Anim. Behav., 1996, 52, 405–417

Paternity and paternity assurance behaviour in the bluethroat,Luscinia s. svecica

CHRISTIN KROKENE*, KRISTIN ANTHONISEN*, JAN T. LIFJELD* &TROND AMUNDSEN†

*Zoological Museum, University of Oslo†Department of Zoology, University of Trondheim

(Received 30 May 1995; initial acceptance 23 August 1995;final acceptance 12 December 1995; MS. number: 4940 )

Abstract. Genetic parentage was examined in the territorial and predominantly monogamous blue-throat, by means of multilocus DNA fingerprinting. Thirty-five per cent (11/31) of nests contained oneor more offspring sired by extra-pair males, in total 20% (30/150) of all offspring. The frequency ofextra-pair paternity was similar in the two study years. All offspring were genetically related to theirputative mother, thus no evidence for intraspecific brood parasitism was found. Extra-pair offspringwere not randomly distributed, but were clustered in certain broods, in which on average 54% of theyoung were sired by extra-pair males. Ten copulations were seen, one of which was an extra-paircopulation. Males guarded their mate intensely around the start of egg laying, staying within 1 m of herfor more than 60% of the time. The male typically followed the female, and not vice versa. Hence, mateguarding was not a cooperative behaviour of the two sexes. Male singing activity declined graduallyduring the breeding season without reaching a peak during the females’ fertile period. Male bluethroatsthus apparently use mate guarding and not song as their primary paternity guard. For a subset of pairs(N=8) the observed behaviour could be related to paternity. These data indicated that, during the fertileperiod, there were more intrusions in territories with extra-pair offspring than in territories with onlylegitimate offspring. There was also a non-significant tendency for males with full paternity in their nestto mate guard more intensely than males with reduced paternity.

? 1996 The Association for the Study of Animal Behaviour

Extra-pair paternity is common in birds but variesacross species (Birkhead & Møller 1992). Inpasserines, reported frequencies range from 0% insome species (e.g. Gyllensten et al. 1990), to morethan 50% in others (e.g. Dunn et al. 1994; Mulderet al. 1994). Although several behavioural, socialand ecological factors that might influence theoccurrence of extra-pair paternity have been sug-gested (for reviews, see Westneat et al. 1990;Birkhead & Møller 1992; Stutchbury & Morton1995), the diversity of paternity patterns in birds isstill poorly understood. First, knowledge of thebasic behaviour underlying paternity patterns isoften incomplete. For example, extra-pair pater-nity may not always arise through extra-pair

copulation. Rapid mate switching and matereplacement may have been overlooked as sourcesof extra-pair paternity in several species (Birkheadet al. 1990; Birkhead & Møller 1992; Oring et al.1992; Pinxten et al. 1993). Second, when extra-pair paternity does result from extra-pair copu-lation, it is essential to separate the effects ofmale–male competition and female choice in orderto understand the selective factors involved(Lifjeld et al. 1994). Correlates of extra-pair pater-nity may be radically different in species withmale-driven and female-driven sperm competition(Lifjeld et al. 1994). The role of the female ininfluencing paternity patterns is largely unknown,although some evidence exists for a few species(Kempenaers et al. 1992; Lifjeld & Robertson1992; Westneat 1992). Thus, studies of avianpaternity should be accompanied by detailedbehavioural studies to reveal the role of the twosexes.

Correspondence: C. Krokene, Zoological Museum,University of Oslo, Sars gate 1, N-0562 Oslo, Norway(email: christin.evensen@toyen.uio.no). T. Amundsenis at the Department of Zoology, University ofTrondheim, N-7055 Dragvoll, Norway.

0003–3472/96/020405+13 $18.00/0 ? 1996 The Association for the Study of Animal Behaviour

405

Sperm competition has led to the evolutionof various anti-cuckoldry strategies in males(Birkhead &Møller 1992). The two most commonpaternity guards in birds are frequent pair copu-lations and mate guarding by close following,which seem to be alternative strategies (Møller &Birkhead 1991). Mate guarding appears to be theprimary paternity guard in territorial passerines.Møller (1991a) also suggested that male songcould function as a paternity assurance behaviourin passerines (the fertility-announcement hypoth-esis). According to that hypothesis, males sing toannounce both their mate’s fertility and their ownwilingness to defend their mate from intrudersseeking extra-pair copulations. In support of thehypothesis, Møller (1991a) reported that, in manyspecies, male singing activity reaches a peakduring the female’s fertile period.We studied genetic parentage in the blue-

throat, a territorial and essentially monogamouspasserine. We investigated whether males usemate guarding or song as a paternity guard, andwhether females promote sperm competition. Fora subset of the data we also related behaviourdirectly to paternity.

METHODS

Field Methods

The study was carried out in Øvre Heimdalen(61)25*N, 8)52*E), southern Norway, at about1100 m above sea level in the sub-alpine veg-etation zone. Further details of the study area canbe found in Vik (1978).We collected blood samples for DNA finger-

printing analysis in 1991 (15 broods) and 1992(16 broods). Blood (approximately 25 ìl fromnestlings and 50 ìl from adults) was collectedin capillary tubes from the brachial vein andimmediately suspended in 1 ml of ‘Queen’s lysisbuffer’ (Seutin et al. 1991). The samples werestored at ambient temperature for a few weeksin the field and then in a refrigerator (4)C) for3–8 months before being analysed in thelaboratory.In 1991, we located most nests during the

nestling period, whereas all but two nests in 1992were found during nest building. For thosenests found after completion of egg laying, weestimated the laying date of the first egg byback-dating, assuming an incubation period of

