210

Short Communications

[Auk, Vol. 100'

Female-FemalePairs and PolygynousAssociationsin a Quebec Ring-billed Gull Colony

MARIE-CHRISTINE LAGRENADE 1 AND PIERRE MOUSSEAU

Centre recherches de dcologiques Montrdal,Universitd Montral,5858, de de Cte des Neiges, bureau 400, Montrdal,Qudbec H3C 3J7, Canada
Recently, female-femalepairs have been discovered in severalspeciesof Laridae: WesternGull (Larus occidentalis; Hunt and Hunt 1977), California Gull

(L. californicus; Conoveret al. 1979),Herring Gull (L.

argentatus; Fitch1979)and Ring-billed Gull (L. delawarensis; Ryder and Somppi 1979, Conover et al. 1979). These pairs usually attend superclutches (i.e. 5-8 eggs).Hunt andHunt (1977),however,havealso found female associations attending normal-sized clutches (1-3 eggs).Mate fidelity and site tenacityin female-female pairshavebeenfoundin Western Gulls (Hunt and Hunt 1977) and in Ring-billed Gulls (Kovacsand Ryder 1981).In view of thesefindings, our objectives thepresent in studywereto determinethe proportionsof female-female pairs and of trios, consisting of a male and two females,that were associated with superclutches and the breeding success, egg-fertilityrate, site tenacity,and mate fidelity of birds in these unusual mating systemsat a Ringbilled Gull colonyat Ile de la Couve, which is located in the Saint-Lawrence Seaway near Montreal

ite Island(1.9% out of 1,911nests; Ryderand Somppi 1979;P < 0.01)andin Washington (1.3%out of 1,669 nests;Conoveret al. 1979; P < 0.05) but is comparable to that in 21 Herring Gull colonies the Great of Lakes (0.3% out of 10,707nests;Shugart1980; P 0.05). All of theseabovepercentages much lower are
than those found in the Santa Barbara Western Gull

colony (10-15%, Hunt et al. 1980). In this latter species, however, contraryto the Ring-billed Gull, a 4-egg clutch is considereda superclutch and is associatedwith female-femalepairs. Four-egg clutcheswere attended predominantly (95%) by male-femalepairs (P < 0.001), suggesting the capacity thesefemalesto lay four eggsand/or of the occurrence egg dumping (Table 1). One feof male-female pair was alsocaptured from a four-egg nest. Five-egg clutcheswere attendedequally by male-femaleor female-femalepairs or one male-two
female associations(P > 0.05). We found that 83%

of six-eggclutches were attendedby female-female

pairs (P > 0.05).

(4528'30"N, 7330'30'vV). colonyhas been preThis viously described by Lagrenade and Mousseau
(1981a).

The finding of onemale-twofemaletrios attending superclutch,es the second is recordof a polygynous

associationin the Ring-billed Gull. Conover et al. (1979) reported the presenceof a male with three females attending a superclutch. Two of the seven one male-two female trios had a double-cup nest. The presenceof double-cup nests has been mentioned only by Shugart(1980)in polygynous Herring Gull trios and by Southern (1978) in a Ring-billed Gull pair. Of the 11 pairs of femalesbanded in 1980, three

Nestscontaining superclutches were markedwith flagsand visited at intervalsof 2-3 days,startingat the end of the laying period (April) of 1980and 1981.
The attendants of such nests were live-trapped in walk-in trapsplacedon the nest(Weaverand Kadlec 1970). Using Ryder's (1978a)discriminantfunction
on bill measurements, we sexed and then banded
and released them. To be sure that all attendants were

caught,eachnestwaswatched periodsof up to for

1 h per visit. We opened all unhatchedeggsto determinewhetheror not theywerefertile.Addledeggs were considered infertile.Therefore, fertility rate the

remained together 1981. in Twoof threenested within 3 m of their 1980 territory, while the other pair moved approximately m away from its 1980 ter40 ritory. Furthermore,two femalesfrom a 1980polygynous associationreturned to the same territory without the male and formed a female-female pair.. There is no significant difference between the number of eggs that disappeared (eaten or stolen) from nests of male-femalepairs and female-female pairs.In polygynous associations, however,therewas

of the eggs(percentage eggs of laid thatwerefertile)

is an underestimate.We define the hatching rate as the percentageof eggs laid that hatched and the

hatchingrate of fertile eggsasthe percentage ferof tile eggsthat hatched.All statistical testsareWilks's Chi-square (also called G-test; Sokal Rohlf1969). and In the 1980and 1981surveys the Ile de la Couve of colony,superclutches accounted lessthan0.14% for of the total nests(14,331).The percentage superof clutches theIle de la Couvecolony significantly in is lower than that in Ring-billed Gull colonies Granat

a significant increase the percentage egg loss. in of

Two-cup nestshad a higher egg-loss rate (0.36) than did one-cupnests(0.28) in thesetrios (P > 0.05). The egg-loss rate of the polygynous associations this of study (0.30) seemssimilar to the one observedfor female-female pairsby Ryder andSomppi(1979;0.25). The hatching of the eggsof female-female rate pairs and polygynous trios was significantly lower (P < 0.01) than that of ma!e-femalepairs (Table 2). This was partly due to the lower egg-fertilityrate (P <

Present address: Departmentof Renewable Resources, Macdonald Campusof McGill University,Ste. Anne-de-Bellevue, Qu4becH9X
1C0, Canada.

