made During the nest-building phase of the subsequent 129 of 329 (39%) total observed feeds. The mother breeding season,about 20% of mated pairs are accom- made only 36 (11%) feeds, whereas the father made panied by a nonbreeder,-75% of which are philopa- 164 (50%) feeds. On a daily basis, ROG’s provisioning tric offspring (unpubl. data). effort (number of feeds) typically exceeded that of the We made detailed observationsduring the breeding mother (14 of 15 days; sign test: P < O.OOl),but not seasonin one territory during both 1994 and 1995. The the father (5 of 15 days; P = 0.30). As in the previous adults had occupied the territory since 1992. All jays year, the rate of parental aggressiontowards the helper were color-banded.In the spring of 1994, the territorv diminished acrossdays (rs = -0.81, P < 0.001; over- was occupied by a mated pair-(female, WOL; male, all rate: 0.2 hrr’; 19 total interactions),and no signif- ROS), along with a son (GOS) from the 1993 breeding icant temporal trend emergedin the proportionof feeds season. Three fledglings were produced in 1994, in- performed by ROG (rs = -0.20, P = 0.47). Provi- eluding ROG. Muitil&s minisatellite DNA finger- sioning was last seen 26 days post-fledge. printing confirmed that WOL and ROS were the par- In 1996, the yearling nonbreeder(YOS) was chased ents of GOS and the three fledglings. In October,WOL by the adults on the three occasions when it ap- disappearedfollowing an injury and was replaced by proached the nest. It was never observed to provision a first-year female (WOB) from an adjoining territory. the sole nestling, his putative half-sibling, which was In the spring of 1995, the territory was occupied by fed 12 times by each parent. Likewise, YOS was never ROS, the same breeding male; ROG, his philopatric observed to provision the young during the post-fledg- son; and WOB, stepmotherto ROG. ROS and WOB ing period, when the father made 90 feeds and the producedone (female) fledgling, LOS. mother made 100 feeds. In 1994, we conducted focal watches of the three In 1994, GOS made several contributions classifi- fledglings for 56 hr during nine days between 1 May able as antipredatorbehavior. While uttering stereotyp- (seven days post-fledging) and 6 June. In 1995, we ical mobbing calls, he performed five of seven record- conducted66 hr of obAphelocoma tion about helping in Gray Jays was how to explain its coerulescens) of Oaxaca, Mexico. Auk 110:207- absence. Why would parents start treating their phil- 214. opatric offspring despoticallyat the onset of the breed- COCKBURN,A. 1996. Why do so many Australian ing season,following nearly a year of extendedparen- birds cooperate:social evolution in the Corvida?, tal care? Why would the parentsapparently not allow p. 451-472. In R. Floyd, A. Sheppard, and F! de their mature young to help feed the nestlingsas do the Barro [eds.], Frontiers of population ecology. parents of so many other speciesin seemingly similar CSIRO Publishing, Melbourne, Australia. EDWARDS,S. V., AND S. NAEEM. 1993. The phyloge- demographiccircumstances? netic componentof cooperativebreeding in perch- Unfortunately, any meaningful analysisof the inclu- ing birds. Am. Nat. 141:754-789. sive fitness consequencesof helping cannot be accom- EKMAN, J., AND B. ROSANDER.1992. Survival en- plished using our long-term data set; we have no way hancementthrough food sharing:a means for pa- of knowing whether mature offspring present on the rental control of natal dispersal.Theor. Pop. Biol. natal territory during the breeding seasonprior to 1994 42:117-129. actually helped their parents provision the fledglings. EKMAN, J., B. SKLEPKOVYCH,AND H. TEGELSTR~M. Now that helping is known to occur, our future work 1994. Offspring retention in the Siberian Jay Per- will be aimed at evaluating the indirect and direct fit- isoreus infaustus: the prolonged brood care hy- ness consequencesof helping. vothesis. Behav. Ecol. 5:245-253. Finally, we encourage other workers to pay in- HEIN’SOHN,R. G., A. COCKBURN,AND R. A. MULDER. creased attention to the post-fledging period. This pe- 1990. Avian cooperativebreeding: old hypotheses riod has been largely ignored in studiesof cooperative and new directions. Trends Ecol. Evol. 5:403- 407. breeding, which have usually been done on popula- KOENIG, W. D., E A. PITELKA,W. J. CARMEN,R. L. tions in which helping during the nestling stage is MUMME, AND M. T STANBACK. 1992. The evo- prevalent (Heinsohn et al. 1990). Our findings prompt lution of delayed dispersal in cooperative breed- the suggestionthat helping may be restricted to the ers. Quart. Rev. Biol. 67: 11l-150. post-fledging period in other semi-social species(e.g., LINDGREN,E 1975. Iakttagelser rorande lavskrikan Western Scrub-Jay,Aphelocoma californica, in south- (Perisoreus infaustus)-huvudsakligen dess hlck- em Mexico, Burt and Peterson 1993). If so, studies ningsbiologi. Fauna Flora 70: 198-210. focusing on this period should help elucidate condi- STRICKLAND,R. D. 1991. Juvenile dispersal in Gray tions favoring cooperative breeding in such species, Jays: dominant brood member expels siblings and should fill remaining gapsin our knowledge of the from natal territory. Can. J. Zool. 69:2935-2945. phylogenetic distribution of cooperativebreeding. STRICKLAND, D., AND H. OUELLET.1993. Gray Jay. In A. Poole, P Stettenheim, and E Gill [eds.], The We thank Dorice Hanes, Gary Hanes, Gord Lewer, birds of North America, No. 40. Academy of Nat- and Louis-Matthieu Strickland for helping-[near]-the- ural Sciences, Philadelphia and American Orni- nests; Kim Lundy and Julie Reider for helping-in-the- thologists’ Union, Washington, DC. lab; and Ron Mumme, Patricia Parker, and two anon- WAITE, T. A., AND R. C. YDENBERG. 1996. Foraging ymous reviewers for helpful comments on the manu- currencies and the load-size decision of scatter- script. hoarding Grey Jays. Anim. Behav. 51:903-916.