Behavioral Ecology Vol. 11 No. 3: 265–273
High fidelity on islands: a comparative study
of extrapair paternity in passerine birds
Simon C. Griffith
Department of Evolutionary Biology, EBC, Uppsala University, Sweden
It is commonly assumed that the intensity of sexual selection is lower in island populations. Extrapair paternity (EPP) is widespread within passerine birds and is indicative of sexual selection. A conservative analysis of the levels of EPP in island and
equivalent mainland populations of passerines reveals that insular populations are indeed characterized by low levels of EPP.
This supports the idea that the intensity of sexual selection is lower on islands. This relationship has previously been predicted,
based on the assumption of low levels of genetic variation for fitness in such populations. The evidence from this analysis
suggests that this is just one of several nonmutually exclusive hypotheses that may explain the high fidelity of island-living
females. Key words: extrapair paternity, island populations, passerines, sexual selection. [Behav Ecol 11:265–273 (2000)]
T
he discovery that extrapair paternity (EPP) is a widespread feature of the most common avian mating systems has revolutionized our understanding of sexual selection
in birds. Revealing the causes of the observed variation in the
reported levels of EPP in different species and populations
will be critical to fully understanding the mechanisms driving
extrapair behavior (Petrie and Kempenaers, 1998). If EPP is
the result of a female’s propensity to participate in extrapair
copulations (EPCs) (given that willing males are available),
then variation in the level of EPP will be determined largely
by the costs and benefits of EPCs to females. The observed
level of EPP in natural populations shows great variation, from
0% (e.g., Gyllensten et al., 1990) to ⬎75% (Mulder et al.,
1994). Several factors have been suggested to account for this
variation at both the species and population level. For example, the breeding system—from social monogamy through polygyny and polyandry to cooperative breeding systems—will
obviously have a profound effect on the female cost–benefit
relationship of engaging in EPCs. The precise nature of such
effects is unclear and will undoubtedly be confounded by lifehistory variables such as annual fecundity of individuals and
life span (Birkhead and Møller, 1992).
Other, more subtle variables, which may differ between species but also between populations of one species, will also affect the level of EPP. It has been suggested that the level of
EPP will be positively related to both the local breeding density (Møller and Birkhead, 1993) and degree of breeding synchrony within populations (Stutchbury and Morton, 1995).
Both of these ideas are still controversial, and clear support
for both ideas is lacking (e.g., see Westneat and Sherman,
1997; Weatherhead and Yezerinac, 1998).
Petrie and Lipsitch (1994) proposed that the degree of polygyny will be related to population levels of variation for fitness-related genes. The degree of polygyny could be viewed
synonymously with the level of EPP, with some males likely to
be mated to multiple females. A comparative analysis of the
levels of allozymic variation in species with differing levels of
EPP supports the predictions of the model (Petrie et al.,
1998); however, in their analyses there is some difficulty in
establishing cause and effect. Additionally, this theory relies
on the assumption that EPP is largely caused by females seeking good genes—an idea that currently has received little direct support (but see Hasselquist et al., 1996; Sheldon et al.,
1997). Other studies have demonstrated that EPP may result
from females seeking direct benefits (e.g., Gray, 1997).
Generally, there are two nonmutually exclusive groups of
populations, which may be expected to have a below average
level of genetic diversity, allowing an indirect test of this hypothesis. The first is those populations that have historically
been through a genetic bottleneck due to a significant numerical decline, often due to human pressure. The second is
naturally occurring island-dwelling populations, which are expected to exhibit low levels of genetic variation due to a combination of founder effects, inbreeding, genetic drift, and low
rates of dispersal ( Jaenike, 1973; Wright, 1931). Frankham
(1997) found that, like other taxa, island-dwelling populations
of birds do indeed have lower levels of genetic variation (measured with allozymes, nuclear DNA markers, and mitochondrial inversions) than those living on larger landmasses. On
this basis, a comparative analysis of EPP in island and mainland populations of birds will test the expectation of Petrie
and Lipsitch (1994) that island populations will exhibit significantly lower levels of EPP.
If island populations do exhibit lower levels of EPP, then
the implications are potentially great for some of the classic
evolutionary studies which have focused on island populations
of passerines, as well as recent studies (e.g., Grant, 1986; Komdeur, 1996; Smith, 1988). Extrapair paternity is a common and
widespread behavior in avian breeding systems (Petrie and
Kempenaers, 1998) and will play a significant role in sexual
selection within such systems. If such an important component of the avian breeding system is insignificant in island
populations, then they must be abnormal in some fundamental way, and the use of island systems to study mating systems
must surely be questioned.
METHODS
Data set
Address correspondence to S. C. Griffith, Department of Evolutionary Biology, EBC, Uppsala University, Norbyvägen 18D, S-752 36
Uppsala, Sweden. E-mail: [email protected].
Received 8 April 1999; revised 25 July 1999; accepted 24 August
1999.
2000 International Society for Behavioral Ecology
Extrapair paternity is widespread and shows great variation in
the Passeriformes (Petrie and Kempenaers, 1998), a homogeneous group well enough represented in the literature to
enable a test of whether island populations do exhibit low
levels of EPP through the comparative approach. I have followed Westneat et al.’s (1990) definitions of pair-bonds and
266
EPP; thus a pair-bond is viewed as an extended social and
sexual association between a male and female, lasting considerably longer than the time taken to copulate. Subsequently,
EPP can result from a mating between a female and a male
other than her pair-bonded mate. This definition excludes
copulations between females and males that have extensive
social associations such as, for example, a multiply mated female in a polyandrous breeding system (e.g., the mating system of the dunnock, Prunella modularis; Burke et al., 1989).
Cooperative breeding is defined as a reproductive system in
which one or more members of a social group of more than
two adults provide care to young that are not their own (Stacey and Koenig, 1990). For cooperative breeders, EPP is actually extragroup paternity; however, for simplicity it will also
be referred to as EPP.
I compiled data on levels of EPP from original sources that
were located by searching the major biological and ornithological journals and searching the BIDS and BIOSIS databases. All published studies (up to the end of January 1999) that
have reported the level of EPP in wild, unmanipulated populations based on reliable genetic techniques (multi- or singlelocus minisatellite fingerprinting or polymerase chain reaction-based microsatellite genotyping) were included. I omitted studies based on allozymes, plumage markers, and sex differences in heritability estimates due to their poor precision
and the ambiguity of such results with regard to the difference
between EPP and intraspecific brood parasitism (see Westneat
and Sherman, 1997). There is no reason to suspect that the
exclusion of these results caused consistent bias with regard
to the hypothesis being tested.
