doi:10.1111/j.1420-9101.2010.02013.x
Inbred women in a small and isolated Swiss village have fewer
children
E. POSTMA, L. MARTINI & P. MARTINI
Institute of Evolutionary Biology and Environmental Studies, University of Zürich, Zürich, Switzerland
Keywords:
Abstract
consanguinity;
family size;
humans;
inbreeding;
inbreeding depression;
isolated populations;
maternal effect;
reproduction;
Switzerland.
Despite overwhelming evidence for a negative effect of inbreeding on fitness
in plants and nonhuman animals, the exact nature of its effect in humans
remains subject to debate. To obtain a better understanding of the effects of
inbreeding on reproductive success in humans, we reconstructed the genealogies of the current inhabitants of a small and isolated Swiss village and used
these to estimate the level of inbreeding of both members of all married
couples, as well as their relatedness (i.e. the level of inbreeding of their
offspring). Although there was no effect of parental relatedness on the number
of children a couple had, we found that inbred mothers, but not inbred
fathers, had significantly fewer children. Thus, although related couples did
not have fewer children themselves, their inbred daughters did leave them
with fewer grandchildren. Thereby, we provide evidence for the existence of
inbreeding depression in human fertility, also in relatively outbred and
egalitarian communities.
Introduction
Darwin married his cousin and feared that the poor
health and early death of several of his ten children were
the result of ‘a serious form of inheritance from my poor
constitution’ (Burkhardt et al., 1997). He therefore
‘wished that the truth of the often repeated assertion
that consanguineous marriages lead to deafness and
dumbness, blindness, etc., should be ascertained’
(Darwin, 1887). Although his request for a large-scale
investigation into the possible detrimental effects of
consanguinity was turned down, since then, the effects
of inbreeding have been subject of continuous investigation into a wide range of species, including humans
(Cavalli-Sforza & Bodmer, 1971; Jain, 1976; Thornhill,
1993; Bittles et al., 2002; Keller & Waller, 2002).
Unlike Darwin, we now know that inbred offspring
have an increased probability of receiving the same
deleterious recessive allele from both of their parents and
that they are more homozygous in general, both of
Correspondence: Erik Postma, Institute of Evolutionary Biology and
Environmental Studies, University of Zürich, Winterthurerstrasse 190,
8057 Zürich, Switzerland.
Tel.: +41 44 635 47 66; fax: +41 44 635 68 18;
e-mail: [email protected]
1468
which may result in reduced fitness, i.e. inbreeding
depression (Charlesworth & Charlesworth, 1987).
Although the evidence for inbreeding depression in
plants and nonhuman animals keeps accumulating
(Keller & Waller, 2002), the exact nature of its effect
on fitness in humans, and on human fertility in particular, remains unclear and subject to debate (e.g. Bittles
et al., 2002).
Adverse effects of inbreeding on a couple’s reproductive success have been found in the form of, for example,
an increased rate of miscarriage or infant mortality and
morbidity (Bittles & Neel, 1994; Ober et al., 1999; Bittles,
2001; Jorde, 2001). Interestingly, however, several of the
studies that found a strong effect of inbreeding on
fecundity found that consanguineous couples in fact
have more, rather than fewer, children (Bittles et al.,
1991; Hussain & Bittles, 1999; Blanco Villegas & Fuster,
2007). A number of nonbiological explanations for this,
at first sight counter-intuitive, pattern have been put
forward. For example, related couples may get married at
an earlier age, have a lower level of education or may
gain socio-economic advantages by keeping wealth and
land within their (extended) family (Bittles et al., 1991;
Al Husain & Al Bunyan, 1997; Hussain & Bittles, 1999;
Blanco Villegas & Fuster, 2007). Furthermore, couples
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
Inbred women have fewer children
may have more children to compensate for a higher rate
of infant mortality (referred to as reproductive compensation) (Hussain & Bittles, 1999; Ober et al., 1999; Overall
et al., 2002). Alternatively, this positive association
between inbreeding and fecundity has been argued to
reflect the negative effects of outbreeding depression, in
which the break-up of coadapted gene complexes or
favourable epistatic relationships (Lynch, 1991) and an
increase in maternal–foetal incompatibilities (Philippe,
1974) reduce fitness of more distantly related couples. In
line with this, a number of recent studies have found a
nonlinear relationship between fertility and a couple’s
relatedness, with fertility being maximized at levels of
inbreeding greater than zero, in both humans (Helgason
et al., 2008), nonhuman animals and plants (Edmands,
2007).
