Department of Biology, Queen's University, Kingston, Ontario K7L 3N6, Canada Polyandry is usually defined as the mating of a female with more than one male during the course of a breeding season (Oring 1982, 1986; see Ford 1983 for a more restrictive definition). Oring (1986) rec- ognized two main types of polyandrous mating sys- tems, classical and cooperative polyandry. Both are rare among avian species, and particularly among pas- serines (Jenni 1974, Oring 1982) where most reported cases are based on limited and sometimes doubtful data (Ford 1983). Recent field studies of color-marked populations have nevertheless provided good evi- dence for the regular occurrence of cooperative poly- andry in the Dunnock (Prunella modularis; Burke et al. 1989), classical polyandry in the American Goldfinch (Carduelis tristis; Middleton 1988), and simultaneous polyandry-polygyny in Smith's Longspur (Calcarius pictus; J. V. Briskie pets. comm.). Isolated cases of se- quential polyandry have been reported also in the Northern Mockingbird (Mimus polyglottos; Fulk et al. 1987), the Black-capped Chickadee (Parus atricapillus; Waterman et al. 1989), and Geospiza finches (Boag and Grant 1984). We documented a case of polyandry in the Com- mon Redpoll (Carduelis f. fiammea), a species whose mating system has never been carefully studied. Us- ing Jeffreys' DNA fingerprinting probe 33.15 (Jeffreys et al. 1985a), we have confirmed that the two adults attending the first nest were the biological parents of the nestlings, and at the second nest we observed that the first male participated in neither mate guarding during nest construction nor mate provisioning dur- ing brooding. Apart from Middleton's (1988) Amer- ican Goldfinch study (and possibly Stokes' [1950] American Goldfinch population, see Middleton 1988: 299), there is no other report of polyandry in car- dueline species, even in well-studied taxa (e.g. New- ton 1973). We speculate that polyandry may be more prevalent in redpolls than our single observation in- dicates. Seutin observed redpolls in Churchill, Manitoba, 27 May to 1 August in 1988, and 1 June to 6 July in 1989. Nests were found usually by searching spruces (Picea balsamea) and stands of willows (Salix spp.). We mist-netted adults attending nests, banded them with standard USFWS metal bands and colored celluloid bands (A. C. Hughes, Middlesex, England), took a blood sample, and made standard morphometric mea- surements. To extract DNA from the blood samples of the fe- male involved in the polygamous relationship, her first mate, and the two nestlings from her first nest, we used an Applied Biosystems nucleic acid extractor as described by Seutin et al. (1991). Aliquots (4 ag) of redissolved DNA were cleaved with 15 units of the restriction enzymes AluI, HaeIII, and MboI for 3 h at 37øC in the appropriate buffers. Restriction fragments were separated on a 30 cm long 0.8% agarose TBE (0.09 M Tris, 0.09 M boric acid, 0.002 M Na2-EDTA, pH 8.0) gel running at 1.3 V/cm for 50 h. Restriction fragments resulting from the digestion of lambda DNA with HindlII were used as molecular size markers. Electrophoresed DNA fragments were depurinated, denatured, and neutralized in situ and then trans- ferred to a charge-modified blotting membrane (Gene Screen Plus(c); DuPont) as described by Seutin et al. (1991). The blot was dried, baked at 80C for 2 h, and incubated overnight at 65øC in a sealed plastic bag with 15 ml of prehybridization solution (Westneat et al. 1988). An aliquot (25 ng) of Jeffreys' 33.15 probe (Jeffreys et al. 1985a) was labeled with alpha 32P-dCTP in a random priming reaction to a specific activity of 1.1 x 109 dpm.ag , added to the prehybridization bag, and hybridization proceeded at 65øC for 12 h with shaking. The blot was then washed two times for 10 min in a 2x SSC, 0.1% SDS solution at room temper- ature, and two times for 30 min in a 2x SSC, 0.5% SDS solution at 65øC with constant shaking. Restric- tion fragments having hornology to the probe were visualized by autoradiography at -70øC for 6 days on a Cronex (c) film with one Cronex (c) intensifying screen. Behavioral observations.