To be applicable to the hypothesis of Petrie and Lipsitch
(1994), an island population should ideally conform to both
of the following characteristics: (1) the population should be
sedentary on the island throughout the year. Those species
that breed on an island but are migratory cannot reasonably
be expected to suffer reduced genetic variation because vagrants are likely to immigrate into the breeding population.
Similarly, the population must be effectively isolated from wintering conspecifics. (2) The island must be reasonably small
so as not to contain too large a population. Fundamental to
the idea of reduced genetic variation within an island population is the effective population size. This will be related to
the size of the island via the ecological carrying capacity and
also to the spatial distribution of individuals within that population. A landmass such as Britain, which has been classed
as an island by some authors (e.g., Fitzpatrick, 1998), obviously cannot realistically be classified as an island with respect
to the hypothesis in question.
Although it may appear that the definition of what constitutes an island appears subjective, in this data set there is a
clear distinction between island and nonisland populations.
Size is potentially the most arbitrary category, and in this data
set there is a considerable difference between the smallest
piece of land classified as mainland (Britain 244,820 km2) and
the largest island (Gotland, Sweden; 57⬚ N, 18⬚ E, 3170 km2;
see Figure 1). Finally, although small patches of suitable habitat within a larger land mass may represent islands to certain
species, I did not account for such situations in this study due
to the obvious difficulty in detecting and defining such ecologically delimited islands.
Analysis
The data set consists of 74 populations from 54 different species. Eight of the populations in the data set are cooperatively
breeding species, which were excluded from the main analysis
because EPP is likely to be affected by different factors in such
breeding systems. Of the remainder, 13 of the populations live
Behavioral Ecology Vol. 11 No. 3
Figure 1
The size of landmass on which the 14 most insular populations are
found. Populations considered to be truly insular in this study are
those on a landmass ⬍10,000 km2. Gotland, Sweden (b) is the
largest landmass to be considered an island, and Britain (a) is the
smallest landmass not classified as an island (see Methods).
on islands and 53 live on larger land masses. For species in
which the level of EPP had been estimated in more than one
population, I calculated a mean for the species based on all
the populations within one of the two categories (either island
or mainland), weighted by sample size of the component estimates.
The hypothesis can be tested using a simple comparison of
the mean level of EPP in the island populations with the mean
level in the mainland populations. The validity of this test
relies on the independent derivation of the island state. The
phylogeny of the order is presented in Table 1, and from the
distribution of island populations it seems most parsimonious
that the island state is in fact independently derived for each
of these populations. In addition, only three of the species are
island endemics with most species living both on islands and
mainland and therefore island living is not a species characteristic. Although for the above reasons it is unnecessary to
control for the effects of nonindependent derivation of island
populations, to be conservative I repeated the analysis using
a pairwise analysis of closely related species or populations
(following Møller and Birkhead, 1992). For each pairwise
comparison I paired the island population with the mainland
population most closely related to it using the molecularbased phylogeny of Sibley and Alquist (1990) (see Table 2).
In each case pairs were constructed with species from the
same tribe or subfamily. Two island populations could not be
paired (Petroica australis and Zosterops lateralis), as they are
the only members of their respective families subjected to genetic parentage studies. The exclusion of these two can only
make the analysis more conservative, as the level of EPP in
both was found to be 0%, and therefore it is impossible for a
comparative population to have a lower level. Four of the 10
pairs are intraspecies comparisons, which is the best test of
the hypothesis. Finally, as discussed above, to conform to the
idea of Petrie and Lipsitch (1994), island populations should
be closed to excessive, regular gene flow, and therefore species that breed on islands but migrate to them annually should
be excluded from the analysis. Four of the 10 island popula-
Griffith • Island levels of EPP
tions are migratory, and therefore I repeated the analysis with
these pairs excluded.
For all the above comparisons mean levels of EPP were established from the available data (again weighted for sample
size), and the pairs were analyzed using the nonparametric
paired Wilcoxon test.
RESULTS
The mean level of EPP in island-dwelling populations of passerines was significantly lower than the level of EPP in populations living on larger landmasses (see Figures 2 and 3a). The
mean level of EPP in island populations was 8.2% (⫾3.4 SEM)
and in mainland populations was 17.6% (⫾1.7nn-Whitney,
z ⫽ 2.52, n ⫽ 66, p ⫽ .01; see Figure 3a).
The pairwise comparison between island and equivalent
mainland populations, which controlled for the possible nonindependent derivation of the island state, confirmed this result (Wilcoxon z ⫽ 2.16, n ⫽ 10, p ⫽ .03; Figure 3b). In 8 of
the 10 pairs, the island population had a lower level of EPP.
Interestingly, the two island populations that did not conform
to the overall relationship were the populations of the collared flycatcher and willow warbler, both on Gotland, the largest island in the data set and also both migratory species. As
shown in Figure 3a, the four migratory island populations exhibit the highest levels of EPP within the island data set. This
observation suggests that, as expected, the island effect does
not apply to migratory species. If the analyses are repeated
excluding these four migratory species, there is a significant
difference between island and mainland species, even though
the sample size is greatly reduced (Wilcoxon z ⫽ 2.81, n ⫽ 6,
p ⫽ .005).
As shown in Figure 2, the accuracy of reported levels of EPP
are dependent on the sample size of offspring used to construct the estimate. For illustrative purposes, the shaded areas
on the graph represent the 95% and 99% confidence intervals around the mean level of all passerine populations
(15.9% ⫾ 1.6), across different sample sizes. As illustrated, the
estimates of island populations are based on sample sizes
equivalent to those of the mainland populations. The four
island populations within the boundaries of the 95% confidence limits around the mean level of EPP for passerines are
the four migratory species (see above); nearly all of the others
(seven out of eight) are significantly lower than the mean
level at the respective sample sizes. The single exception is an
estimate of 0% based on a small sample of offspring.