13 days and 24-h egg-laying intervals (personalobservations). We estimated the age of broodsfound after hatching from growth curves of youngof known age (Rangbru 1994). Adults were agedas second year or older according to Svensson(1984). Bluethroats breed rather synchronously(for comparison with other species, see Stutchbury& Morton 1995).We studied the birds’ behaviour in 1992 only,

when we followed them throughout the breedingseason. Males were caught in mist nets soon afterterritory establishment. We used playback of malebluethroat song to attract territorial males to thenets. Some females were caught in mist netsduring nest building, but the majority were notcaptured until the eggs had hatched in order toreduce the risk of nest desertion. Males andfemales were marked with a numbered aluminiumring and a unique combination of three colouredleg rings. We used similar colours on all males toavoid any potential bias in the data arising fromfemale preferences for particular colours (cf.Burley et al. 1982). Although the majority offemales were not ringed until after the eggs hadhatched, we could identify them by their plumage.At capture we also recorded the birds’ winglength, tarsus length and body mass.Mate-guarding behaviour was studied inten-

sively for nine breeding pairs. We observed eachpair for 20 min between 0500 and 1100 hourseach day, from the time of pair formation until 3days after clutch completion. We defined pairformation as the first day the male was observedtogether with his mate. All pairs were observeddaily by one of two observers (C.K. and K.A.)using a rotating schedule to avoid any pair beingobserved at the same time every morning. Anobservation bout of 20 min started whenever onemember of the pair was located within theterritory. The distance between the pair memberswas recorded every second min. Distances wererecorded as: <1 m, 1–5 m, 6–10 m, 11–20 m or>20 m. We also recorded whether one or bothpartners were out of sight. During each 2-mininterval we recorded if either member of the pairflew 5–20 m or more than 20 m away, and whetherit was followed by its mate within 10 s. Becausefew flights of more than 20 m were recorded, thetwo variables were later combined for analysis,resulting in one variable ‘flights longer than 5 m’.Intrusions and copulations were also noted. Inaddition, we recorded whether the male sang

Animal Behaviour, 52, 2406

during each 2-min interval and we counted thenumber of song flights. Singing activity wasexpressed as the number of 2-min intervals withsong during a 20-min bout. All vocalizationssimilar to the typical territorial song wererecorded, but not presumed contact or warningcalls. Before pair formation and after they hadbeen colour-marked, males were observed for20 min each day to record singing activity andsong flights.For analysis of possible temporal changes in

behaviour, we defined the fertile period as begin-ning 3 days before the first egg was laid (cf.Björklund & Westman 1983; Carlson et al. 1985),and to end the morning the penultimate egg waslaid (Johnson 1986; Birkhead & Møller 1992).Observations on the latter day were included inthe fertile period, as bluethroats often laid theireggs after the watches were terminated.

DNA Fingerprinting Methods

A volume of 500 ìl of the blood suspension wasincubated overnight at 37)C on a rocking traywith 100 ìl proteinase K (10 mg/ml) and 100 ìl10% SDS, washed once with an equal volume of a1:1 mixture of phenol and chloroform and oncewith an equal volume of chloroform, and theDNA subsequently precipitated with isopropanol.The DNA was collected on a fine-tipped glass rod,washed in 70% ethanol, air-dried and resuspendedin 500 ìl 1#TE-buffer (0.01 TRIS, 0.001

EDTA, pH 7.5). A volume corresponding toapproximately 20 ìg DNA was then digestedovernight at 37)C with 50 units of the restrictionenzyme HaeIII. The digested DNA was precipi-tated with 2.5 volumes cold absolute ethanol andcentrifuged for 30 min. The precipitate waswashed with 70% ethanol, dried and resuspendedin 30 ìl TE-buffer. The DNA concentration wasestimated by running the samples with standardsof known concentration on a test gel.Approximately 3–5 ìg of DNA was loaded on

to a 20#40 cm 0.8% agarose gel in 1#TAE-buffer with a molecular size marker (lambdaphage DNA digested with HindIII) in the outer-most lanes. Whenever there was enough digestedDNA from the putative parents they were loadedtwice on the gel (on both sides of their nestlings)to facilitate comparisons across lanes. The gelwas electrophoresed at 1.2 volts/cm with buffer-circulation for approximately 45 h.

Gels were washed in 0.25 HCl for 10–15 min,in a denaturing solution (0.4 NaOH, 1.5

NaCl) for 60–75 min and in a neutralizingbuffer (0.5 Tris-Cl, 1.5 NaCl, pH 7.5) for60 min. The DNA was then Southern blottedon to Immobilon-N (Millipore) or Hybond-N(Amersham) hybridization transfer membranes.The membranes were hybridized with the mini-satellite probe 33.15 (Jeffreys et al. 1985) accord-ing to the protocol described in Stacy et al. (1991)and Stacy & Jakobsen (1993), and then exposed toKodak X-OMAT film at "80)C for 2–14 dayswith or without intensifying screens.

Scoring the Fingerprints

Clearly resolved bands (>2–4 kb) on the auto-radiographs were marked on an acetate sheetoverlay with permanent markers, using differentcolours for paternally and maternally derivedbands. Bands were considered equal if they hadmigrated so that their centres were less than0.5 mm apart, and if they did not differ greatly instrength. When lanes differed markedly in theamount of DNA, faint bands in the stronger laneswere ignored. Nestling bands that were notpresent in the fingerprints of either parent werescored as novel fragments.The degree of band sharing between two

individuals was expressed by the band sharingcoefficient D, which was calculated as 2NAB/(NA+NB), where NAB is the number of bandsshared between birds A and B, and NA and NB arethe number of bands in birds A and B, respect-ively (Wetton et al. 1987). The coefficient rangesfrom 0 (no bands shared) to 1 (all bands shared).The fingerprints of the families from 1991 werescored by J.T.L. Those from 1992 were scoredindependently by C.K. and K.A. Band sharingcoefficients obtained by the latter scorers werehighly significantly correlated (mother–offspring:Pearson correlation, r=0.76, N=83, P=0.0001;father–offspring: Spearman rank correlation,rS=0.87, N=83, P=0.0001), as were the numberof bands not found in either parent, i.e. novelfragments (rS=0.70, N=83, P=0.0001), and there-fore the mean values for the two scorers wereused. We used the number of novel fragments andband sharing coefficients to determine parentage.

Statistics

For statistical analysis, parametric tests wereused on normally distributed variables. Otherwise

Krokene et al.: Paternity in the bluethroat 407

non-parametric tests were used. All test resultsreport two-tailed probabilities.

RESULTS

On average, there were 29.9&9.9 (X&;range=7–51) scorable bands per fingerprint(N=210). We analysed 150 offspring, of which 65had a fingerprint profile that completely matchedthose of their putative parents. Their band sharingcoefficients with the putative mother averaged0.59&0.01 (X&; range=0.40–0.71), and withthe putative father 0.61&0.01 (range=0.44–0.78).Thus, all D-values were 0.40 or higher, and wetherefore consider 0.40 to be the lower limit for

band sharing with a genetic parent. The numberof novel fragments per offspring showed a discon-tinuous and bimodal distribution, with offspringhaving either three or fewer, or five or more novelfragments (Fig. 1). All offspring with up to threenovel fragments (N=55) had a band sharingvalue of 0.40 or higher with both putative parents(Fig. 1; mother–offspring: D=0.63&0.01, range=0.42–0.82, and father–offspring: D=0.64&0.01,range=0.48–0.80). Hence, we conclude that theywere all related to their putative parents. Off-spring with more than three novel fragments(N=30) shared a similar proportion of bands withtheir putative mother (D=0.56&0.02, range=0.42–0.69), as did those with fewer novel frag-ments, suggesting that they were all related to

24

0.9

00

Ban

d sh

arin

g co

effi

cien

t (D

)

14

0.4

0.8

0.7

0.6

0.5

0.3

0.2

0.1

6 8 10 12 16 18 20 2242

(a)

24

0.9

00

Novel fragments14

0.4

0.8

0.7

0.6

0.5

0.3

0.2

0.1

6 8 10 12 16 18 20 2242

(b)

Figure 1. The relationship between band sharing coefficients with (a) the putative mother and (b) the putative fatherand the number of novel fragments in the offspring fingerprint (N=150). The dashed lines indicate the criteria forexcluding parentage.