--On 18 July 1988, two Com- mon Redpoll nests were found ca. 200 m apart. The first nest (nest A) contained two large nestlings 8-9 days old. The second nest (nest B) consisted of only a few dry twigs of spruce and willow intermingled with pieces of dry grass; its construction had probably been started on the previous evening. No adult was seen close to either nest. The next day (19 July) at 0930, two adult birds were seen arriving close to nest B. The male (i.e. the bird that was not carrying lining material) had no red coloration on the breast and sides, and it was there- fore assumed to be a second year (SY) bird (Molau 1985, Seutin et al. 1989). The second bird was carrying ptarmigan feathers and was assumed to be female. She went to the nest, sat for approximately 10 min, and when she flew off, she was followed immediately by the male that was perched in a willow 1 m away. This behavior is typical of male carduelines involved in mate guarding (e.g. Newton 1973; G. Seutin pers. observ.). No other redpoll was seen nearby. Later the same morning, a male and a female were captured and color-banded at nest A. The female was captured as she left the nest when Seutin approached. Based on the shape of their central rectrices, we de- termined both birds to be more than two years old (ASY; Molau 1985); the male was a typical red-breast- ed bird. On the next visit, 21 July at 0600, nest A was empty. The adult male and the fledglings were seen neither that day nor during the following week (6 h spent within 500 m of the nest). At 0700, nest B was found unoccupied, completely lined and containing 1 egg. After 8 min of observation at nest B, the female that was color-marked at nest A arrived at the nesting bush carrying lining material; an unbanded SY bird si- multaneously arrived some 15 m away. The birds did not move significantly for 9 min, at which point the watch had to be terminated. On 27 July, at 1100, nest B contained a full clutch of 5 eggs. During a 1-h watch, the banded female did not leave the nest and was provisioned once by an unbanded SY bird. A 45-min watch on 31 July yielded one feeding of the brooding female by an unbanded SY bird. At the end of both watches, the female was flushed so that her identity as the female previously attending nest A was confirmed. Several other red- polls were observed on both days, but not the male that was color-banded at nest A. Seutin left Churchill on 1 August before the eggs in nest B hatched so we have no data on the fate of that nest. It was found empty in 1989. Some nests, known to have been abandoned in 1988, were found in 1989 still containing eggs, and nests from previous years were found in 1988 and 1989 containing whole eggs, eggshell fragments (a sign of clutch predation), or dead nestlings of all ages. Neither nest was used in 1989, and none of the 1988 adults or nestlings were seen in 1989. Assessment of parentage through DNA fingerprinting.- On the DNA autoradiograph (Fig. 1), it is possible to match every band present in an offspring with a band present in one of the adults. Two bands were consid- ered to match when they had the same electrophoretic position and similar autoradiographic intensity. As- suming the probabilit of band sharing between un- related individuals in the population to be 0.20 (see below), the probability that two unrelated individuals will share 8, 9, or 11 bands as did the adults and offspring (Table 1) is 2.4 x 10 6, 4.7 x 10 7, and 1.8 x 10 8 (Jeffreys et al. 1985b). These results indicate that both adults are related to the offspring and are probably their parents. The actual relatedness of in- dividuals can be estimated from the similarity be- tween their DNA fingerprints, using the statistic D = 2N^/(N^ + NB), where N^ and NB are the number of fragments scored in individuals A and B, and N^B is the number of fragments shared by both (Jeffreys et al. 1985b, Wetton et al. 1987). The overall similarity between the adult patterns is 0.20 (Table 1). This value is similar to that for the comparison of unrelated in- dividuals with the same or similar probes in humans and