Therefore, the above results are not unduly affected by a
consistent bias in the unreliability of either island or mainland
population estimates. In the pairwise comparative data there
was no difference between the mean sample size used for island and mainland estimates of EPP (Wilcoxon z ⫽ 0.45, n ⫽
10, p ⫽ .65). For 8 of the 10 pairs in the comparative analysis
there are estimates of breeding density for both the island
and mainland populations (see Table 1). In five cases the island populations bred at a higher density, whereas in the other three the mainland populations bred at higher density, but
in most cases the differences were not great. There is no significant difference between the breeding density of the mainland and island populations used in this analysis (Wilcoxon z
⫽ 0.98, n ⫽ 9, p ⫽ .33). Most convincingly, the differences
between the breeding densities of the island and mainland
populations of the four species that are represented by both
mainland and island populations (i.e., Parus major, P. caeruleus, Phylloscopus trochilus, and Passer domesticus) are negligible (see Table 1).
267
DISCUSSION
The finding that insular populations exhibit significantly lower levels of EPP than other populations is consistent with the
model proposed by Petrie and Lipsitch (1994). However,
there are several viable alternatives that are difficult to exclude. First, the set of species or populations inhabiting islands today (and therefore available to study) may not reflect
a nonrandom set of all passerines. Such populations may be
present due to a characteristic in which suitability for an insular existence covaries with a significantly lower-than-average
level of EPP. For example, Sorci et al. (1998) found that, after
anthropogenic introduction to New Zealand, dichromatic species were more likely to become extinct. The interpretation
was that strong sexual selection made a species less competitive and unable to force its way into a new environment.
An alternative idea which may bear some relevance comes
from the detailed studies of speciation in Hawaiian Drosophila
and is argued to be responsible for asymmetric sexual isolation in species pairs on different islands. It was suggested that,
after a founder event, when a population is still small, females
are unable to satisfy all their mate choice requirements, and
this subsequently leads to a shift in the choosiness of females
within sister species pairs (Kaneshiro, 1976). In the given example of Drosophila, courtship patterns are quite complicated,
and it appears that only some parts of the courtship are lost
(Hoikkala and Kaneshiro, 1993). It is possible that such
founder effects may lead to a reduction in the choosiness of
females even in species in which courtship is more simple or
based primarily on one or two simple traits, particularly if
there are costs associated with female choosiness. Either of
these two routes of selection may have led to differences between the island populations and mainland populations used
in this analysis. However, with our present limited knowledge
about the extent of genetic variation for both female choosiness and promiscuity in birds, it is difficult to predict the importance of such effects. It is also hard to imagine how such
processes could account for such significant differences between populations of the same species (e.g., Passer and Parus), particularly where gene flow is not restricted (Griffith et
al., 1999; Verhulst and van Eck, 1996).
An alternative to differences between populations caused
by selection is the idea that the observed differences between
island and mainland populations are the result of phenotypic
plasticity in response to the island environment. For example,
a female in an island environment may be less likely to seek
EPCs than a female in a mainland population. It is not known
whether such differences are due to selection or environmental determination (or a combination of both), but several
studies have revealed systematic differences between islanddwelling birds and their mainland counterparts. Male coloration has been shown to be drabber on islands (Fitzpatrick,
1998), and island and mainland populations show divergence
in many reproductive traits. For example, great tits on offshore islands in Denmark bred later, laid smaller clutches, and
laid larger eggs than those on the mainland (Wiggins et al.,
1998). Low genetic diversity, lower food abundance, and density dependence were given as possible reasons for these differences in the reproductive ecology of the different populations.
The evidence for lower genetic variation of island versus
mainland populations is quite clear. Island populations of
mammals, birds, reptiles, insects, and plants had 29% lower
heterozygosity compared with mainland populations of the
same species (Frankham, 1997). This gives some support to
the idea that low genetic variation may lead to reduced levels
of EPP through the route proposed by Petrie and Lipsitch
(1994); however, this interpretation should be drawn with cau-
Behavioral Ecology Vol. 11 No. 3
268
Table 1
Phylogeny of the passeriformes (based on Sibley and Alquist, 1990)
Family
Formicariidae
Maluridae
Meliphagidae
Pardalotidae
Eopsaltriidae
Laniidae
Vireonidae
Muscicapidae
Subfamily
Tribe
Malurinae
Manorinae
Acanthizinae
Malurini
Manorini
Sericornithini
Lanius
Turdinae
Muscicapinae
Muscicapini
Saxicolini
Sturnidae
Certhiidae
Paridae
Sturnini
Troglodytinae
Parinae
Species
MS Status Area
Cercomacra tyrannina
Malurus cyaneus
Manorina melancephala
Sericornis frontalis
Petroica australis
Lanius bucephalus
Vireo olivaceus
V. solitarius
Turdus grayi
Sialia mexicana
S. sialis
Ficedula albicollis*
F. hypoleuca*
Oenanthe oenanthe*
Luscinia svecica*
Sturnus vulgaris
Troglodytes aedon
Campylorhynchus griseus
C. nuchalis
Parus atricapillus
P. caeruleus*
r
r
r
r
r
m
m
m
r
p
p
m
m
m
m
p
m
r
r
r
p
p
p
p
p
p
p
r
m
m
m
m
m
r
m
m
m
r
m
m
m
r
r
m
r
r
n
r
