770 ShortCommunications [Auk,Vol. 104

ProlongedSperm StorageDuration in DomesticatedCanaries

T. R. BIRKHEAD ZoologyDepartment, The University, Sheffield S10 2TN, UnitedKingdom

Relative to most mammals, female can retain weekspreviously. The other 8 birdsin the aviary were their fertility for a considerabletime after mating and subsequentlypaired and confirmed to be females.Es- separationfrom the male (Lake 1975).Maximum sperm timated sperm storageduration was 35 days. storagedurations range from about 6 days in Ringed (4) A female canarywas paired to a male European Turtle-Doves (Streptopeliarisoria), 16 days in ducks, Goldfinch (C. ),and the male removed once and 35 days in chickens to 72 days in turkeys (see the clutch was complete.A single goldfinch mule was Birkhead 1987 for a review). No published informa- reared: this took 42 days (14 days of incubation and tion is available on the duration of sperm storagein 28 daysto rear the chick).The female canarywas then .Here I report maximum sperm storageval- paired immediately with a male siskin, and within a ues for female domesticated Common Canaries (Ser- week a clutch had been laid. One chick was reared inus canaria). from this clutch, but this was alsoa goldfinch rather I collecteddata from a questionnaireto breedersof than a siskin mule. Estimatedsperm storageduration canariesand "mules" (i.e. hybrids between a British was 49 days. male x canary female). Six independent in- (5 and 6) The situation was identical to that in (2) stanceswere reported; in each case female canaries above, except that after the second brood had been laid fertile eggsseveral days or weeksafter separation reared and the chicks removed, the female laid a third from the males. Routinely, some breeders separate clutch without being remated. The entire third clutch the male and female after the clutch has been laid, was fertile, but it is not known whether the eggs leaving the female to incubate and rear the young hatched. Estimated sperm storage duration was 35 alone (Carr 1959, Dodwell 1986). After the first brood days for the secondclutch and 68 days for the third had been reared,some females laid fertile eggswith- clutch. out being remated. Other instancesinvolved casesof Although theseindividual values are conservative, unexpectedpaternity when femaleshad been paired they are likely to lie near the extreme for canaries. with males of two different finch speciesin succes- Nonetheless,the maximum value, 68 days, is similar sion. In each case described below I give the most to the maximum record for any species:72 days conservativeestimate of the duration of sperm stor- in domesticatedturkeys (Lorenz 1950) and 60 days in age. Grey-facedPetrels (Pterodroma macroptera, Imber 1976; (1) A female canary was paired to a male Eurasian see Perrins and Brooke 1976, Hatch 1987 for infor- Linnet (Cardueliscannabina), but the linnet died before mation for other Procellariiformes). any eggs were laid. Approximately 3 weeks later the Information on the duration of sperm storage is female canarywas paired to a male Common essentialif one is to calculateaccurately the period (C.fiammea); they built a nestand a clutch was started over which a female bird might be fertilized. In par- after 7 days. Of the clutch of 4 eggs, 2 were fertile. ticular, it is important to know the timing and du- When the chicks fledged they were clearly linnet ration of this period to determine whether extrapair rather than redpoll mules, i.e. the linnet had fertilized copulationsoccur when they could fertilize eggs.Sev- the eggs,not the redpoll. Sperm storageduration was eral workers, including me (Birkhead 1982), have as- estimatedat 28 days. sumed, in the absenceof data, that sperm storage (2) A male and female canary were separatedafter duration in passerinesis considerably less than the egg laying and the female left to incubate(13-14 days) values presented here suggest. and rear the young alone. After the chicks were re- The adaptive significanceof the prolonged sperm moved (aged 21 days), the female laid a secondclutch storage that occurs in some birds is poorly under- "within days";one egg was fertile and subsequently stood, but it may be important in speciesin which hatched. Estimated sperm storage duration was 35 females copulate infrequently but produce large days. clutches (e.g. some Galliformes; see Birkhead et al. (3) Ten canaries,all thought to be female, were 1987)and in seabirds,where the femalemay be away maintained in an aviary over winter. In early Feb- from the colony and her partner for extended periods ruary one bird (obviouslya male) was heard singing before laying (see Imber 1976, Hatch 1983). There is and removed. Four weeks later one of the females no compellingreason why canariesshould have pro- was paired to a male Eurasian Siskin (C. )in a longed sperm storage.Wild canariesproduce clutches separatecage; this pair built a nest and started to lay of 4-5 eggs (Bannerman and Bannerman 1968), and within a week. From the clutchof 5 eggs3 were fertile the male and the female are unlikely to be apart for and proved to be canaries,not siskin mules. The fe- long periods.In all other carduelinefinches that have male could have been fertilized only by the male been studied, and in domesticated canaries, mate canary, which had been removed from the aviary 4 guarding is well developed; the male remains with October1987] ShortCommunications 771 the female continuouslyuntil she startsto incubate --, L. ATKIN,& A. P. MOLLER.1987. Copulation (Tsuneki 1961;Newton 1972,pers. comm.). One pos- behaviour in birds. Behaviour 101: 101-138. sible explanation for the results on canariesis that CARR,V. A.V. 1959. Mule and hybrid birds.London, selectivebreeding during the periodof domestication Cage Birds. (ca.350 yr; Dodwell 1986)has produced incidentally CLAYTON,G.A. 1972. Effectsof selectionon repro- a well-developedsperm storage ability (Clayton 1972, duction in avian species.J. Reprod. Fert. Suppl. Lake 1975). This idea is plausible becausebreeders 15: 1-21. have undoubtedly selectedfor individuals that bred DODWELL,G.T. 1986. The completebook of canaries. successfullyin captivity, and prolongedsperm stor- London, Merehurst Press. age contributesto successfulbreeding. In addition, HATCrt,S. A. 1983. Mechanismand ecologicalsig- spermstorage durations differ significantlybetween nificanceof spermstorage in the Northern Ful- different strains of chickens selected for different traits mar with reference to its occurrence in other birds. (Tanejaand Gowe 1961), indicating that sperm storage Auk 100: 593-600. duration hasa geneticbasis. Another explanationfor 1987. Copulationand mate guardingin the theseresults is that they representa few very rare Northern Fulmar. Auk 104: 450-461. occurrences at the extreme end of the distribution of IMBER,M. J. 1976. Breeding biology of the Grey- spermstorage duration in canaries.A systematicstudy facedPetrel Pterodromamacroptera gouldi. Ibis ! ! 8: of sperm storagein canariesand other passerines, 51-64. particularlywild species,might be interesting. LAKE,P. E. 1975. Gamete production and the fertile I am extremelygrateful to the individuals who pro- period with particular reference to domesticated videdinformation on which thisnote is based: J. May- birds. Syrup. Zool. Soc. London 35: 225-244. lan, R. Moat, T. Roberts,J. Robson,K. Swarm, and A. LORENZ,F.W. 1950. Onsetand duration of fertility C. Wyatt. I also thank Drs. M. de L. Brooke, K. M. in turkeys.Poultry Sci. 29: 20-26. Cheng,A. P. Moller, and I. Newton for commenting NEWTON, I. 1972. . London, Collins. on the manuscript. PERRINS, C. M., & M. DE L. BROOKEß 1976. Manx Shearwatersin the Bay of Biscay.Bird Study 23: LITERATURE CITED 295-299. TANEJA,G. C., & R. S. GOWE.1961. Effectof varying BANNERMAN,D. A., & W. M. BANNERMAN. 1968. Birds dosesof undiluted semen on fertility in the do- of the Atlantic islands.Edinburgh. mestic fowl. Nature London 191: 828-829. BIRKHEAD,T. g. 1982. Timing and duration of mate TSUNEKI,K. 1961. Territory in flocksof the caged guardingin MagpiesPica pica. Anita. Behav.30: canaries,with special reference to its causation 277-283. and defence.Japanese J. Ecol. 11: 19-26. ß 1987. Behaviouralaspects of sperm compe- tition in birds. Adv. Stud. Behav. in press. Received28 January1987, accepted 22 April 1987.