several avian species (Burke and Bruford 1987, Morton et al. 1990, Meng et al. 1990, P. T. Boag and B. N. White, unpubl. data). It is considered as the probability of band sharing between unrelated red- polls. D .... tive between adults and offspring ranged from 0.53 to 0.64 (Table 1). These values are very close to that expected for a parent-offspring pair, D = 0.58 (Jeffreys et al. 1985b, Burke and Bruford 1987). Be- cause we have no knowledge of the segregation pat- tern of the individual autoradiographic bands, some assumptions involved in the calculation of the ex- pected D-value might be violated (Lynch 1988). We feel, however, that if violations exist, they involve only a small proportion of the resolved bands and that our estimate is realistic. We conclude that the two adults caught near nest A are the biological par- ents of the nestlings. A genetic assessment of par- entage of the female's second brood could not be done because the male attending nest B could not be caught for blood sampling, and the clutch was not hatched when Seutin left Churchill. In their reviews of polyandry in birds, Jenni (1974) and Oring (1982) mentioned seven passerine species for which polyandry has been reported. Ford (1983) judiciously pointed out that most of these reports are based on equivocal data or are cases of mate switching between breeding attempts (see Ford [1983] for a dis- cussion of sequential monogamy with mate switching vs. sequential polyandry). Unequivocal evidence for the regular or occasional occurrence of polyandry in wild passerines has now been collected for a few species (see above). Our evidence for the occurrence of sequential poly- andry in the Common Redpoll is as follows. An ASY female was observed on the same day feeding almost fully grown nestlings at one nest (nest A) and build- ing a second nest (nest B) 200 m away. DNA finger- prints showed that the female and a color-banded ASY male were the biological parents of the offspring in nest A. That male was never observed close to nest B; instead, an unbanded SY bird accompanied the female both times she was observed bringing nesting material to nest B, and a similar bird (presumably the same putative mate) was later observed twice feeding the brooding female. Unfortunately, we did not have the DNA samples to evaluate the relative role of the two males in the female's second clutch. Whether or not the second male fathered the second clutch (in part or in whole), our observations indicate polyan- dry at least from a behavioral point of view. There is no previous report of polygamy in the Common Redpoll or in the closely related Hoary Red- poll (C. hornemanni). This is not surprising because only two color-marked populations of these species have been studied, and in both cases with limited M N1 N2 F  M N1 N2 F  M N1 N2 F kb 9.4 6.7 4.4 Fig. 1. DNA fingerprinting of a Common Redpoll family using Jeffreys' 33.15 probe. DNA samples (4 were digested with AluI, HaeIlI, and MboI. All bands present in an offspring (N1 and N2) are matched by a band of similar mobility and intensity in one or both parents (M: male; F: female). Some faint bands that could be scored on the autoradiographic film are not visible on this print. Poor resolution precluded scoring the lower bands in the MboI lanes. Lambda DNA digested with HindIII was used as molecular size marker; units of calibration are kilobases (kb). effort and for a limited time (Nystr6m and Nystr6m 1987; G. Seutin unpubl. data). Further, study of suc- cessive breeding attempts of individual redpolls is difficult because nests are usually not reused for sec- ond clutches (Jehl and Smith 1970, G. Seutin unpubl. data), and successive nests are probably built some distance apart. Circumstantial evidence indicates that individual female redpolls may even raise broods hundreds of kilometers apart during a single breed- ing season (e.g. Peiponen 1957). Middleton (1988) reported sequential polyandry in a population of the American Goldfinch as a regular alternative mating tactic present at low frequency in a predominantly monogamous population. His report is most interesting