r
r
r
r
r
m
m
r
m
m
m
r
m
P. cristatus
P. major*
Aegithalidae
Hirundinidae
Hirundininae
Zosteropidae
Sylviidae
Acrocephalina
Passeridae
Fringillidae
P. montanus
Panurus biarmicus
Psaltriparus minimus
Progne subis
Delichon urbica
Hirundo ariel
Hirundo rustica
Tachycineta bicolor
Zosterops lateralis
Acrocephalus arundinaceus
A. paludicola
A. schoenobaenus
A. vaughani taiti
Phylloscopus sibilatrix
P. trochilus*
Passerinae
Passer domesticus*
Motacillinae
Prunellinae
Anthus spinoletta
Prunella collaris
P. modularis
Taeniopygia guttata
Fringilla coelebs
Carpodacus mexicanus*
Loxioides bailleui*
Emberiza citrinella*
E. schoeniclus*
Miliaria calandra*
Passerculus sandwichensis*
Passerina cyanea*
Geospiza scandens*
Dendroica petechia
Setophaga ruticilla
Wilsonia citrina
Cardinalis cardinalis
Agelaius phoeniceus
Estrilidinae
Fringillinae
Estrilidini
Fringillini
Carduelini
Emberizinae
Emberizini
Parulini
Cardinalini
Icterini
c
c
c
i
⬍5
i
3170
i
⬍5
Density
150, 600
600
160, 50
27
38
c
c
m
i
m
i
40
60
8
22, 8
40
50, 65
c
i
⬍5
c
m
i
m
i
3170
⬍5
140, 59
80
120
100
c
c
i
140
100
i
i
610
⬍5
i
⬍5
500
60
200
n
EPP
Referencea
15
181
85
137
62
99
19
37
37
207
83
459
456
73
150
154
790
222
191
53
1808
466
121
1006
924
112
187
59
142
135
203
553
378
122
553
70
345
23
56
229
68
536
305
1052
110
133
82
47
80
20
123
216
44
160
63
159
355
108
356
37
1255
0
76.2
5.9
12.4
0
10.1
57.8
2.7
37.8
19
8.4
15.5
8.0
11
20
9.1
8.4
4.1
4.2
17
12
6.6
12.4
11.5
5.5
0.9
14.4
0
18.3
17.3
14
22.5
53.1
0
3.1
36
7.8
0
0
15.7
27.9
13.6
1.3
5.2
0
0.8
2.4
17
8.8
0
37.4
54.6
0
23.1
34.9
7.5
36.6
39.8
26.7
13.5
28
1
32
48
33
37
12
70
70
22
20
21
34
35,75,76
19
18
68,69
15
13
14
39
41,42,43
40
44
41,71,72
40,45
46
23
38
8
2,3
4
4,5,6,7
9,10
47
24
25
26,27
28
29
74,29
30
67
66
11
16
17
65
64
63
58
49
50
51
52
54
73
55
56
57
53
59,60,61
MS, migration status (r ⫽ resident, m ⫽ migratory, p ⫽ partial migrant); status: c ⫽ cooperative breeder, i ⫽ island population, m ⫽
mainland populations (only given for species where an island population also exists); area ⫽ area of the landmass inhabited by island
populations (km2); n ⫽ number of offspring on which the estimate of extrapair paternity (EPP) is based; EPP is percent offspring.
* Species used in the pairwise comparison.
a References for data on extrapair paternity: 1. Fleischer et al. (1997); 2. Whittingham and Lifjeld (1995); 3. Riley et al. (1995); 4. Magrath
and Elgar (1997); 5. Primmer et al. (1995); 6. Smith et al. (1991); 7. Møller and Tegelström (1997); 8. Wagner et al. (1996); 9. Dunn and
Robertson (1993); 10. Lifjeld et al. (1993); 11. Reyer et al. (1997); 12. Yamagishi et al. (1992); 13. Haydock et al. (1996); 14. Piper et al.
Griffith • Island levels of EPP
269
Figure 2
Different population estimates of the level of extrapair paternity (EPP) in the passerines against sample size. The shaded areas represent the
95 and 99% confidence intervals around the passerine mean level of EPP (15.9%) at varying sample sizes to illustrate the effect of sample
size on the reliability of estimates. The confidence intervals were calculated following Rohlf and Sokal (1969).
tion. Petrie and Lipsitch’s (1994) theory concerns additive
genetic variance for traits related to fitness, but most estimates
of genetic variation are based on allozyme studies and studies
of supposedly neutral genetic markers. There is some discussion of the extent to which allozyme heterozygosity correlates
with additive genetic variation for traits related to fitness (Butlin and Treganza, 1998), and, indeed, any relationship that
does exist is likely to be weakened by low effective population
sizes (Lande, 1988).
Several observations point to the fact that, in at least some
cases, the island effect is unlikely to be caused by low genetic
variance. The level of heterozygosity and allelic diversity at
microsatellite loci in the island population of house sparrows
on Lundy off the southwest coast of England were no different
from that in several mainland populations. This observation
was in line with expectation, given the effective population
size and the number of immigrants per generation (Griffith
et al., 1999). Similarly, the population of great tits breeding
on Vlieland, an island in the Dutch Wadden Sea, is highly
unlikely to be genetically different from the mainland population, given that each year, approximately 25% of the breeding females arrived on the island as immigrants from mainland populations (Verhulst and van Eck, 1996). Naturally,
some of these island populations are small, isolated ones that
in some cases have been through substantial genetic bottlenecks: for example, the study population of the New Zealand
←
(1995); 15. Soukup and Thompson (1998); 16. Hartley et al. (1995); 17. Burke et al. (1989); 18. Krokene et al. (1996); 19. Currie et al.
(1998); 20. Dickinson (1998); 21. Meek et al. (1994); 22. Stutchbury et al. (1998); 23. Hoi and Hoi-Leitner (1997); 24. Hasselquist et al.
(1995); 25. Schulze-Hagen et al. (1993); 26. Langefors et al. (1998); 27. Buchanan and Catchpole (in press); 28. Brooke and Hartley (1995);
29. Gyllensten et al. (1990); 30. Fridolfsson et al. (1997); 31. Bjørnstad and Lifjeld (1997); 32. Mulder et al. (1994); 33. Whittingham et al.
(1997); 34. Sheldon and Ellegren (1999); 35. Ellegren et al. (1995); 36. Rätti et al. (1995); 37. Ardern et al. (1997); 38. Bruce et al. (1996);
39. Otter et al. (1994); 40. Krokene et al. (1998); 41. Gullberg et al. (1992); 42. Kempenaers et al. (1992); 43. Kempenaers et al. (1997); 44.
Lens et al. (1997); 45. Verboven and Mateman (1997); 46. Orell et al. (1997); 47. Robertson (1996); 48. Põldmaa et al. (1995); 49. Sundberg
and Dixon (1996); 50. Dixon et al. (1994); 50a. Petren et al. (1999); 51. Hartley et al. (1993); 52. Freeman-Gallant (1996); 53. Ritchison et
al. (1994); 54. Westneat (1990); 55. Yezerinac et al. (1995); 56. Perreault et al. (1998); 57. Stutchbury et al. (1997); 58. Fleischer et al.