Animal Behaviour, 52, 2408

their putative mother. However, band sharingwith the putative father was low (D=0.24&0.02,range=0.09–0.46) and comparable to that for pairmates (D=0.29&0.02, range=0.14–0.52, N=31).With only three exceptions (Fig. 1) the bandsharing values with the putative father were belowthe critical limit of 0.40, and we conclude thatthese offspring were unrelated to the putativefather.The three ambiguous offspring fulfil one cri-

terion for paternity exclusion (i.e. more than threenovel fragments) but not the other (D-values§0.40). To assess the probability that these off-spring were related to both putative parents weneed to know the likelihood that all novel frag-ments in these offspring (i.e. 5.5, 10 and 11,respectively) arose from mutations. Minisatellitefragments have a mutation rate of about one outof 100–300 meiotic events (Jeffreys et al. 1985;Burke & Bruford 1987; Westneat 1990; Lifjeld etal. 1993). Calculating a true mutation rate is notstraightforward here as the number of novel frag-ments is subject to scoring error and there weresome discrepancies between the scorers. However,if we assume that all 120 offspring with up to threenovel fragments were related to their putativeparents and that 55 of these had at least one novelfragment each (i.e. including all offspring witha score of 0.5 novel fragment), then the truemutation rate per individual must be lower than0.458 (=55/120). If we conservatively use thisvalue as the mutation rate, the probability of nmutations in one particular individual is 0.458n

(Westneat 1993a). Hence, in our sample of 150offspring we would expect to find fewer than 2,0.06 and 0.03 individuals with 5.5, 10 and 11mutations, respectively. It is therefore reasonableto conclude that the 10 and 11 novel fragmentswere caused by mismatched parentage rather thanmutations. Since both offspring had high bandsharing with the putative mother (D§0.60), weconclude that both were related to the mother butnot to the putative father.The offspring with 5.5 novel fragments is still a

border-line case, but it seems more likely that itwas unrelated to the putative father. This isbecause the band sharing with the putative fatherwas only at the border of exclusion, whereas thenumber of novel fragments was within the rangeof exclusion. The frequency distribution of novelfragments was discontinuous around four frag-ments (Fig. 1), whereas the band sharing values

showed a more continuous distribution. Somepair mates, that is, putatively unrelated individ-uals, even had band sharing values above 0.40.This suggests that the criterion of novel fragmentshas a better discrimination power than theband sharing limit. Thus, the most parsimonioussolution in this case is to consider the offspringunrelated to the putative father. In sum, we con-clude that all offspring with more than three novelfragments resulted from extra-pair paternity, andthat there were no cases of intraspecific broodparasitism.Overall, 35% (11/31) of the nests contained

offspring sired by extra-pair males, 20% (30/150)of all offspring analysed. In 1991, extra-pair pater-nity accounted for 27% (4/15) of nests and 19%(13/67) of nestlings. In 1992, the correspondingvalues were 44% (7/16) for nests and 20% (17/83)for nestlings. The proportion of nests with extra-pair offspring did not differ between years(÷2=0.18, df=1, P=0.54). There was one poly-gynous male in each of the 2 study years. Thepolygynous male in 1991 lost paternity in hissecondary nest (one of four young), but not in hisprimary nest. The polygynous male in 1992 hadfull paternity in his two nests.There were relatively many broods with a high

percentage of extra-pair offspring (Fig. 2). Aspotential benefits of extra-pair paternity tofemales predict differential patterns for the distri-bution of extra-pair paternity among broods(see Birkhead et al. 1990; Westneat et al. 1990;Kempenaers & Dhondt 1993; Lifjeld et al. 1993),we wanted to test whether the extra-pair offspringwere randomly distributed across the 31 broods.Using the method described in Lifjeld et al.(1993), and assuming that each offspring had a

25

00

Nu

mbe

r of

nes

ts

1–25

Extra-pair offspring (%)

26–50 51–75

20

15

10

5

76–100

Figure 2. Percentage of extra-pair offspring in 31 blue-throat broods.

Krokene et al.: Paternity in the bluethroat 409

probability of 0.20 of being sired extra-pair, wecalculated the expected number of broods withone or more extra-pair offspring to be 19.95. Thisvalue differed significantly from the observednumber of 11 broods (÷2=11.27, df=1, P<0.001).If we assumed that the two nests of a polygynousmale were not independent observations, and thusshould be counted as one, the observed number ofbroods with extra-pair paternity was still signifi-cantly lower than expected (÷2=10.12, df=1,P<0.01). We therefore conclude that the extra-pair offspring were not randomly distributed, butwere over-represented in particular nests. Thesenests had on average 54% of the young sired byextra-pair males.We did not try to identify the genetic fathers of

all extra-pair offspring as the pool of potentialmales was considerable. Nevertheless, on mostgels we ran a few lanes with DNA from nearestneighbours or males observed as intruders. Wewere able to assign paternity for only oneextra-pair offspring. In that case the male was aneighbour with uncertain mating status.Cuckolded and non-cuckolded males did not

differ with respect to age (÷2=0.03, df=1,P=0.87), body mass (t="0.73, df=29, P=0.47),tarsus length (t="1.41, df=29, P=0.17), con-dition (residuals from regression of body mass ontarsus length; t="0.44, df=29, P=0.65) or winglength (t=0.13, df=29, P=0.90). Neither wasthere any difference in the timing of egg laying(t=0.23, df=29, P=0.82; egg laying standardizedaccording to the median egg-laying date for eachyear).