January 1983]

Short Communications

211

TABLE Percentage occurrence the different drastically 1. of of reducedsince1969,the pesticide hypothmatingsystems foundin superclutches the Ring- esis may not be applicableto the Ile de la Couvge at billed Gull colonyof Ile de la Couve (Quebec) colony. during the breeding seasons 1980and 1981. (2) Ryder (1978b) consideredwidowed fertilized femalesforming bonds as a possibleorigin of feClutch size male-femalepairs. Although the low egg fertility of female-femalepairs doesn't fully support this hy4 eggs 5 eggs 6 eggs pothesis, our observationof a 1980polygynous trio Mating system (n = 19) (n = 21) (n = 6) becominga female-female pair in 1981does.For feMale-female 94.74 ***' 47.62 0.00 male-femalepairs exhibiting mate tenacity, this hyOne malepothesiscould explain the origin, but this behavior two females 0.00 28.57 16.67 doesnot seemto have adaptive significance, these as Two females 5.26 23.81 83.33 femaleswould have better breeding success they if } n -- total nests. found a male mate instead of maintaining the bond ' *** -- significantly different from the occurrence other mating of systemsfor the same clutch size at P < 0.001. for a secondyear.
(3) Female-female pairs could be an adaptive responseto an excessof females in the colony (Hunt and Hunt 1977). After removing males from several Ring-billed Gull colonies, Conover (pers. comm.) found a higher frequencyof female-female pairs in
these colonies than in control colonies and in pre-

0.01), but the low hatchingrate of the fertile eggsof the unusual mating systemssuggests that poor parentalattentionand caremay alsohavebeen a factor. In as much as our egg-fertility rate is an underestimate, our data are much like thosereportedby Conover et al. (1979; 67%) and by Ryder and Somppi (1979; 69%). Hunt and Hunt's study(1977)onceagain standsout, with a very low 14% rate of fertile eggs in superclutches.

vious years.Conoveralsodiscovered that the breeding population of Ring-billed Gulls in Washington state is highly skewed. Thus, a female that fails to find a male mate may associate with another unmatedfemale.They coulddefenda territory, get ferby Three hypotheses havebeenproposed explain tilized promiscuously alreadymatedmales,build to the occurrence polygynousand female associa- a nest, and share incubation and chick-rearing duof ties. Such associations indispensableif birds are are tions in this normallymonogamous bird. single-parentgulls in colo(1) Toxic chemicalcomponents such as DDT and going to breed, because raise young (Ryder 1978b). its metabolitesinduce in male embryosof Western nies cannotsuccessfully and Californiagullshormonalfeminization,leading Knowing that female-female pairs, in this present with a to a suppression their breedingbehavior(Fry and study, predominantlyattend superclutches of low rate,thatsome exhibitmate Toone 1981). Such an explanationcould be the key relatively egg-fertility leads to abnormalpairs in Great Lakes'Herring Gull col- and sitetenacity,and thatsomearepromiscuous onieswherethese birdsarehighlycontaminated with us to support this third hypothesis. The infrequency of these unusual mating systems PCBsand otherorganochlorine pollutants (Gilmanet that the Ring-billed Gull shouldstill be conal. 1977). During the breeding seasonof 1979, we suggests species. superclutches As have examinedeggs,chicks,and adultsfor the presence sidereda monogamous in since1941(Moffitt 1942), of contaminantsand found that the gulls are rela- been observed this species pairs and polygynous tively free of toxicresidues (Mousseau Lagren- it is likely that female-female and ade 1980). As the use of DDT in Quebec has been associations have always existed. Ring-billed Gulls

TABLE Fertility,hatching, egg-loss 2. and ratesof thedifferent matingsystems observed theRing-billed at Gull colony the Ile de la Couve(Quebec) of duringthe breedingseasons 1980and 19812 of
Hatching
rate of

Fertility

Hatching

fertile

Egg-loss

Mating system
Male-female(n = 117) One male-two females(n = 36) Femlae-female(n = 64)
Number of fertile eggsper eggslaid. Number of hatchedeggsper eggslald. Number of hatched eggsper fertile eggs. Number of disappearedeggsper eggslaid.

rate '
0.85** 0.50 0.56

rate d
0.69** 0.33 0.38

eggs t
0.82 0.67 0.67

rate f
0.09 0.31'* 0.09

** - significantly differentfrom corresponding in other matingsystems P < 0.01. rate at n - total eggs,

212

Short Communications

[Auk, Vol. 100

& P. MOUSSEAU. 1981a. Re-

are opportunistic both feedinghabits (Lagrenade LAGRENADE, M.-C., in

and Mousseau 1981b) and habitat selection (Soots and Landin 1978). Therefore, such birds could have

production des Golands a bec cercl a l'ile de la Couve, Quebec (Canada). Naturaliste can.