(1994); 59. Westneat (1993); 60. Gray (1996); 61. Weatherhead and Boag (1995); 62. Barber et al. (1996); 63. Hill et al. (1994); 64. Sheldon
and Burke (1994); 65. Birkhead et al. (1990); 66. Griffith et al. (submitted); 67. Wetton and Parkin (1991); 68. Pinxten et al. (1993); 69.
Smith and von Schantz (1993). 70. Morton et al. (1998); 71. Lubjuhn et al. (1993); 72. Strohbach et al. (1998); 73. Petren et al. (1999); 74.
Bjørnstad and Lifjeld (1997); 75. Lifjeld et al. (1991); Rätti et al. (1995). Density estimates were taken from the original source except for
Luscinia svecica (Krokene C, personal communication) and Ficedula albicollis (Sheldon B, personal communication).
270
Behavioral Ecology Vol. 11 No. 3
Figure 3
(a) Population estimates of extrapair paternity (EPP) in island and mainland populations of passerines with the mean (⫾SE). The four
island-dwelling populations that are migratory are indicated by filled symbols. (b) Pairwise comparisons of the level of EPP in island and
mainland populations.
black robin (approximately 150 pairs) was established from
just 5 individuals (Ardern et al., 1997).
Although the difference in genetic variation between island
and mainland populations appears to be a likely candidate for
at least some of the observed differences, there are some nonmutually exclusive alternatives that are testable and require
further investigation. Petrie and Lipsitch (1994) focused on
the potential benefits to participation in EPCs, but the level
of EPP will also depend on the costs of EPP for females. For
a socially monogamous passerine, perhaps the greatest cost is
the loss of male help in rearing offspring. Monogamy is assumed to be so prevalent among passerines due to the constraints of egg laying and the advantages of biparental care to
the successful production of offspring (Gowaty, 1996). Several
studies have revealed a negative relationship between a male’s
certainty of paternity and the proportion of food he brings to
the nest relative female provisioning (see Wright, 1998). Given the potential cost of such punishment to the female in
terms of both current and future reproduction, perhaps in a
more ecologically constrained island environment the costs
are proportionally higher and not worth paying. This idea
could be tested by manipulating ecological aspects of island
populations to reduce the constraints and make them comparable to those encountered by mainland populations.
Other characteristics of insular populations may also adjust
the cost–benefit equation for female participation in EPCs.
For example, Slagsvold and Lifjeld (1994) proposed that EPP
should be lower in populations or species in which females
can assess the quality of males over a long period. Perhaps in
small, insular populations individuals have good opportunity
to thoroughly judge each other, and therefore they will mate
assortatively with respect to quality, resulting in low levels of
EPP. It is difficult to determine the extent to which this idea
may explain the difference between island and mainland populations in the level of EPP. However, it is interesting that
within island populations, those with the highest levels of EPP
were those that migrated to the breeding area and only arrived a short time before pair formation.
Although breeding density may vary between mainland and
island populations, within a species this variation is apparently
limited (certainly within this data set), and there does not
seem to be an overall trend for islands to have either abnormally low or high breeding densities. In fact, in this data set
island populations are more likely to be of a higher density
than corresponding mainland populations. This leads to an
expectation of higher levels of EPP (Westneat and Sherman,
1997) on islands and therefore suggests that the island effect
is stronger than any density dependence of EPP. Certainly, any
affect of islands on EPP is unlikely to be caused by the effects
of breeding density.
Unfortunately, sufficient data were not available to investigate the possibility that breeding synchrony may contribute to
low levels of EPP on islands. The level of breeding synchrony
is a potential confounding variable; however, there is still disagreement over whether it will increase (Stutchbury and Morton, 1995) or decrease (Birkhead and Biggins, 1987; Sherman
and Morton, 1988) the level of EPP. The idea that there is a
simple relationship between the level of breeding synchrony
and the level of EPP has provoked controversy and has yet to
receive any unequivocal empirical support (see Westneat and
Sherman, 1997; Weatherhead and Yezerinac, 1998). It is difficult to imagine why island populations may exhibit the sort
of consistent significant differences in the degree of breeding
synchrony that would be needed to cause low levels of EPP,
even assuming there is a relationship between synchrony and
EPP. Finally, of the island populations cited in this study, several specifically cast doubt on the ability of the synchrony hypothesis to explain the low levels of EPP in their populations.
For example, the common occurrence of asynchronous second and third broods means that the populations have both
synchronous and asynchronously breeding individuals (e.g.,
Griffith et al., 1999; Petren et al., 1999; Verboven and Mateman, 1997).
Griffith • Island levels of EPP
In conclusion, it appears that island populations are characterized by extremely low levels of EPP relative to populations on larger landmasses. This result is consistent with other
observations suggesting that the intensity of sexual selection
is lower in island populations. Although the cause of this relationship may be at least partly due to the low genetic variation of some island populations, it may also be largely determined by nongenetic factors which require further investigation. That island populations are so different with respect to
such an important aspect of behavioral ecology is of great
importance from both a conservation view and because island
populations are often used as model populations for the study
of behavioral and evolutionary ecology.
Thanks to Terry Burke, Anders Møller, Ian Owens, Marion Petrie, Ben
Sheldon, and Dave Westneat for discussion and ideas and two anonymous referees for comments on the manuscript. This work was supported by the UK Natural Environment Research Council and a grant
from the Swedish Natural Sciences Research Council to Ben Sheldon.
REFERENCES
Ardern SL, Ma W., Ewen JG, Armstrong DP, Lambert DM, 1997. Social
and sexual monogamy in translocated New Zealand robin populations detected using minisatellite DNA. Auk 114:20–126.
Barber CA, Robertson RJ, Boag PT, 1996. The high frequency of extrapair paternity in tree swallows is not an artifact of nestboxes. Behav
Ecol Sociobiol 38:425–430.
Birkhead TR, Biggins JD, 1987. Reproductive synchrony and extrapair copulations in birds. Ethology 74:320–334.
Birkhead TR, Burke T, Zann R, Hunter FM, Krupa A, 1990. Extrapair paternity and intraspecific brood parasitism in wild zebra finches Taeniopygia guttata, revealed by DNA fingerprinting. Behav Ecol
Sociobiol 27:315–324.
Birkhead TR, Møller AP, 1992. Sperm competition in birds: evolutionary causes and consequences. London: Academic Press.
Bjørnstad G, Lifjeld JT, 1997. High frequency of extra-pair paternity
in a dense and synchronous population of Willow warblers Phylloscopus trochilus. J Avian Biol 28:319–324.