in the context of our observation in the Common Redpoll as the two species are con- geners, share several attributes of their breeding bi- ology, and in both cases polyandry was noticed in females with previous breeding experience. Middle- ton (1988) suggested that polyandry in goldfinches evolved from monogamy through increasing male parental care and female desertion (Oring 1986). His model requires a male-biased sex ratio in the popu- lation. As presented, Middleton's (1988) system strongly benefits females and second males. By switching mates, females increase their probability of successfully rear- ing a second brood if time is limiting, and second males gain the opportunity to leave progeny in a season where they were initially unpaired. First males have the advantage of mating with experienced fe- males, who in the goldfinch have a higher reproduc- tive success than younger females (Middleton 1979), but first males also risk being abandoned with the offspring. Middleton (1988) suggests that this risk is offset by the greater chance of survival of the male's genetic investment obtained through breeding with an experienced female. In redpolls, females have the capacity to double- clutch (Peiponen 1957, Troy and Shields 1979, this report). Because they face a short breeding season, there is a great advantage for them to start their sec- ond nest as early as possible, possibly through se- quential polyandry as presented by Middleton (1988). We have no strong evidence that a male-biased sex ratio exists in redpolls, that ASY females have a higher fertility than SY females, or that experienced males are able to singlehandedly rear young to indepen- dence; all these conditions are necessary in Middle- ton's model. Nevertheless, we feel these conditions are probably met--at least in some years in some pop- ulations--and consequently, Middleton's (1988) model will apply to redpolls as well as to goldfinches. It is hard to determine the frequency at which poly- andry might occur in redpolls because, in both our population and in the literature, there are no other data on the identity of the successive mates of an individual female during a breeding season. Indirect evidence suggests that polyandry might be more fre- TABLE 1. Number of bands scored (N) and percent- age of shared bands (D) in the DNA fingerprinting analysis.  AluI HaeIII MboI lative NM 18 19 20 57 NF 18 11 15 44 Nm 19 17 19 55 No2 18 15 19 52 DMF 0.17 0.20 0.23 0.20 Do  0.49 0.72 0.72 0.64 Do 2 0.61 0.71 0.62 0.64 Dm 0.59 0.50 0.47 0.53 DFo2 0.50 0.46 0.65 0.54 ß M: male; F: female; O1: offspring 1; 02: offspring 2. quent than our single record indicates. We observed single males caring for fledglings more frequently than we observed single females, a fact that might reflect females' desertion and involvement in second breeding attempts with other males. Of the family groups with adults that were observed or mist-netted in Churchill in 1988 and 1989, four were accompanied only by a male (all ASY males, 2 Common and 2 Hoary redpolls), one group by a female (a SY Common Red- poll), and two groups by both parents. This distri- bution is not statistically different from a uniform distribution (expected ratio 1:1:1; G = 1.83, P > 0.1), and it might have resulted from causes other than female desertion, but it is in the direction expected if polyandry occurs at a significant frequency. Fur- ther, the single males that attended fledglings were all ASY birds as expected in Middleton's (1988) mod- el. We feel that polyandry might be more prevalent in cardueline finches (including redpolls) than is gen- erally recognized, especially in species with short or unpredictable breeding seasons. That it has been re- ported in only two cardueline populations probably reflects the lack of within-season site fidelity of fe- males in several species (e.g. Newton 1973). We thank E. Guglich for the extraction of samples, J. Quinn for help with the DNA fingerprinting pro- tocol, and A. Jeffreys for the 33.15 probe. N. L. Ford, D. Lank, and an anonymous referee kindly reviewed earlier versions of the manuscript. This work was supported by Natural Sciences and Engineering Re- search Council of Canada grants to P. T. Boag, B. N. White, and L. M. Ratcliffe, and grants from the F. M. Chapman Memorial Fund, American Museum of Nat- ural History, the A. B. Kelly Memorial Fund, Province of Quebec Society for the Protection of Birds, and the Department of Indian and Northern Affairs, Ottawa. G. Seutin was supported by doctoral scholarships from the Natural Sciences and Engineering Research Council of Canada and from the Fonds pour la For- mation de chercheurs et l'aide h la recherche, Quebec. LITERATURE CITED BOAt, P. T., & P. R. GatqT. 