Territorial Intrusions and Copulations

We observed 31 intrusions and all observedterritorial intruders were males. Intrusion rate wasexpressed as the number of intrusions per 20 min.The intrusion rate differed significantly betweenthe pre-fertile, fertile and post-fertile periods(repeated measures ANOVA: F2,8=5.24, P=0.02),with the highest intrusion rates recorded duringthe fertile period (Fig. 3). There was a significantcontrast in intrusion rate between the pre-fertileand the fertile period (t="2.39, df=8, P=0.029)as well as between the fertile and the post-fertileperiod (t=3.08, P=0.007).In some cases intruders were identified as

neighbouring males (N=4) or as non-neighbours(N=3), but in most cases they were unmarked

individuals (N=9) or not successfully identified(N=15). The intruders either sang or perchedsilently in the territory. Twenty-six (84%) of allintrusions occurred during the females’ fertileperiod, and in 14 of these cases, the intruder waschased away by the resident male. In three cases,the territory owner did not react to the intruder,apparently because the owner did not detect him.In the remaining nine cases, the territory ownerwas out of sight.We saw 10 copulations, one of which was an

extra-pair copulation apparently initiated by theintruding male. The extra-pair copulation tookplace 2 days before the female laid her first egg(i.e. on day "2). Whether the copulation led tocloacal contact could not be determined withcertainty. The DNA analyses revealed that thenest contained no illegitimate offspring. The ninewithin-pair copulations occurred from day "6 today 1, with no apparent peak in frequency.

Mate Guarding

If males guard their partners in order to protecttheir paternity, we should expect more intensemate guarding during the females’ fertile periodthan in the pre- and post-fertile periods. Twopredictions concerning male behaviour can bederived from the mate-guarding hypothesis: (1)the male should stay closer to his mate, and (2) ahigher proportion of female movements should befollowed by the male during the female’s fertile

0.5

0Pre-fertile

Intr

usi

ons

per

20 m

in

Fertile

0.4

0.3

0.2

0.1

Post-fertile

Figure 3. Mean number of intrusions per 20 min innine bluethroat territories during the pre-fertile, fertileand post-fertile periods of the resident females. Barsindicate 1 .

Animal Behaviour, 52, 2410

period than during the pre- and post-fertileperiods.To test the first prediction, we examined dis-

tances between mates in the pre-fertile, fertile andpost-fertile periods. In many cases either the male(especially in the post-fertile period) or the female(especially in the pre-fertile period) was out ofsight. This posed a methodological problem inanalysing intra-pair distances, because a birdrecorded as out of sight might have been with itsmate but hidden in the vegetation. We thereforerearranged the data into only two distance cat-egories (‘<1 m’ and ‘>1 m’), assuming that theintra-pair distance was >1 m whenever one of thebirds was out of sight. The time pair mates stayedcloser than 1 m varied significantly between theperiods, with a distinct peak in the fertile period(repeated measures ANOVA: F2,8=6.84, P=0.009). The increase from the pre-fertile to thefertile period in time spent together was significant(t="3.26, df=8, P=0.006), as was also thedecrease from the fertile to the post-fertile period(t=3.03, df=8, P=0.009). The intensity of mateguarding was at its highest from day "3 to day 1(Fig. 4), when the pair mates were closer than 1 mfor an average of 47% of the time. On days "2and "1 this value exceeded 60%.To test the second prediction of the mate-

guarding hypothesis, we examined the number oftimes females flew off and their mates followedthem during the pre-fertile and fertile periods.

Males followed their mates more frequently dur-ing the fertile period than during the pre-fertileperiod (Wilcoxon signed-rank test on proportionsof followings: Z="2.20, N=6, P=0.03; Fig. 5a).Three males followed their mate every time sheinitiated a flight during the fertile period. Therewere no differences between the two periods in thenumber of flights initiated by females (Z="0.11,N=6, P=0.92; Fig. 5a), or by males (Z="0.67,N=6, P=0.50; Fig. 5b), and males did not initiatefewer flights than females in either period(Z="0.94, N=5, P=0.35 and Z="0.85, N=7,P=0.40, respectively). On some occasions themale flew long distances during observations, andprobably left the territory. For the post-fertileperiod, we had too few data for any meaningful

Pro

port

ion

of

tim

e se

xes

<1 m

apa

rt

0.9

0–11

Day

0.4

0.3

0.2

0.1

1

1 1 3 4

6

8

4

5 8

8

8

6

8

7

8 6

3 1 1 1

6 90.8

0.7

0.6

0.5

–9 –7 –5 –3 –1 1 3 5 7 9 11

Figure 4. Intensity of mate guarding in bluethroats inrelation to start of egg laying (=day 0). Bars indicate1 , numbers above bars indicate sample size.

3

0

Tim

e fe

mal

e fl

ew/m

ale

foll

owed

Femaleflew

0.5

2.5

2

1.5

1

(a)

3

0

0.5

2.5

2

1.5

1

(b)

Male followed

Femaleflew

Male followed

Femaleflew

Male followed

Femaleflew

Male followed

Pre-fertile Fertile

Pre-fertile Fertile

Figure 5. Mean number of flights (per 20-min obser-vation period) initiated by (a) females and (b) males inthe pre-fertile and fertile period, and mean number oftimes (a) the males followed their mates and (b) thefemales followed their mates (per 20 min) in the pre-fertile and fertile periods, respectively. Bars indicate 1 ,N=6.

Krokene et al.: Paternity in the bluethroat 411

tests, since females were incubating most of thetime and males were seldom observed.If the female benefits from mate guarding by the

male (e.g. for protection against extra-pair copu-lations or harassment) she should also follow themale whenever he initiates a flight. However,females seldom followed their mates, and theproportion of times the females followed theirmates did not differ significantly between the twoperiods (Z="1.00, N=6, P=0.32; Fig. 5b). Infact, only one female followed her mate during thefertile period, and only for 10% of the flights.Females never followed their mates when theyinitiated a flight during the pre-fertile period.There were no significant correlations between

the proportion of time the mates were less than1 m apart and how often the male followed hismate (rS=0.31, N=8, P=0.56), or between theproportion of time the mates were less than 1 mapart and the intrusion rate (rS=0.23, N=8,P=0.54). However, the sample sizes were small,and we cannot rule out the possibility of a truepositive relationship in either case.