108: 119-130. highly adaptablemating strategies. The origin of female-female pairs and polygynous trios is still un & -1981b. Alimentation des poussins de Golands bec cercl de l'ile de la Couclear,but mate fidelity in theseassociations seems to indicate an adaptive response an excess feto of ve, Quebec (Canada). Naturaliste can. 108: 131138. malesin the colony. We are grateful to Robert Pichet and Alfred La- MOFFITT, 1942. A nestingcolonyof Ring-billed J. Gulls in California. Condor 44: 105-107. grenadefor their field assistance and to Michael R. Conover for his helpful review of the manuscript. MOUSSEAU, P., & M.-C. LAGRENADE. 1980. Succbs This studywas supported the Centrede recherby de reproduction et contaminantspresents chez chescologiques Montrealand the salaries field de for le Gotland a bec cercl du sud-ouest du Quebec

assistance Summer Canada Projects. by

LITERATURE CITED

(Canada). Rapp. prpar par le Centre de rech. col. de Montreal pour le Serv. can.faune, Envir.
Canada.

CONOVER,M. R., D. E. MIrLER, & G. L. HUNT, JR.

1979. Female-female pairs and other unusual reproductive associations in Ring-billed and

RYDER, J.P. 1978a. SexingRing-billed Gulls externally. Bird-banding49: 218-222. 1978b. Possibleorigins and adaptive value

Californiagulls.Auk 96: 6-9. FitCH, M. 1979. Monogamypolygamy, and female-female pairs in Herring Gulls. Proc.Colonial Waterbird Group 3: 44-48.
FRY, C. M., & C. K. TOONE. 1981. DDT-induced

of female-female pairingin gulls.Proc.Colonial

Waterbird Group 2: 138-145. , & P. L. SOMPPI. 1979. Female-femalepair-

feminizationof gull embryos. Science 213:922924.

ing in Ring-billed Gulls. Auk 96: 1-5. SHUGART, W. 1980. Frequencyand distribution G. of polygyny GreatLakes in HerringGullsin 1978.
Condor 82: 426-429.

GILMAN A. P., G. A. Fox, D. B. PEAKALLS. M. TEEPLE, T. g. CARROLL G. T. HAYMES. 1977. &

SOKAL, R., & F. J. ROHLF. 1969. Biometry. San R.

Francisco Freeman.

opment and managementof avian habitat on levels of Great Lakes Herring Gulls. J. Wildl. dredged material islands. U.S. Army Engineer Mgmt. 41: 458-468. water ways experiment station. Dredged mateHUNT, G. L., JR., & M. W. HUNT. 1977. Femalerial research program. female pairing in WesternGulls (LarusoccidenSOUTHERN, E. 1978. Ring-billed Gull pair with W. talis) in southern California. Science 196: 14661467.

Reproductive parametersand egg contaminant

SOOTS,R. F., JR., & M. C. LANDIN.

1978. Devel-

two nests. Wilson

Bull. 90: 299-301.

J. C. WINGFIELDA. NEWMAN, & D. S. FARNER. 1980. Sex ratio of Western Gulls on Santa

WEAVER,D. K., & J. A. KADLEC. 1970. A method

for trappingbreedingadultgulls.Bird-banding
41: 28-31.

Barbara Island, California. Auk 97: 473-479. KovAcs, K. M., & J.P. RYDER. 1981. Nest-site te-

Received February 1 1982, accepted May 1982. 27

nacity and mate fidelity in female-female pairs of Ring-billed Gulls. Auk 98: 625-627.

A Record of the Siberian Flycatcher(Muscicapasibirica) from Bermuda:an ExtremeExtra-limitalVagrant

DAVID B. WINGATE

Government AquariumandMuseum,P.O. Box 145, Flatts3, Smith'sParish,Bermuda

On 28 September1980 at 1300I was checkingthe Somerset Long Bay Nature Reservein Sandy'sParish, Bermuda migratorybirds when I discovered for a medium-sized flycatcher feedingalongthe edgeof a grassy clearing. First impressions suggested an Empidonax flycatcher,but on closerapproachI was puzzled to observea generallydark plumage with extensive white spotting theupperpartsand dark on spottingor streakingon the underpartsforming a

partialbreastband. Anotherstrikingfeaturewasthe largesizeof the eyes,enhanced white eye rings. by This combinationof characters gave it the general appearance a juvenal-plumaged of EasternBluebird (Sialiasialis)exceptfor the lack of blue on wings and tail. The behavior however wasunquestionably that of a flycatcher. perched It conspicuously low dead on snagsoverhanging clearing,lookedaround in the flycatcher fashion,and made frequentflycatching