Unlike most studies on nonhuman animals, which
focus much of their attention on the effects of inbreeding
at an individual level (and thus the degree of relatedness
of their parents), most studies on the effects of inbreeding
in humans solely test for an effect at the level of the
couple. So, rather than testing for an association between
a parent’s coefficient of inbreeding and fecundity, they
compare the reproductive success of related and unrelated couples. Although this approach will detect effects
of inbreeding on prenatal or neonatal viability, it misses
the non-negligible effects of inbreeding that may only
manifest themselves later in life, for example when
inbred offspring become reproductively active themselves. Indeed, in a long-term study on the effects of
inbreeding in adult Hutterites, an Anabaptist sect, Ober
et al. (1999) found that although more and less inbred
women had the same number of children, inbred women
had longer interbirth intervals, as well as longer intervals
to a recognized pregnancy. Similarly, Robert et al. (2009)
found an effect on family size of the inbreeding level of
the father, at least at later age, in a pre-industrial
Canadian population.
Gaining a better understanding of the effects of
inbreeding in humans requires estimates of inbreeding
depression in relatively outbred and egalitarian populations while accounting for any confounding nonbiological factors as much as possible. Furthermore, rather than
focussing solely on parental relatedness (consanguinity),
the effects of inbreeding throughout the complete life
cycle have to be taken into account. Here, we therefore
use genealogical data from a small and isolated village in
the Swiss Alps and simultaneously test for the effects of
parental relatedness and both maternal and paternal
inbreeding on family size.
Materials and methods
Genealogical data
Using data from, among others, parish and town registries, we were able to reconstruct genealogies for most of
1469
the families living in two adjacent small (between 500
and 700 inhabitants) and isolated villages in the southern
part of the Swiss Alps (Cavergno and Bignasco). It is
worth noting that, compared to bigger cities, such
isolated mountain villages are relatively egalitarian and,
until recently, have seen relatively little social and
economic change.
For the great majority of individuals, we had information on the identity of their father and mother, as well as
on their year of marriage, the number of children they
had and how many of these children reached adulthood.
The total number of children and the number of children
who reached adulthood were highly correlated (Spearman rank correlation: q = 0.91, P < 0.0001) and provided qualitatively and quantitatively very similar
results. As the number of children who reached adulthood is less complete and accurate (for example owing to
emigration), we present here only the results for the total
number of children. Immigrant couples, which may have
had children born outside the village, were not included
in the analyses (also see below).
To anonymize the data, all individuals received a
unique numerical identifier before any analyses were
performed. We included all couples where the woman
was born before 1954 (assuming a maximum age at the
onset of menopause of 55), assuring our estimate of the
number of children refers to completed family sizes.
Those couples where the year of birth of the woman was
not known were included if they had got married before
1974. A total of 28 couples in which either the husband
or the wife had been married previously were excluded.
Estimating relatedness and inbreeding
The probability of two individuals sharing an allele that is
identical by descent (IBD) for a given locus is given by the
coefficient of kinship and provides a measure of the
degree of relatedness or consanguinity of any two
individuals. The probability of an individual inheriting
two alleles that are IBD is equal to the parental coefficient
of kinship and is referred to as an individual’s coefficient
of inbreeding (Falconer & Mackay, 1996, pp. 82–88).
Using ENDOG v4.6 (Gutiérrez & Goyache, 2005), we
calculated for all couples the coefficient of kinship with
their partner (i.e. the coefficient of inbreeding of their
offspring) as measure of their relatedness, here referred
to as R, as well as the coefficient of inbreeding of both
parents (i.e. the coefficient of kinship of the parents’
parents), here referred to as Fmaternal and Fpaternal.