1984. Darwin's finches (Geospiza) on Isla Daphne Major, Galapagos: breeding and feeding ecology in a climatically variable environment. Ecol. Monogr. 54: 463-489. BURKE, T., & W. M. BRUFORD. 1987. DNA finger- printing in birds. Nature 327: 149-152. --, N. B. DAVIES, W. M. BRUFORD, & B. J. HATCH- WELL. 1989. Parental care and mating behaviour of polyandrous Dunnocks Prunella modularis re- lated to paternity by DNA fingerprinting. Nature 338: 249-251. FORD, N.L. 1983. Variation in mate fidelity in mo- nogamous birds. Pp. 329-356 in Current orni- thology, vol. 1 (R. F. Johnston, Ed.). New York, Plenum Press. FULK, K. R., C. A. LOGAn, & L. E. HYATT. 1987. Poly- andry in a female Northern Mockingbird. Wilson Bull. 99: 286-288. JEFFREYS, A. J., V. WItSON, & S. L. THEIN. 1985a. In- dividual-specific "fingerprints" of human DNA. Nature 316: 76-79. , J. F. Y. BROOKFIELD, & R. SEMEONOFF. 1985b. Positive identification of an immigration test-case using human DNA fingerprints. Nature 317: 818- 819. JEHL, J. R., JR., & B. A. SMITH. 1970. Birds of the Churchill region, Manitoba. Manitoba Mus. Man Nat., Spec. Publ. No. 1. Winnipeg. JENNI, D. A. 1974. Evolution of polyandry in birds. Am. Zool. 14: 129-144. LYNCH, M. 1988. Estimation of relatedness by DNA fingerprinting. Mol. Biol. Evol. 5: 584-599. MEC, A., R. E. CARTER, & D. T. PAretiN. 1990. The variability of DNA fingerprints in three species of swan. Heredity 64: 73-80. MIDDLetON, A. L.A. 1979. Influence of age and hab- itat on reproduction by the American Goldfinch. Ecology 60: 418-432. ß 1988. Polyandry in the mating system of the American Goldfinch, Carduelis tristis. Can. J. Zool. 66: 296-299. MOLAU, U. 1985. Grfisiskkomplexet i Sverige. Vfir Ffigelvrld 44: 5-20. MORTON, E. S., L. FOP, MA, & M. BaUN. 1990. Ex- trapair fertilizations and the evolution of colonial breeding in Purple Martins. Auk 107: 275-283. NEWTON, I. 1973. Finches. New York, Taplinger Publ. Co., Inc. NYSTRM, B., & H. NYSTRM. 1987. Biotopval och hckning hos grfisiskor Carduelis fiammea och sni3siskor C. hornemanni i Ammarnsomrfidet, si3dra Lappland. Vr Ffigelvrld 46: 119-128. ORING, L. W. 1982. Avian mating systems. Pp. 1-92 in Avian biology, vol. 6 (D. S. Farner, J. R. King, and K. C. Parkes, Eds.). New York, Academic Press. --. 1986. Avian polyandry. Pp. 309-351 in Cur- rent ornithology, vol. 3 (R. F. Johnston, Ed.). New York, Plenum Press. PEIPONEN, V. 1957. Wechselt der birkenzeisig Car- duelis fiammea (L.), sein brutgebiet whrend des sommers? Ornis Fennica 34: 41-64. SEUTIN, G., P. T. BOA, & L. M. RATCtFFE. 1989. Plumage variability in the North American red- polls (Aves: Carduelinae): the importance of sex- ual and age dimorphisms. MuskOx 39:116-121. , B. N. WHITE, & P. T. BOA. 1991. Preservation of avian blood and tissue samples for DNA anal- yses. Can. J. Zool. 68: 82-90. STOKES, A.W. 1950. Breeding behavior of the gold- finch. Wilson Bull. 62: 107-127. TROY, D. M., & G. F. SHIELDS. 1979. Multiple nesting attempts by Alaskan redpolls. Condor 81: 96-97. WATERMAN, J., A. DESROCHERS, & S. HANNON. 1989. A case of polyandry in the Black-capped Chick- adee. Wilson Bull. 101: 351-353. WESTNEAT, D. F., W. A. NOON, H. K. REEVE, & C. F. AQUADRO. 1988. Improved hybridization con- ditions for DNA 'fingerprints' probed with M13. Nucleic Acids Res. 16: 4161. WETTON, J. H., R. E. CARFa, D. T. PAmO, & D. WAL- TERS. 1987. Demographic study of a wild House Sparrow population by DNA fingerprinting. Na- ture 327: 147-149. Received 2 February 1990, accepted 7 July 1990.