Brooke M, Hartley IR, 1995. Nesting Henderson reed-warblers (Acrocephalus vaughani taiti) studied by DNA fingerprinting: unrelated
coalitions in a stable habitat? Auk 112:77–86.
Bruce JP, Quinn JS, Sloane SA, White BN, 1996. DNA fingerprinting
reveals monogamy in the bushtit, a cooperatively breeding species.
Auk 113:511–516.
Buchanan KL, Catchpole CK, 1999. Extra-pair paternity in the socially
monogamous sedge warbler (Acrocephalus schoenobaenus) as revealed by multilocus DNA fingerprinting. Ibis 142:12–20.
Burke T, Davies NB, Bruford MW, Hatchwell BJ, 1989. Parental care
and mating behaviour of polyandrous dunnocks Prunella modularis
related to paternity by DNA fingerprinting. Nature 338:249–251.
Butlin RK, Treganza T, 1998. Levels of genetic polymorphism: marker
loci versus quantitative traits. Phil Trans R Soc Lond B 353:187–
198.
Currie DR, Burke T, Whitney RL, Thompson DBA, 1998. Male and
female behaviour and extra-pair paternity in the wheatear. Anim
Behav 55:689–703.
Dickinson JL, 1998. Extrapair paternity, inclusive fitness, and withingroup benefits of helping in western bluebirds. Mol Ecol 7:95–105.
Dixon A, Ross D, O’ Malley SLC, Burke T, 1994. Paternal investment
inversely related to degree of extra-pair paternity in the reed bunting. Nature 371:698–700.
Dunn PO, Robertson RJ, 1993. Extra-pair paternity in polygynous
tree-swallows. Anim Behav 45:231–239.
Ellegren H, Lifjeld JT, Slagsvold T, Primmer CR, 1995. Handicapped
males and extrapair paternity in pied flycatchers: a study using microsatellite markers. Mol Ecol 4:739–744.
Fitzpatrick S, 1998. Intraspecific variation in wing length and male
plumage coloration with migratory behaviour in continental and
island populations. J Av Biol 29:248–256.
Fleischer RC, Tarr CL, Pratt TK, 1994. Genetic structure and mating
system in the palila, an endangered Hawaiin honeycreeper, as assessed by DNA fingerprinting. Mol Ecol 3:383–392.
271
Fleischer RC, Tarr CL, Morton ES, Sangmeister A, Derrickson KC,
1997. Mating system of the dusky antbird, a tropical passerine, as
assessed by DNA fingerprinting. Condor 99:512–514.
Frankham R, 1997. Do island populations have less genetic variation
than mainland populations? Heredity 78:311–327.
Freeman-Gallant CR, 1996. DNA fingerprinting reveals female preference for male parental care in Savannah sparrows. Proc R Soc
Lond B 263:157–160.
Fridolfsson A, Gyllensten UB, Jakobsson S, 1997. Microsatellite markers for paternity testing in the willow warbler Phylloscopus trochilus:
high frequency of extra-pair young in an island population. Hereditas 126:127–132.
Gowaty PA, 1996. Battle of the sexes and origins of monogamy. In:
Partnerships in birds: the study of monogamy (Black JM, ed). Oxford: Oxford University Press; 21–52.
Grant PR, 1986. Ecology and evolution of Darwin’s finches. Princeton,
New Jersey: Princeton University Press.
Gray EM, 1996. Female control of offspring paternity in a western
population of red-winged blackbirds (Agelaius phoeniceus). Behav
Ecol Sociobiol 38:267–278.
Gray EM, 1997. Female red-winged blackbirds accrue material benefits from copulating with extra-pair males. Anim Behav 53:625–639.
Griffith SC, Stewart IRK, Dawson DA, Owens IPF, Burke T, 1999. Contrasting levels of extra-pair paternity in mainland and island populations of the house sparrow (Passer domesticus): is there an ‘island
effect’? Biol J Linn Soc
Gullberg A, Tegelström H, Gelter HP, 1992. DNA fingerprinting reveals multiple paternity in families of great and blue tits (Parus
major and P. caeruleus). Hereditas 117:103–108.
Gyllensten UB, Jakobsson S, Temrin H, 1990. No evidence for illegitimate young in monogamous and polygynous warblers. Nature 343:
168–170.
Hartley IR, Davies NB, Hatchwell BJ, Desrochers A, Nebel D, Burke
T, 1995. The polyandrous mating system of the alpine accentor,
Prunella collaris. II. Multiple paternity and parental effort. Anim
Behav 49:789–803.
Hartley IR, Shepherd M, Robson T, Burke T, 1993. Reproductive success of polygynous male corn buntings (Miliaria calandra) as confirmed by DNA fingerprinting. Behav Ecol 4:310–317.
Hasselquist D, Bensch S, von Schantz T, 1995. Low frequency of extrapair paternity in the polygynous great reed warbler, Acrocephalus
arundinaceus. Behav Ecol 6:27–38.
Hasselquist D, Bensch S, von Schantz T, 1996. Correlation between
male song repertoire, extra-pair paternity and offspring survival in
the great reed warbler. Nature 381:229–232.
Haydock J, Parker PG, Rabenold KN, 1996. Extra-pair paternity uncommon in the cooperatively breeding bicolored wren. Behav Ecol
Sociobiol 38:1–16.
Hill GE, Montgomerie R, Roeder C, Boag P, 1994. Sexual selection
and cuckoldry in a monogamous songbird: implications for sexual
selection theory. Behav Ecol Sociobiol 35:193–199.
Hoi H, Hoi-Leitner M, 1997. An alternative route to coloniality in the
bearded tit: females pursue extrapair fertilizations. Behav Ecol 8:
113–119.
Hoikkala A, Kaneshiro KY, 1993. Change in the signal-response sequence responsible for asymmetric isolation between Drosophila
planitibia and Drosophila silvestris. Proc Natl Acad Sci USA 90:5813–
5817.
Jaenike JR, 1973. A steady state model of genetic polymorphism on
islands. Am Nat 107:793–795.
Kaneshiro KY, 1976. Ethological isolation and phylogeny in the planitibia subgroup of Hawaiian Drosophila. Evolution 30:740–745.