Unlike their molecular counterparts, pedigree-based
estimates of inbreeding and relatedness measure F and R
relative to the individuals in the base population (i.e. all
individuals with unknown parents), which are assumed
to be unrelated and outbred. Consequently, the amount
of information on an individual’s ancestors sets a limit to
the minimum F that can be detected. Consequently,
individuals with F = 0 for which limited pedigree
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
1470
E. POSTMA ET AL.
information is available may still be inbred, and couples
with R = 0 may still be related.
Even to detect the closest form of inbreeding (offspring
from a brother–sister or parent–offspring marriage,
resulting in F = 0.25), we need to be able to go back at
least two generations. It is for this reason that studies
typically only include those individuals for whom all four
grandparents are known. However, in systems in which
such close inbreeding is rare and most inbreeding events
are relatively distant (as is the case in humans), a dataset
including all individuals for which all four grandparents
are known will still contain a very large number of
noninformative zero inbreeding coefficients. While we
may choose to further restrict the data set (e.g. including
only individuals where all eight great-grandparents are
known) to exclude more noninformative inbreeding
coefficients equal to zero, this may also result in the
exclusion of some highly informative close inbreeding
events.
Here, we chose to include all parents and offspring
with F > 0 in our analysis, irrespective of the amount of
pedigree information available. Additionally, we used
ENDOG to calculate for all parents and their offspring the
number of equivalent generations on which their
inbreeding coefficient was based, providing a standardized measure of the amount of pedigree information
available (Maignel et al., 1996; also see González-Recio
et al., 2007). Following González-Recio et al. (2007), we
additionally included all parents and offspring with F = 0
for which all four grandparents were known and for
which the number of equivalent generations was 4 or
more. Note that the results were relatively insensitive to
the exact location of this cut-off value and that any cutoff value between three and five equivalent generations
provided similar results. Furthermore, including only
individuals with F > 0 and ⁄ or couples with R > 0 (so
excluding all F = 0 and R = 0) or excluding all estimates
of F and R based on less than that four equivalent
generations (irrespective of whether F or R are greater
than zero) resulted in smaller datasets but provided
quantitatively very similar results.
Given that, as pointed out earlier, inbreeding coefficients are dependent on pedigree depth. González-Recio
et al. (2007) suggested to regress fitness against the
increase in inbreeding per generation (DF), rather than
against F itself. However, in our case, F and DF were
strongly correlated (r > 0.9), and the use of DF instead of
F provided virtually identical results. As the interpretation of the slope of the regression of fitness against DF (or
DR) is less straightforward, we present the results for F
(and R) only.
Statistical analyses
We tested for an effect of the relatedness of a couple (R,
i.e. the level of inbreeding of their children), as well as
both the paternal and the maternal level of inbreeding
(Fpaternal and Fmaternal, respectively) on the total number
of children they had (Ntotal). Ntotal was square-root
transformed [(Ntotal + 1)], as suggested by a Box–Cox
transformation. Shapiro–Wilk tests for normality showed
no evidence for significant deviations from normality of
the residuals for any of the models.
Analyses using the full data set showed a significant
decline in family size over time (year of marriage:
b ± SE = )0.0030 ± 0.00058, P < 0.001). Year of marriage was therefore included as a covariate in all models.
Furthermore, to account for the possible nonindependence of brothers and sisters (who have the same
inbreeding coefficient and share a household for the
early parts of their lives), we included paternal and
maternal family (defined as a unique combination of the
ID of both parents of the husband and the wife,
respectively) as random effects in all models (irrespective
of their statistical significance). Whereas in the full data
set (so including couples with insufficient pedigree
information, see above) paternal family explained a
significant 31% of variation in Ntotal (v21 = 20.9,
P < 0.001), maternal family explained as little as 6.0%
(v21 = 0.601, P = 0.438). However, although paternal
family explained quantitatively similar amounts of variation in the more restrictive models including R or F (see
below), it never reached statistical significance in these
models.
We first tested simultaneously for an effect of R,
Fmaternal and Fpaternal on Ntotal. Note that sample sizes are
greatly reduced for this model as only couples for which
we have sufficient pedigree information for both the
husband and the wife (and thus for their offspring) could
be included (see above). Additionally, we therefore fitted
models including R, Fhusband and Fwife separately.