Singing Activity and Song Flights

Møller’s (1991a) fertility-announcement hy-pothesis predicts that a male’s singing activityshould peak during his mate’s fertile period.Singing activity varied significantly betweenthe pre-fertile, fertile and post-fertile periods(repeated measures ANOVA: F2,8=34.12, P<0.001), with a marked decline from the pre-fertileto the post-fertile period. The decline in singingactivity from the pre-fertile to the fertile periodwas significant (t=5.21, df=8, P<0.001) as well asthat from the fertile to the post-fertile period(t=2.94, df=8, P=0.010). This pattern is incon-sistent with the fertility-announcement hypoth-esis. Similarly, the number of song flights alsodeclined from the pre-fertile to the post-fertileperiod, though the decline only approached stat-istical significance (repeated measures ANOVA:F2,8=3.22, P=0.07). Both singing activity (Fig.6a) and number of song flights (Fig. 6b) increasedslightly around pair formation, but the increaseswere not statistically significant (Wilcoxon signed-rank tests: Z="0.84, N=8, P=0.40 and Z="0.70, N=8, P=0.48, respectively).Møller’s (1991a) fertility-announcement hy-

pothesis further predicts that the intrusion rateshould be inversely related to the singing activity

of the territorial male. The relationships betweenthe intrusion rate and the singing activity and thefrequency of song flights were in the expecteddirection, but not statistically significant (rS="0.51, N=8, P=0.18 and rS="0.63, N=8,P=0.10, respectively).

Behavioural Correlates of Paternity

Behaviour data were available for nine pairs,but only eight pairs could be used to study therelationship between behaviour and paternity,because one brood was predated before bloodsamples of the nestlings could be taken. Four ofthe eight males had reduced paternity and fourhad full paternity in their nest.

8

0

Sin

gin

g ac

tivi

ty 6

4

2

(a)

2.5

0S

ong

flig

hts

2

1.5

1

(b)

Befo

re p

air f

orm

atio

nPr

e-la

ying

, pre

-fert

ilePr

e-la

ying

, fer

tile

Egg

layi

ng

Incu

batio

n

Figure 6. (a) Mean singing activity of nine male blue-throats from territory establishment until incubation.Singing activity is defined as the number of 2-minintervals where song was recorded during a 20-minobservation bout. (b) Mean frequency of song flights(flights per 20 min) for nine male bluethroats fromterritory establishment until incubation. Bars indicate1 .

Animal Behaviour, 52, 2412

Males with full paternity tended to follow theirpartners more often than males with extra-pairoffspring, and there was also a tendency towardshigher intrusion rates in the territories of maleswith extra-pair offspring but neither result wasstatistically significant (Table I). Males withextra-pair offspring in their nest did not differfrom males with full paternity with respect tosinging activity or frequency of song flights(Table I).

DISCUSSION

Paternity

The percentage of extra-pair paternity docu-mented in this study (20% of nestlings) is high,but not exceptionally high for a socially mono-gamous passerine (Birkhead & Møller 1992).Higher frequencies of extra-pair paternity havebeen recorded, for example in tree swallows,Tachycineta bicolor (Lifjeld et al. 1993; Dunnet al. 1994), indigo buntings, Passerina cyanea(Westneat 1987a, 1990), white-crowned sparrows,Zonotrichia leucophrys (Sherman & Morton 1988)and reed buntings, Emberiza schoeniclus (Dixonet al. 1994). On the other hand, several specieshave lower frequencies of extra-pair paternity,with less than 5% of the offspring being siredextra-pair, as for example in dunnocks, Prunellamodularis (Burke et al. 1989), zebra finches,Taeniopygia guttata (Birkhead et al. 1990) andgreat reed warblers, Acrocephalus arundinaceus(Hasselquist et al. 1995). In willow warblers,Phylloscopus trochilus, and wood warblers,P. sibilatrix, no extra-pair offspring have beenrecorded despite a considerable number of broods

being analysed (Gyllensten et al. 1990). Thisvariation among species, many of which have abreeding biology similar to that of the bluethroat,is largely unexplained. To understand paternitypatterns, it is important to know whether extra-pair paternity arises through rapid mate switch-ing, mate replacement or extra-pair copulation(Sims et al. 1987; Birkhead &Møller 1992), and towhat extent the behaviour is under male or femalecontrol (Lifjeld et al. 1994).In birds, testis size is correlated with the level of

extra-pair paternity (Møller & Briskie 1995). Inbluethroats, the testis size is close to the valueexpected for its body mass (Møller 1991b). Hence,the moderate level of extra-pair paternity found inthis study is consistent with this pattern.Extra-pair paternity is particularly likely to

result from rapid mate switching in species inwhich eggs are laid soon after pair formation, forexample in migrant birds such as the bluethroat,which form seasonal pair bonds. This is becausesperm can remain viable in the female’s reproduc-tive tract for several days (Birkhead & Møller1992), a time potentially long enough for rapidmate switching to result in extra-pair paternity. Inthe present study we do not know for sure whetherfemales that raised extra-pair offspring had beenpaired with another male before they settledwith their mate. But, since the time from pairformation to the start of egg laying was as muchas 6 days on average and a last-male-sperm-precedence pattern seems to be general for birds(Birkhead & Møller 1992), it seems unlikely thatrapid mate switching can explain the many casesof extra-pair paternity. Besides, only one possiblecase of mate switching was observed, but this wasbefore the supposed fertile period, and the female

Table I. Behavioural characteristics of bluethroat pairs with and without extra-pair offspring in their nest

Variable NExtra-pairoffspring

No extra-pairoffspring U-statistic P

Intrusion rate 8 1.41&0.56 0.31&0.20 1.5 0.06Proportion <1 m apart 8 0.34&0.06 0.32&0.14 7.5 0.88Female flights 7 3.06&0.47 1.77&0.54 2.0 0.16Proportion of male following 7 0.77&0.10 1.00&0.00 1.5 0.08Male flights 7 1.42&0.25 1.67&0.67 6.0 1.00Proportion of female following 7 0.00&0.00 0.03&0.03 4.0 0.25Song activity 8 2.60&0.45 2.73&0.57 6.0 0.56Song flights 8 1.17&0.91 0.63&0.32 7.0 0.77

Values are given as X& for 20-min observation periods.