Kempenaers B, Verheyen GR, Dhondt AA, 1997. Extrapair paternity
in the blue tit (Parus caeruleus): female choice, male characteristics,
and offspring quality. Behav Ecol 8:481–492.
Kempenaers B, Verheyen GR, Van den Broeck M, Burke T, Van Boeckhoven C, Dhondt AA, 1992. Extra-pair paternity results from female
preference for high-quality males in the blue tit. Nature 357:494–
496.
Komdeur J, 1996. Influence of age on reproductive performance in
the Seychelles warbler. Behav Ecol 7:417–425.
Krokene C, Anthonisen K, Lifjeld JT, Amundsen T, 1996. Paternity
and paternity assurance behaviour in the bluethroat, Luscinia s svecica. Anim Behav 52:405–417.
Krokene C, Rigstad K, Dale M, Lifjeld JT, 1998. The function of ex-
272
trapair paternity in blue tits and great tits: good genes or fertility
insurance? Behav Ecol 9:649–656.
Lande R, 1988. Genetics and demography in biological conservation.
Science 241:1455–1460.
Langefors A, Hasselquist D, von Schantz T, 1998. Extra-pair fertilizations in the sedge warbler. J Avian Biol 29:134–144.
Lens L, van Dongen S, van den Broeck M, van Broeckhoven C,
Dhondt AA, 1997. Why female crested tits copulate repeatedly with
the same partner: evidence for the mate assessment hypothesis. Behav Ecol 8:87–91.
Lifjeld JT, Dunn PO, Robertson RJ, Boag PT, 1993. Extra-pair paternity in monogamous tree swallows. Anim Behav 45:213–229.
Lifjeld JT, Slagsvold T, Lampe HM, 1991. Low frequency of extra-pair
paternity in pied flycatchers revealed by DNA fingerprinting. Behav
Ecol Sociobiol 29:95–101.
Lubjuhn T, Curio E, Muth SC, Brun J, Epplen JT, 1993. Influence of
extra-pair paternity on parental care in great tits (Parus major). In:
DNA fingerprinting: state of the science (Pena S, ed). Basel: Birkhäuser; 379–385.
Magrath MJL, Elgar MA, 1997. Paternal care declines with increased
opportunity for extra-pair matings in fairy martins. Proc R Soc
Lond B 264:1731–1736.
Meek SB, Robertson RJ, Boag PT, 1994. Extra-pair paternity and intraspecific brood parasitism in eastern bluebirds revealed by DNA
fingerprinting. Auk 111:739–744.
Møller AP, Birkhead TR, 1992. A pairwise comparative method as
illustrated by copulation frequency in birds. Am Nat 139:644–656.
Møller AP, Birkhead TR, 1993. Cuckoldry and sociality: a comparative
study of birds. Am Nat 142:118–140.
Møller AP, Tegelström H, 1997. Extra-pair paternity and tail ornamentation in the barn swallow. Behav Ecol Sociobiol 41:353–360.
Morton ES, Stutchbury BJM, Howlett JS, Piper WH, 1998. Genetic
monogamy in blue-headed vireos and a comparison with a sympatric vireo with extrapair paternity. Behav Ecol 9:515–524.
Mulder RA, Dunn PO, Cockburn RA, Lazenby-Cohen KA, Howell MJ,
1994. Helpers liberate female fairy-wrens from constraints on extrapair mate choice. Proc R Soc Lond B 255:223–229.
Orell M, Rytkonen S, Launonen V, Welling P, Koivula K, Kumpulainen
K, Bachmann L, 1997. Low frequency extra-pair paternity in the
willow tit Parus montanus as revealed by DNA fingerprinting. Ibis
139:562–566.
Otter K, Ratcliffe L, Boag PT, 1994. Extra-pair paternity in the blackcapped chickadee. Condor 96:218–222.
Perreault S, Lemon, RE, Kuhnlein U, 1998. Patterns and correlates
of extrapair paternity in American redstarts (Septophaga ruticilla).
Behav Ecol 8:612–621.
Petren K, Grant BR, Grant PR, 1999. Low extrapair paternity in the
cactus finch (Geospiza scandens). Auk 116:252–256.
Petrie M, Doums C, Møller AP, 1998. The degree of extra-pair paternity increases with genetic variability. Proc Natl Acad Sci USA 95:
9390–9395.
Petrie M, Kempenaers B, 1998. Extra-pair paternity in birds: explaining variation between species and populations. TREE 13:52–58.
Petrie M, Lipsitch M, 1994. Avian polygyny is most likely in populations with high variability in heritable male fitness. Proc R Soc Lond
B 256:275–280.
Pinxten R, Hanotte O, Eens M, Verheyen RF, Dhondt AA, Burke T,
1993. Extra-pair paternity and intraspecific brood parasitism in the
European starling, Sturnus vulgaris: evidence from DNA fingerprinting. Anim Behav 45:795–809.
Piper WH, Parker PG, Rabenold KN, 1995. Facultative dispersal by
juvenile males in the cooperative stripe-backed wren. Behav Ecol 6:
337–342.
Põldmaa T, Holder K, 1997. Behavioural correlates of monogamy in
the noisy miner, Manorina melanocephala. Anim Behav 54:571–578.
Primmer CR, Møller AP, Ellegren H, 1995. Resolving genetic relationships with microsatellite markers: a parentage testing system for the
swallow Hirundo rustica. Mol Ecol 4:493–498.
Rätti O, Hovi M, Lundberg A, Tegelström H, Alatalo RV, 1995. Extrapair paternity and male characteristics in the pied flycatcher. Behav
Ecol Sociobiol 37:419–425.
Reyer H, Bollmann K, Schläpfer AR, Schymainda A, Klecack G, 1997.
Ecological determinants of extrapair fertilizations and egg dumping
in Alpine water pipits (Anthus spinoletta). Behav Ecol 8:534–543.
Riley HT, Bryant DM, Carter RE, Parkin DT, 1995. Extra-pair fertil-
Behavioral Ecology Vol. 11 No. 3
izations and paternity defence in house martins, delichon urbica.
Anim Behav 49:495–509.
Ritchison G, Klatt PH, 1994. Mate guarding and extra-pair paternity
in northern cardinals. Condor 96:1055–1063.
Robertson BC, 1996. The mating system of the Capricorn silvereye
(PhD dissertation). St. Lucia, Brisbane: University of Queensland.