All statistical analyses were performed using R version
2.9.1 (R Development Core Team, 2008). Linear mixed
models were run using the lme4 package (Bates et al.,
2008). The significance of the fixed effects as well as their
95% confidence intervals was obtained from 100.000
MCMC simulations, performed using the pvals.fnc function in the languageR package (Baayen, 2007). Significance of the random effects was assessed using likelihood
ratio tests.
Results
Descriptive statistics
In total, we had data on completed family size for 465
couples. The great majority of these couples got married
in the 19th century (earliest: 1745; 1st quartile: 1820;
median: 1851; 3rd quartile: 1894; latest: 1950). The
median completed family size was 4 (minimum: 0; 1st
quartile: 2; 3rd quartile 7; maximum: 16).
The distribution of parental and offspring F, as well as
the amount of pedigree information on which these
estimates are based, is depicted in Fig. 1. It is clear from
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
Inbred women have fewer children
1471
Table 1 Statistical results for a mixed effects model, explaining
variation in the number of children a couple has had by parental
relatedness (R), paternal inbreeding (Fpaternal) and maternal
inbreeding (Fmaternal). Provided are for each covariate the estimate
of the slope and its standard error (b ± SE), its 95% confidence or
credible interval (95% CI) and its level of significance (P). The latter
two are based on MCMC simulations. Also provided is the percentage of variance explained by the two random effects, their 95%
confidence intervals, as well as their significance based on a
likelihood ratio test.
Fixed effects b ± SE
Fig. 1 Inbreeding coefficients plotted against the amount of pedigree information available. Included are the estimated inbreeding
coefficients of all individuals and their offspring (given by the
coefficient of kinship of their parents). The amount of pedigree
information on which an individual’s inbreeding coefficient was
based is expressed in the equivalent number of generations. In grey
are the coefficients of inbreeding that were excluded because of
insufficient pedigree information. For more information, see
Materials and methods.
this figure that, provided that there is sufficient pedigree
information available, all individuals are to some degree
inbred (i.e. F > 0), albeit often only very little. The data
points that were excluded from subsequent analyses
(F = 0 and equivalent number of generations < 4, also
see Materials and methods) are depicted in grey.
The mean inbreeding level across all men was 0.0063
and across those that fitted our criteria for inclusion
mean F was 0.0197 (n = 74). Mean F across all women
was lower (0.0030) but similar when only women with
sufficient pedigree information were included (0.0190,
n = 148). The latter is because there are relatively more
women than men for which we cannot obtain an
informative estimate of F because their parents and
grandparents are unknown. This may be the result of
female-biased immigration or of the fact that a woman’s
birth name was not always recorded when she got
married. Mean R was 0.0065 across all couples and
0.0194 across couples with sufficient pedigree information (n = 153).
)0.0172 ± 0.0036
95% CI
P
)0.023 to )0.0096
Year of
marriage
R
Fpaternal
Fmaternal
< 0.001
)0.706 ± 3.522
3.967 ± 3.928
)10.689 ± 3.703
Random
effects
% variance
explained
95% CI
v21
Paternal
family
Maternal
family
47.2
0 to 0.1810
1.95 0.16
0
0 to 0.1501
0
)8.80 to 5.74
)3.72 to 11.09
)18.31 to )2.78
0.69
0.31
0.0076
P
1.000
tion in family size (Table 1). There was no significant
quadratic effect of year of marriage (P = 0.33), and
inclusion of the quadratic term provided very similar
estimates for the effects of relatedness and inbreeding.
To maximize statistical power, we subsequently tested
for an effect of R, Fpaternal and Fmaternal in three separate
models. Again, we found a significantly negative effect of
maternal
inbreeding
level
on
family
size
(b ± SE = )9.02 ± 3.02, P = 0.0036), but no effect of
parental relatedness (b ± SE = 1.11 ± 2.29, P = 0.55) or
paternal inbreeding level (b ± SE = )1.66 ± 2.40,
P = 0.41). Furthermore, there was no evidence for a
quadratic effect of either parental relatedness (P = 0.83),
maternal inbreeding (P = 0.70) or paternal inbreeding
(P = 0.78), although it should be noted that statistical
power was very low. In all three models, there was again
a highly significant effect of the year of marriage, and
again, paternal but not maternal family explained a
substantial, but nonsignificant, amount of variation in
Ntotal (results not presented).