Krokene et al.: Paternity in the bluethroat 413

involved had no extra-pair offspring in her nest.Furthermore, mate replacement during the fertileperiod (Pinxten et al. 1993) can also be excludedas no such replacements were observed.Colegrave et al. (1995) showed that the opti-

mum time for extra-pair copulations to result inextra-pair paternity in the zebra finch was aroundday "2 to day 1, which fits well with our obser-vation of a peak in mate guarding during thesedays in the bluethroat. Therefore, the behaviouralevidence, with a peak in territorial intrusion bymales, and intense mate guarding around the startof egg laying, points to extra-pair copulations asthe major source of extra-pair paternity. Oneextra-pair copulation was also recorded.The fact that all intruders were males, and

that nearly all intrusions coincided with femalefertility, suggests that males aimed for extra-paircopulations. Similar observations of territorialintrusion by males seeking extra-pair copulationshave been reported from several other species, forexample, cattle egrets, Bubulcus ibis (Fujioka &Yamagishi 1981), yellow warblers, Dendroicapetechia (Ford 1983), house martins, Delichonurbica (Lifjeld & Marstein 1994), and severalwaterfowl species (McKinney et al. 1984). Inmany birds, males seeking extra-pair copulationsoften approach females secretively, movingquietly and stealthily, using the vegetation ascover, for example, pied flycatchers, Ficedulahypoleuca (Björklund & Westman 1983), magpies,Pica pica (Buitron 1983), yellow warblers (Ford1983), indigo buntings (Westneat 1987b). Thisdescription seems to fit well for bluethroats. Sincewe never saw females outside their territory, itseemed as if males sought females for extra-paircopulations and not vice versa. Even if mostextra-pair copulations take place on the female’sown territory, however, a role of female choicecannot be dismissed because the number of maleintrusions may be so high that most femaleswould have several opportunities for engaging inextra-pair copulations. In other species, femalesactively seek extra-pair copulations outside theirterritories, for example blue tits, Parus caeruleus(Kempenaers et al. 1992) and black-cappedchickadees, P. atricapillus (Smith 1988). This mayalso be the case for bluethroats, since recenttelemetry studies have revealed that bluethroatfemales do visit the territories of extra-pairmales during the pre-laying period (Smiseth &Amundsen 1995).

Paternity Guards

Bluethroats copulate under cover of vegetation,so we may have missed some cases. However, itseems clear that bluethroats do not copulate atvery high frequencies (sensu Birkhead et al. 1987).Typically, solitary and monogamous passerineshave rather low copulation rates (Birkhead et al.1987).The mate-guarding hypothesis is supported by

the observations that a male stayed closer to hismate and followed her almost every time sheinitiated a flight during the fertile period. Malesand females initiated a similar number of flightsduring the fertile period, which contrasts withsome other studies (e.g. magpies, Birkhead 1982)where females initiated flights more often thantheir mates. The purpose of male flights mighthave been to forage. This seems, however, lesslikely when the males flew long distances, possiblyout of the territory, since there was probablysufficient food inside the territory. A more reason-able explanation for long flights is that the malessought extra-pair copulations. One cost for malesseeking extra-pair copulations outside their terri-tory is the risk of being cuckolded themselves(Birkhead & Møller 1992). Since the nine pairsbred extremely synchronously, with 5 daysbetween the earliest and the latest nest, males mayface a trade-off between the benefit of gainingextra-pair copulations and the risk of beingcuckolded.The observation that both the level of singing

activity and the number of song flights werehighest in the pre-fertile period is in accordancewith the general view that males sing in order toacquire a mate. It does not support the fertility-announcement hypothesis (Møller 1991a; but seealso Sheldon 1994). However, since singingactivity did not decrease until a few days after pairformation, song may also have other functions,for instance to attract secondary mates or femalesfor extra-pair copulations, to retain females(Merilä & Sorjonen 1994) and/or to advertise theterritory to neighbouring males (cf. Slagsvoldet al. 1994). Merilä & Sorjonen (1994) found thatthe singing activity and the rate of song flightsof male bluethroats in Finland peaked beforeor during pair formation and ceased shortlythereafter, which is in accordance with our study.If mate guarding functions to protect paternity,

the presence of the male should reduce the

Animal Behaviour, 52, 2414

occurrence of extra-pair copulations. For mateguarding to be effective, males should be closeenough to their partners to disrupt any extra-paircopulation attempt (Birkhead & Møller 1992).Males stayed within 1 m of their mates much ofthe time during the fertile period, and this distanceis probably short enough to prevent any extra-paircopulation. However, there was no indication thatmales with full paternity spent more time close totheir mates than did males with extra-pair off-spring in their nest. Males with full paternity,however, seemed to follow their females moreoften than males with reduced paternity. Thus,there was an indication that mate guarding hadsome effect in terms of paternity protection, butany firm conclusion cannot be drawn from thisrestricted sample.Other studies have reported a negative relation-

ship between mate guarding and paternity (e.g.Gowaty & Bridges 1991; Kempenaers et al. 1992,1995). Johnsen & Lifjeld (1995) found that blue-throat males that were experimentally made lessattractive to females guarded their mates moreintensely than control males. This finding suggeststhat males adjust their effort in mate guarding tothe perceived risk of losing paternity. Thus it wassurprising to find that males that lost paternityguarded their mates less, not more, intensely thanmales with full paternity.Female bluethroats did not seem to cooperate

in mate guarding, since they did not follow theirmates during the fertile period. It may be thatfemales do not need the male for protection, eitheragainst predators or against males seeking extra-pair copulations (see Björklund et al. 1992). Iffemales are interested in being courted by extra-pair males, mate guarding may mirror a sexualconflict where the pair male seeks to deter hismate from engaging in extra-pair events.To sum up, our data show that sperm compe-

tition is relatively intense in the bluethroat. Thebehavioural evidence strongly suggests that spermcompetition arises through extra-pair copulations.The most prominent anti-cuckoldry tactic inmales is mate guarding during the female’s fertileperiod. The role of females in promoting spermcompetition remains obscure, though there areseveral lines of evidence consistent with a femalechoice mechanism: the fact that females did notcooperate in mate guarding, evidence that femalesdo visit other males’ territories (Smiseth &Amundsen 1995), and that males with reduced

sexual attractiveness mate guard more (Johnsen &Lifjeld 1995).Finally, the association between extra-pair

paternity and mate guarding in territorial passer-ines should be addressed. The bluethroat is anexample of a territorial passerine with a highfrequency of extra-pair paternity and a high inten-sity of mate guarding. A similar association hasbeen found in several other species, for example,the eastern bluebird, Sialia sialis (Gowaty &Bridges 1991), the blue tit (Kempenaers et at.1992, 1995) and the red-winged blackbird,Agelaius phoeniceus (Westneat 1993b). Similarly,mate guarding seems to be relaxed in species withlow intensities of sperm competition, such as theNorwegian population of pied flycatchers (Lifjeldet al. 1991; Chek et al. 1996). However, this doesnot mean that mate guarding and extra-pairpaternity are positively associated traits. There arealso examples of mate-guarding species with lowintensities of sperm competition, for examplewillow warblers (Gyllensten et al. 1990; Arvidsson1992) and great reed warblers (Hasselquist &Bensch 1991; Hasselquist et al. 1995). Further-more, the reed bunting has an exceptionally highlevel of extra-pair paternity but males do not mateguard (Dixon et al. 1994). Hence, there is no clearpattern between the occurrence of mate guardingand the intensity of sperm competition in com-parisons across territorial passerines. One reasoncould be that the causality between mate guardingand sperm competition could go either way.Future studies should test experimentally whethermate guarding is shaped by the intensity of spermcompetition and to what extent mate guardinginfluences sperm competition. The role of femalesin sperm competition may also differ betweenspecies and influence these associations, and there-fore particular attention should be paid to thebehaviour of females.