Rohlf J, Sokal RR, 1969. Statistical tables. New York: Freeman and
Company.
Schulze-Hagen K, Swatschek I, Dyrcz A, Wink M, 1993. Multiple paternity in broods of aquatic warblers Acrocephalus paludicola: first
results of a DNA fingerprinting study. J Ornithol 134:145–154.
Sheldon BC, Burke T, 1994. Copulation behavior and paternity in the
chaffinch. Behav Ecol Sociobiol 34:149–156.
Sheldon BC, Ellegren H, 1999. Sexual selection resulting from extrapair paternity in collared flycatchers. Anim Behav 57:285–298.
Sheldon BC, Merilä J, Qvarnstrom A, Gustafsson L, Ellegren H, 1997.
Paternal genetic contribution to offspring condition predicted by
size of male secondary sexual character. Proc R Soc Lond B 264:
297–302.
Sherman PW, Morton ML, 1988. Extra-pair fertilizations in mountain
white-crowned sparrows. Behav Ecol Sociobiol 22:413–420.
Sibley CG, Ahlquist JE, 1990. Phylogeny and classification of birds: a
study in molecular evolution. New Haven, Connecticut: Yale University Press.
Slagsvold T, Lifjeld JT, 1994. Polygyny in birds—the role of competition between females for male parental care. Am Nat 143:59–94.
Smith JNM, 1988. Determinants of lifetime reproductive success in
the song sparrow. In: Reproductive success (Clutton-Brock TH, ed).
Chicago: University of Chicago Press; 154–172.
Smith HG, Montgomerie R, Põldmaa T, White BN, Boag PT, 1991.
DNA fingerprinting reveals relation between tail ornaments and
cuckoldry in barn swallows, Hirundo rustica. Behav Ecol 2:90–98.
Smith HG, von Schantz T, 1993. Extra-pair paternity in the European
starling: the effect of polygyny. Condor 95:1006–1015.
Sorci G, Møller AP, Clobert J, 1998. Plumage dichromatism of birds
predicts introduction success in New Zealand. J Anim Ecol 67: 263–
269.
Soukup SS, Thompson CF, 1998. Social mating system and reproductive success in house wrens. Behav Ecol 9:43–48.
Stacey PB, Koenig WD, 1990. Cooperative breeding in birds: long
term studies of ecology and behavior. Cambridge: Cambridge University Press.
Strohbach S, Curio E, Bathen A, Epplen JT, Lubjuhn T, 1998. Extrapair paternity in the great tit (Parus major): a test of the good genes
hypothesis. Behav Ecol 9:388–396.
Stutchbury BJ, Morton ES, 1995. The effect of breeding synchrony
on extra-pair mating systems in songbirds. Behavior 132:675–690.
Stutchbury BJM, Morton ES, Piper WH, 1998. Extra-pair mating system of a synchronously breeding tropical songbird. J Avian Biol 29:
72–78.
Stutchbury BJM, Piper WH, Neudorf DL, Tarof SA, Rhymer JM, Fuller
G, Fleischer RC, 1997. Correlates of extra-pair fertilization success
in hooded warblers. Behav Ecol Sociobiol 40:119–126.
Sundberg J, Dixon A, 1996. Old, colourful male yellowhammers, Emberiza citrinella, benefit from extra-pair copulations. Anim Behav 52:
113–122.
Verboven N, Mateman AC, 1997. Low frequency of extra-pair fertilizations in the great tit Parus major revealed by DNA fingerprinting.
J Avian Biol 28:231–239.
Verhulst S, van Eck HM, 1996. Gene flow and immigration rate in an
island population of great tits. J Evol Biol 9:771–782.
Wagner RH, Schug MD, Morton ES, 1996. Confidence of paternity,
actual paternity and parental effort by purple martins. Anim Behav
52:123–132.
Weatherhead PJ, Boag PT, 1995. Pair and extra-pair mating mating
success relative to male quality in red-winged blackbirds. Behav Ecol
Sociobiol 37:81–91.
Weatherhead PJ, Yezerinac SM, 1998. Breeding synchrony and extrapair mating in birds. Behav Ecol Sociobiol 43:217–219.
Westneat DF, 1990. Genetic parentage in the indigo bunting: a study
using DNA fingerprinting. Behav Ecol Sociobiol 27:67–76.
Westneat DF, 1993. Polygyny and extrapair fertilizations in eastern
red-winged blackbirds (Agelius phoeniceus). Behav Ecol 4:49–59.
Westneat DF, Sherman PW, Morton ML, 1990. The ecology and evolution of extra-pair copulations in birds. Curr Orn 7:331–369.
Griffith • Island levels of EPP
Westneat DF, Sherman PW, 1997. Density and extra-pair fertilizations
in birds: a comparative analysis. Behav Ecol Sociobiol 41:205–215.
Wetton JH, Parkin DT, 1991. An association between fertility and cuckoldry in the house sparrow, Passer domesticus. Proc R Soc Lond B
245:227–233.
Whittingham LA, Dunn PO, Magrath RD, 1997. Relatedness, polyandry and extra-group paternity in the cooperatively-breeding
white-browed scrubwren (Sericornis frontalis). Behav Ecol Sociobiol
40:261–270.
Whittingham LA, Lifjeld JT, 1995. Extra-pair fertilizations increase the
opportunity for sexual selection in the monogamous house martin
Delichon urbica. J Avian Biol 26:283–288.
Wiggins DA, Møller AP, Sørensen MFL, Brand LA, 1998. Island bio-
273
geography and the reproductive ecology of great tits Parus major.
Oecologia 115:478–482.
Wright S, 1931. Evolution in Mendelian populations. Genetics 16:97–
159.
Wright J, 1998. Paternity and paternal care. In: Sperm competition
and sexual selection (Birkhead TR, Møller AP, eds). London: Academic Press; 117–139.
Yamagishi S, Nishiumi I, Shimoda C, 1992. Extra-pair fertilization in
monogamous bull-headed shrikes revealed by DNA fingerprinting.
Auk 109:711–721.
Yezerinac SM, Weatherhead PJ, Boag PT, 1995. Extra-pair paternity
and the opportunity for sexual selection in a socially monogamous
bird (Dendroica petechia). Behav Ecol Sociobiol 37:179–188.