Discussion
Effects of parental relatedness and inbreeding
When we included R, Fpaternal and Fmaternal in the model,
we found no effect of either parental relatedness or
paternal inbreeding on family size. However, there was a
significantly negative effect of the inbreeding level of the
mother on the total number of children she had
throughout her life (Table 1, Fig. 2). Furthermore, family
sizes were found to have declined significantly over time,
and paternal but not maternal family explained a
substantial (but nonsignificant) proportion of the varia-
Most studies on the effects of inbreeding in humans have
treated the couple as the unit of interest and hence tested
solely for an association between the degree of relatedness of a couple (R) and their reproductive success. This is
in contrast to studies on nonhuman animals, which
typically focus much more on the effects of inbreeding
(F) on an individual level. Here, we combined these two
approaches and tested whether couples in which husband and wife are closely related have fewer children
than more distantly related couples and, at the same
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
1472
E. POSTMA ET AL.
(a)
(b)
(c)
Fig. 2 Effects of relatedness and inbreeding on family size. Whereas there was no effect of (a) parental relatedness or (b) paternal inbreeding,
there was a significantly negative effect of (c) maternal inbreeding on family size. Plotted are the residuals of the full model presented in
Table 1, excluding the covariate depicted.
time, whether there is an effect of the inbreeding levels of
the two parents themselves.
Our finding that inbred mothers have fewer children,
but inbred fathers do not, is in accordance with the fact
that all prenatal and a large proportion of the post-natal
parental investment is provided by the mother. Consequently, the number of genes that are involved in
parental investment and ⁄ or care that is expressed in
women is many times higher than in men, all of which
may carry deleterious recessive alleles that may be
expressed in inbred individuals. This finding is also in
line with studies that have quantified the effects of
paternal and maternal inbreeding in nonhuman animals.
In great tits (Parus major), for example, hatching success
depends not only on the inbreeding level of the egg, but
also on that of the mother (Van Noordwijk & Scharloo,
1981). Similarly, in song sparrows (Melospiza melodia),
inbred females, but not inbred males, have a lower
lifetime reproductive success (Keller, 1998). Although
inbreeding may well have negative effects on sperm
quality and quantity, and thereby reduce fecundity of
inbred males (Roldan et al., 1998; Asa et al., 2007;
Fitzpatrick & Evans, 2009), this may only become
important in species where the level of sperm competition is high (Simmons et al., 2004; Zajitschek et al., 2009).
Furthermore, there is some evidence that the negative
effects of inbreeding on sperm quality or quantity in
humans may only become apparent at a relatively late
age (Robert et al., 2009).
Although the differential effect of maternal and paternal inbreeding makes biological sense, it should be noted
that statistical power is relatively weak (as illustrated by
the 95% confidence intervals in Table 1), and hence, we
cannot exclude the existence of a negative effect of
paternal inbreeding. Furthermore, the absence of an
effect of paternal inbreeding, and of parental relatedness
for that matter, could also be explained by extra-pair
paternity (see Simmons et al., 2004), which introduces
errors in the pedigree along the paternal line. Neverthe-
less, although this may account for the absence of an
effect of paternal inbreeding and parental relatedness, it
cannot explain the significantly negative effect of maternal inbreeding on reproductive success. In fact, pedigree
errors along the paternal line would indirectly also affect
the accuracy of the maternal inbreeding coefficients,
making our estimate of the strength of maternal inbreeding depression, if anything, more conservative.
Our main interest here lies in the evolutionary consequences of inbreeding in humans, i.e. in the association
between parental relatedness and inbreeding, and fitness.
Although we find that inbred women have fewer
children, the data do not allow us to further dissect the
causality of this relationship. Hence, we can only speculate about why maternal inbreeding reduces family size.
Inbred women may, for example, experience a higher
risk of miscarriage or have lower conception rates or
higher peri-implantation loss rates (Ober et al., 1999).
More indirectly, inbreeding may reduce female reproductive success through, for example, the age at marriage. If women who marry at an earlier age have more
children, we may find a negative effect of inbreeding on
family size if inbred women typically marry at a later age.