ACKNOWLEDGMENTS

We thank S. Dale, B. Kempenaers, P. Krokene,B. Sheldon, L. Whittingham, an anonymousreferee and members of the Behavioural Ecologygroup at the Zoological Museum in Oslo, forcomments on the manuscript. We are grateful toJ. E. Stacy and U. H. Refseth for assistance andadvice in the laboratory, and A. Jeffreys forsupplying the probe. The study was funded by theNorwegian Research Council (NFR).

Krokene et al.: Paternity in the bluethroat 415

REFERENCES

Arvidsson, B. L. 1992. Copulation and mate guarding inthe willow warbler. Anim. Behav., 43, 501–509.

Birkhead, T. R. 1982. Timing and duration of mateguarding in magpies, Pica pica. Anim. Behav., 30,277–283.

Birkhead, T. R. & Møller, A. P. 1992. Sperm Compe-tition in Birds: Evolutionary Causes and Consequences.London: Academic Press.

Birkhead, T. R., Atkin, L. & Møller, A. P. 1987. Copu-lation behaviour of birds. Behaviour, 101, 101–138.

Birkhead, T. R., Burke, T., Zann, R., Hunter, F. M. &Krupa, A. P. 1990. Extra-pair paternity and intra-specific brood parasitism in wild zebra finchesTaeniopygia guttata, revealed by DNA fingerprinting.Behav. Ecol. Sociobiol., 27, 315–324.

Björklund, M. & Westman, B. 1983. Extra-pair copu-lation in the pied flycatcher (Ficedula hypoleuca).Behav. Ecol. Sociobiol., 13, 271–275.

Björklund, M., Møller, A. P., Sundberg, J. & Westman,B. 1992. Female great tits, Parus major, avoid extra-pair copulation attempts. Anim. Behav., 43, 691–693.

Buitron, D. 1983. Extra-pair courtship in black-billedmagpies. Anim. Behav., 31, 211–220.

Burke, T. & Bruford, M. W. 1987. DNA fingerprintingin birds. Nature, Lond., 327, 149–152.

Burke, T., Davies, N. B., Bruford, M. W. & Hatchwell,B. J. 1989. Parental care and mating behaviour ofpolyandrous dunnocks Prunella modularis related topaternity by DNA fingerprinting. Nature, Lond., 338,249–251.

Burley, N., Krantzberg, G. & Radman, P. 1982. Influ-ence of colour-banding on the conspecific preferencesof zebra finches. Anim. Behav., 30, 444–455.

Carlson, A. Hillström, L. & Moreno, J. 1985. Mateguarding in the wheatear Oenanthe oenanthe. OrnisScand., 16, 113–120.

Chek, A. A., Lifjeld, J. T. & Robertson, R. J. 1996. Lackof mate guarding in a territorial passerine bird with alow intensity of sperm competition, the pied flycatcher(Ficedula hypoleuca). Ethology, 102, 134–145.

Colegrave, N., Birkhead, A. P. & Lessells, C. M. 1995.Sperm precedence in zebra finches does not require aspecial mechanism of sperm competition. Proc. R.Soc. Ser. B, 259, 223–228.

Dixon, A., Ross, D., O’Malley, S. L. C. & Burke, T.1994. Parental investment inversely related to degreeof extra-pair paternity in the reed bunting. Nature,Lond., 371, 698–700.

Dunn, P. O., Whittingham, L. A., Lifjeld, J. T.,Robertson, R. J. & Boag, P. T. 1994. Effect ofbreeding density, synchrony, and experience onextrapair paternity in tree swallows. Behav. Ecol., 5,123–129.

Ford, N. L. 1983. Variation in mate fidelity in mono-gamous birds. Curr. Ornithol., 1, 329–356.

Fujioka, M. & Yamagishi, S. 1981. Extramarital andpair copulations in the cattle egret. Auk, 98, 134–144.

Gowaty, P. A. & Bridges, W. C. 1991. Behavioral,demographic, and environmental correlates of extra-pair fertilizations in eastern bluebirds, Sialia sialis.Behav. Ecol., 2, 339–350.

Gyllensten, U. B., Jakobsson, S. & Temrin, H. 1990. Noevidence for illegitimate young in monogamous andpolygynous warblers. Nature, Lond., 343, 168–170.

Hasselquist, D. & Bensch, S. 1991. Trade-off betweenmate-guarding and mate attraction in the polygynousgreat reed warbler. Behav. Ecol. Sociobiol., 28, 187–193.

Hasselquist, D., Bensch, S. & von Schantz, T. 1995. Lowfrequency of extrapair paternity in the polygynousgreat reed warbler, Acrocephalus arundinaceus. Behav.Ecol., 6, 27–38.

Jeffreys, A. J., Wilson, V. & Thein, S. L. 1985.Individual-specific ‘fingerprints’ of human DNA.Nature, Lond., 316, 76–79.

Johnsen, A. & Lifjeld, J. T. 1995. Unattractive malesguard their mates more closely: an experiment withbluethroats (Aves, Turdidae: Luscinia s. svecica).Ethology, 101, 200–212.

Johnson, A. L. 1986. Reproduction in the female. In:Avian Physiology (Ed. by P. D. Sturkie), pp. 404–431.New York: Springer-Verlag.

Kempenaers, B. & Dhondt, A. A. 1993. Why do femalesengage in extra-pair copulations? A review of hypoth-eses and their predictions. Belg. J. Zool., 123, 93–103.

Kempenaers, B., Verheyen, G. R., Van den Broeck, M.,Burke, T., Van Broeck-Hoven, C. & Dhondt, A. A.1992. Extra-pair paternity results from female prefer-ence for high-quality males in the blue tit. Nature,Lond., 57, 494–496.

Kempenaers, B., Verheyen, G. R. & Dhondt, A. A.1995. Mate guarding and copulation behaviour inmonogamous and polygynous blue tits: do malesfollow a best-of-a-bad-job strategy? Behav. Ecol.Sociobiol., 36, 33–42.