Energetic Consequencesof Sexual Size Dimorphism in White Ibises (Eudoclmus albus)

KEITH L. BILDSTEIN Departmentof Biology,Winthrop College, Rock Hill, SouthCarolina 29733 USA, and Belle W. BaruchInstitute for Marine Biology,University of SouthCarolina, Columbia, South Carolina 29208 USA

Sexual dimorphism is pronounced in White Ibises males and females intertwine their bills and necks (Eudocimusalbus): males are heavier than females,and during pair formation(Palmer 1962, Rudegeair 1975), they have substantiallylonger tarsi, bills, and wing and during copulation,when malesextend their bills chords(Kushlan 1977a).Sexual size dimorphism may over femalesand engagein twig pulling (Rudegeair result either from natural selectionacting to reduce 1975).A larger bill and body in malesundoubtedly intersexualfood competition or to increaseclutch size is also important for intrasexual fighting during in females (cf. Carothers 1984), or from sexual selec- courtship,mating, and nestbuilding, when malesfight tion acting to enhance the mating successof appro- over matesand attemptnest-site takeovers (Rudegeair priately sized individuals. Kushlan (1977a)proposed 1975,Frederick 1986). Fighting ability in maleshas that the size dimorphismhe reportedfor White Ibises alsobeen linked to increasedextrapair copulations in resulted from sexualselection acting to increasebody this sometimespromiscuous species (Frederick 1985a). size and bill length in males. Although female ibisesparticipate in aggressivein- Male ibisesuse their bills during courtship,when teractionsat the colonysite, female aggression is con-