Indeed, we find a significantly negative effect of the
maternal age at marriage on completed family size
()0.041 ± 0.0079, P < 0.0001), but not of paternal age
at marriage ()0.0086 ± 0.0065, P = 0.51). Yet, there is
only a weak and nonsignificant positive effect of
inbreeding
on
maternal
age
at
marriage
(b = )0.061 ± 0.039, P = 0.20) (note that age at marriage
was reciprocally transformed to improve normality).
Furthermore, when we include maternal age at marriage
as an additional covariate in the model testing for an
effect of maternal inbreeding on family size, the effect of
maternal inbreeding is reduced only slightly (maternal
inbreeding: b = )7.24 ± 3.10, P = 0.022; maternal age at
marriage: b = )19.08 ± 14.8, P = 0.046). Although this
suggests that age at marriage might at least partly
contribute to the negative effect of inbreeding on
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
Inbred women have fewer children
reproductive success in women, again we can only
speculate about the underlying mechanism.
Unlike earlier studies on the effects of inbreeding in
humans, we did not find that related couples had fewer
(or more) children (Bittles & Neel, 1994; Bittles, 2001;
Jorde, 2001). However, the great majority of studies on
the effects of inbreeding in humans have been performed
in communities with exceptionally high rates of inbreeding in, among others, India (mean F = 0.01–0.05) and
Pakistan (mean F = 0.02–0.03) (Bittles et al., 2002).
Importantly, these countries are also characterized by
large amounts of, potentially confounding, variation in
socio-economic status (Hussain & Bittles, 1999, 2002;
Bittles et al., 2002). Estimates of the effect of inbreeding
in less inbred and more homogeneous societies remain
however scarce (but see Ober et al., 1999; Helgason et al.,
2008; Robert et al., 2009). The few studies using data
from populations more similar to the one described here
(i.e. more homogenous and egalitarian, and less inbred)
typically found small and often complex effects of a
couple’s relatedness on total number of children or no
effect at all (Ober et al., 1999; Overall et al., 2002;
Helgason et al., 2008; Robert et al., 2009). The absence
of a strong negative effect of inbreeding in this study may
at least partly be attributed to reproductive compensation, especially if family sizes are smaller than the
reproductive maximum (Ober et al., 1999; Overall et al.,
2002). However, our finding that more inbred mothers
do indeed have fewer children, and thus do not show
complete compensation, suggests that reproductive compensation may not be the sole explanation. Furthermore,
several recent studies on humans and other animals have
found significantly nonlinear effects of relatedness on
reproductive success, with reproductive success being
maximized at intermediate levels of relatedness
(Edmands, 2007; Helgason et al., 2008). Although here
we do not find any significantly nonlinear effects,
statistical power is very low.
Although it will never be possible to completely rule
out the role of nonbiological factors in driving the
observed negative effect of maternal inbreeding found
here, it should first of all be noted that, except for a few
cases (two men, who were brothers, and one woman with
F = 0.125), we find very little close inbreeding, and many
couples with R > 0 would probably have been unaware of
the fact that they were related. Given that most inbreeding events are relatively distant, correlations between
their relatedness and their socio-economic status are
expected to be very weak at most. Second, we tested for
an effect of parental relatedness and both maternal and
paternal inbreeding simultaneously. As we can expect
socio-economic and other nonbiological effects (e.g.
giving rise to systematic year-to-year differences in
inbreeding level and family size) to act on the family
level (i.e. on both the wife and the husband), we would
expect to find similar effects of maternal and paternal
inbreeding, and of parental relatedness. Third and finally,
1473
by including maternal and paternal family as random
effects in our model, we were able to control at least partly
for the effects of variation in cultural or socio-economic
differences among families. With respect to the latter, it is
interesting to note that whereas we found paternal family
to explain a substantial and significant proportion of the
variation in family size, there was no effect of the
maternal family on a couple’s reproductive success. In
other words, brothers had similar family sizes, whereas
sisters did not. One potential explanation for this pattern
could be the inheritance system in which sons benefit
more from the wealth of their parents than daughters.