Lifjeld, J. T. & Marstein, B. 1994. Paternity assurancebehaviour in the house martin Delichon urbica.J. Avian Biol., 25, 231–238.

Lifjeld, J. T. & Robertson, R. J. 1992. Female control ofextra-pair fertilization in tree swallows. Behav. Ecol.Sociobiol., 31, 89–96.

Lifjeld, J. T., Slagsvold, T. & Lampe, H. M. 1991. Lowfrequency of extra-pair paternity in pied flycatchersrevealed by DNA fingerprinting. Behav. Ecol. Socio-biol., 29, 95–101.

Lifjeld, J. T., Dunn, P. O., Robertson, R. J. & Boag,P. T. 1993. Extra-pair paternity in monogamous treeswallows. Anim. Behav., 45, 213–229.

Lifjeld, J. T., Dunn, P. O. & Westneat, D. F. 1994.Sexual selection by sperm competition in birds: male–male competition or female choice? J. Avian Biol., 25,244–250.

McKinney, F., Cheng, K. M. & Bruggers, D. 1984.Sperm competition in apparently monogamous birds.In: Sperm Competition and the Evolution of AnimalMating Systems (Ed. by R. L. Smith), pp. 523–545.Orlando, Florida: Academic Press.

Merilä, J. & Sorjonen, J. 1994. Seasonal and diurnalpatterns of singing and song-flight activity in blue-throats (Luscinia svecica). Auk, 111, 556–562.

Møller, A. P. 1991a. Why mated songbirds sing somuch: mate guarding and male announcement ofmate fertility status. Am. Nat., 138, 994–1014.

Animal Behaviour, 52, 2416

Møller, A. P. 1991b. Sperm competition, sperm deple-tion, parental care, and relative testis size in birds.Am. Nat., 137, 882–906.

Møller, A. P. & Birkhead, T. R. 1991. Frequent copu-lations and mate-guarding as alternative paternityguards in birds: a comparative study. Behaviour, 118,170–186.

Møller, A. P. & Briskie, J. V. 1995. Extra-pair paternity,sperm competition and the evolution of testis size inbirds. Behav. Ecol. Sociobiol., 36, 357–365.

Mulder, R. A., Dunn, P. O., Cockburn, A., Lazenby-Cohen, K. A. & Howell, M. J. 1994. Helpers liberatefemale fairy-wrens from constraints on extra-pairmate choice. Proc. R. Soc. Ser. B, 255, 223–229.

Oring, L. W., Fleischer, R. C., Reed, J. M. & Marsden,K. E. 1992. Cuckoldry through stored sperm in thesequentially polyandrous spotted sandpiper. Nature,Lond., 359, 631–633.

Pinxten, R., Hanotte, O., Eens, M., Frans, R., Dhondt,A. A. & Burke, T. 1993. Extra-pair paternity andintraspecific brood parasitism in the Europeanstarling, Sturnus vulgaris: evidence from DNAfingerprinting. Anim. Behav., 45, 795–809.

Rangbru, B. 1994. Sexual selection in the bluethroat,Luscinia s. svecica: relationship between male bright-ness and mating success: a field study. Cand. scient.thesis, University of Trondheim.

Seutin, G., White, B. N. & Boag, P. T. 1991. Preser-vation of avian blood and tissue samples for DNAanalyses. Can. J. Zool., 69, 82–90.

Sheldon, B. C. 1994. Song rate and fertility in thechaffinch. Anim. Behav., 47, 986–987.

Sherman, P. W. & Morton, M. L. 1988. Extra-pairfertilizations in mountain white-crowned sparrows.Behav. Ecol. Sociobiol., 22, 413–420.

Sims, M. E., Ball, G. F. & Cheng, M.-F. 1987. Spermcompetition after sequential mating in the ringedturtle-dove. Condor, 89, 112–116.

Slagsvold, T., Dale, S. & Sætre, G.-P. 1994. Dawnsinging in the great tit (Parus major): mate attraction,mate guarding, or territorial defence? Behaviour, 131,115–138.

Smiseth, P. T. & Amundsen, T. 1995. Female blue-throats (Luscinia s. svecica) regularly visit theterritories of extrapair males before egg laying. Auk,112, 1049–1053.

Smith, S. M. 1988. Extra-pair copulations in black-capped chickadees: the role of the female. Behaviour,107, 15–23.

Stacy, J. E. & Jakobsen, K. S. 1993. Testing of nylonmembranes for DNA-fingerprinting with multilocusprobes. Int. J. Genome Res., 2, 159–165.

Stacy, J. E., Ims, R. A., Stenseth, N. C. & Jakobsen,K. S. 1991. Fingerprinting of diverse species withDNA probes generated from immobilized single-stranded DNA templates. Nucleic Acids Res., 19,4004.

Stutchbury, B. & Morton, E. G. 1995. The effect ofbreeding synchrony on extra-pair mating systems insongbirds. Behaviour, 132, 675–690.

Svensson, L. 1984. Identification Guide to EuropeanPasserines. Stockholm: Svensson, L.

Vik, R. 1978. The lake Øvre Heimdalsvatn: a subalpinefreshwater ecosystem. Holarct. Ecol., 1, 81–320.

Westneat, D. F. 1987a. Extra-pair fertilizations in apredominantly monogamous bird: genetic evidence.Anim. Behav., 35, 877–886.

Westneat, D. F. 1987b. Extra-pair copulations in apredominantly monogamous bird: observations ofbehaviour. Anim. Behav., 35, 865–876.

Westneat, D. F. 1990. Genetic parentage in the indigobunting: a study using DNA fingerprinting. Behav.Ecol. Sociobiol., 27, 67–76.

Westneat, D. F. 1992. Do female red-winged blackbirdsengage in a mixed mating strategy? Ethology, 92,7–28.

Westneat, D. F. 1993a. Polygyny and extrapair fertiliz-ations in eastern red-winged blackbirds (Agelaiusphoeniceus). Behav. Ecol., 4, 49–60.

Westneat, D. F. 1993b. Temporal patterns of within-pair copulations, male mate-guarding, and extra-pairevents in eastern red-winged blackbirds (Agelaiusphoeniceus). Behaviour, 124, 267–290.

Westneat, D. F., Sherman, P. W. &Morton, M. L. 1990.The ecology and evolution of extra-pair copulationsin birds. Curr. Ornithol., 7, 331–369.

Wetton, J. H., Carter, R. E. & Parkin, D. T. 1987.Demographic study of a wild house sparrow popu-lation by DNA fingerprinting. Nature, Lond., 327,147–149.

Krokene et al.: Paternity in the bluethroat 417