Although social scientists continue to prefer nonbiological explanations (e.g. Shor & Simchai, 2009), evidence in support for the idea that the taboo on incest that
exists in many societies has evolved as a mechanism to
avoid inbreeding depression continues to accumulate
(Maynard Smith, 1978; Lieberman et al., 2003). However, whereas the latter assumes that mating with a
relative results in fewer offspring, we found that closely
related partners did not have smaller families than those
that were more distantly related. Importantly, however,
we found that the daughters of parents who are closely
related (that is daughters who were inbred themselves)
did have fewer children than those with less closely
related parents. Consequently, although related parents
do not have fewer children, their inbred daughters do,
providing further evidence for selection for inbreeding
avoidance, also in humans.
Acknowledgments
We are grateful to Lukas Keller and Barbara Tschirren for
stimulating discussions and constructive comments. EP is
funded by the Swiss National Science Foundation (SNSF)
(grant 31003A-116794).
References
Al Husain, M. & Al Bunyan, M. 1997. Consanguineous
marriages in a Saudi population and the effect of inbreeding
on prenatal and postnatal mortality. Ann. Trop. Paediatr. 17:
155–160.
Asa, C., Miller, P., Agnew, M., Rebolledo, J.A.R., Lindsey, S.L.,
Callahan, M. & Bauman, K. 2007. Relationship of inbreeding
with sperm quality and reproductive success in Mexican gray
wolves. Anim. Conserv. 10: 326–331.
Baayen, R.H. 2007. Analyzing Linguistic Data: A Practical Introduction to Statistics Using R. Cambridge University Press,
Cambridge.
Bates, D.M., Maechler, M. & Dai, B. (2008) lme4: Linear MixedEffects Models Using S4 Classes. R package version 0.99937528. http://lme4.r-forge.r-project.org/.
Bittles, A.H. 2001. Consanguinity and its relevance to clinical
genetics. Clin. Genet. 60: 89–98.
Bittles, A.H. & Neel, J.V. 1994. The costs of human inbreeding
and their implications for variations at the DNA level. Nat.
Genet. 8: 117–121.
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY
1474
E. POSTMA ET AL.
Bittles, A.H., Mason, W.M., Greene, J. & Rao, N.A. 1991.
Reproductive behavior and health in consanguineous marriages. Science 252: 789–794.
Bittles, A.H., Grant, J.C., Sullivan, S.G. & Hussain, R. 2002. Does
inbreeding lead to decreased human fertility? Ann. Hum. Biol.
29: 111–130.
Blanco Villegas, M.J. & Fuster, V. 2007. Differential reproductive
pattern in a rural Spanish region (La Cabrera, Leon): consequences for potential natural selection. Ann. Hum. Biol. 34:
664–672.
Burkhardt, F., Harvey, J., Porter, D.M. & Topham, J.R. (eds)
1997. The Correspondence of Charles Darwin – Volume 10, 1862.
Cambridge University Press, Cambridge.
Cavalli-Sforza, L.L. & Bodmer, W.F. 1971. The Genetics of Human
Populations. Freeman, San Francisco.
Charlesworth, D. & Charlesworth, B. 1987. Inbreeding depression and its evolutionary consequences. Annu. Rev. Ecol. Syst.
18: 237–268.
Darwin, F. (ed.) 1887. The Life and Letters of Charles Darwin,
Including an Autobiographical Chapter. John Murray, London.
Edmands, S. 2007. Between a rock and a hard place: evaluating
the relative risks of inbreeding and outbreeding for conservation and management. Mol. Ecol. 16: 463–475.
Falconer, D.S. & Mackay, T.F.C. 1996. Introduction to Quantitative
Genetics, 4th edn. Longman, Harlow.
Fitzpatrick, J.L. & Evans, J.P. 2009. Reduced heterozygosity
impairs sperm quality in endangered mammals. Biol. Lett. 5:
320–323.
González-Recio, O., López de Maturana, E. & Gutiérrez, J.P.
2007. Inbreeding depression on female fertility and calving
ease in Spanish dairy cattle. J. Dairy Sci. 90: 5744–5752.
Gutiérrez, J.P. & Goyache, F. 2005. A note on ENDOG:
a computer program for analysing pedigree information.
J. Anim. Breed. Genet. 122: 172–176.
Helgason, A., Palsson, S., Guobjartsson, D.F., Kristjansson, P. &
Stefansson, K. 2008. An association between the kinship and
fertility of human couples. Science 319: 813–816.
Hussain, R. & Bittles, A.H. 1999. Consanguineous marriage and
differentials in age at marriage, contraceptive use and fertility
in Pakistan. J. Biosoc. Sci. 31: 121–138.
Hussain, R. & Bittles, A.H. 2002. Assessment of association
between consanguinity and fertility in Asian populations.
J. Health Popul. Nutr. 22: 1–12.
Jain, S.K. 1976. Evolution of inbreeding in plants. Annu. Rev.
Ecol. Syst. 7: 469–495.
Jorde, L.B. 2001. Consanguinity and prereproductive mortality
in the Utah Mormon population. Hum. Hered. 52: 61–65.
Keller, L.F. 1998. Inbreeding and its fitness effects in an insular
population of song sparrows (Melospiza melodia). Evolution 52:
240–250.
Keller, L.F. & Waller, D.M. 2002. Inbreeding effects in wild
populations. Trends Ecol. Evol. 17: 230–241.
Lieberman, D., Tooby, J. & Cosmides, L. 2003. Does morality
have a biological basis? An empirical test of the factors
governing moral sentiments relating to incest Proc. R. Soc. Lond.
B 270: 819–826.
Lynch, M. 1991. The genetic interpretation of inbreeding
depression and outbreeding depression. Evolution 45: 622–629.
Maignel, L., Boichard, D. & Verrier, E. 1996. Genetic variability
of French dairy breeds estimated from pedigree information.
Interbull Bull. 14: 49–54.
Maynard Smith, J. 1978. The Evolution of Sex. Cambridge
University Press, Cambridge.
Ober, C., Hyslop, T. & Hauck, W.W. 1999. Inbreeding effects on
fertility in humans: evidence for reproductive compensation.
Am. J. Hum. Genet. 64: 225–231.
Overall, A.D.J., Ahmad, M. & Nichols, R.A. 2002. The effect of
reproductive compensation on recessive disorders within
consanguineous human populations. Heredity 88: 474–479.
Philippe, P. 1974. Amenorrhea, intrauterine mortality and
parental consanguinity in an isolated French Canadian population. Hum. Biol. 46: 405–424.
R Development Core Team 2008. R: A Language and Environment
for Statistical Computing. R Foundation for Statistical Computing, Vienna.
Robert, A., Toupance, B., Tremblay, M. & Heyer, E. 2009. Impact
of inbreeding on fertility in a pre-industrial population. Eur. J.
Hum. Genet. 17: 673–681.
Roldan, E.R.S., Cassinello, J., Abaigar, T. & Gomendio, M.
1998. Inbreeding, fluctuating asymmetry, and ejaculate
quality in an endangered ungulate. Proc. R. Soc. Lond. B
265: 243–248.
Shor, E. & Simchai, D. 2009. Incest avoidance, the incest taboo,
and social cohesion: revisiting Westermarck and the case of
the Israeli kibbutzim. Am. J. Sociol. 114: 1803–1842.
Simmons, L.W., Firman, R.C., Rhodes, G. & Peters, M. 2004.
Human sperm competition: testis size, sperm production and
rates of extrapair copulations. Anim. Behav. 68: 297–302.
Thornhill, N.W. (ed.) 1993. The Natural History of Inbreeding and
Outbreeding. University of Chicago Press, Chicago.
Van Noordwijk, A.J. & Scharloo, W. 1981. Inbreeding in an
island population of the great tit. Evolution 35: 674–688.
Zajitschek, S.R.K., Lindholm, A.K., Evans, J.P. & Brooks, R.C.
2009. Experimental evidence that high levels of inbreeding depress sperm competitiveness. J. Evol. Biol. 22: 1338–
1345.
Received 29 December 2009; revised 18 March 2010; accepted 6 April
2010
ª 2010 THE AUTHORS. J. EVOL. BIOL. 23 (2010) 1468–1474
JOURNAL COMPILATION ª 2010 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY