?  2006 Nature Publishing Group 
 
Outbred embryos rescue inbred half-siblings in
mixed-paternity broods of live-bearing females
Jeanne A. Zeh1 & David W. Zeh1
Females commonly mate with more than one male1, and poly-
andry has been shown to increase reproductive success in many
species2?4. Insemination by multiple males shifts the arena for
sexual selection from the external environment to the female
reproductive tract, where sperm competition or female choice of
sperm could bias fertilization against sperm from genetically
inferior5 or genetically incompatible males6,7. Evidence that
polyandry can be a strategy for avoiding incompatibility comes
from studies showing that inbreeding cost is reduced in some
egg-laying species by postcopulatory mechanisms that favour
fertilization by sperm from unrelated males8?10. In viviparous
(live-bearing) species, inbreeding not only reduces offspring
genetic quality but might also disrupt feto-maternal interactions
that are crucial for normal embryonic development11?13. Here we
show that polyandry in viviparous pseudoscorpions reduces
inbreeding cost not through paternity-biasing mechanisms
favouring outbred offspring, but rather because outbred embryos
exert a rescuing effect on inbred half-siblings in mixed-paternity
broods. The benefits of polyandry may thus be more complex for
live-bearing females than for females that lay eggs.
Consanguineous mating is a significant source of fitness
depression, and inbred offspring are more likely to be homozygous
either for deleterious, recessive alleles or at loci with heterozygote
advantage14. In live-bearing species, inbreeding can also disrupt the
complex sequence of two-way, immunological interactions between
fetal and maternal tissues15. Unfortunately, investigating inbreeding
effects on fetal loss is hindered in most viviparous animals by
the intrusive methods required to detect early stage, spontaneous
abortion. Unlike most terrestrial arthropods, pseudoscorpions are
viviparous. Embryos develop in an external, translucent brood sac
and draw nutritive fluid from the mother?s reproductive tract16.
Previous research exploited this ?external-womb? form of viviparity
to establish that polyandry in the pseudoscorpion, Cordylochernes
scorpioides, significantly enhances female lifetime reproductive
success by reducing the rate of spontaneous abortion of entire
broods17. Here we investigate the effect of inbreeding on abortion
rate and reproductive success in C. scorpioides, and assess whether
females reduce this cost by mating with both a related and an
unrelated male.
Virgin females were randomly assigned to one of five treatments,
four involving a single mating with each of two males: two non-
siblings of the female (NN), two full-siblings of the female (SS), a
non-sibling first male and a full-sibling second male (NS), or vice
versa (SN). Unlike the SS treatment, male pairs in NN replications
were non-brothers, the usual situation for polyandrous females in
nature18. For a conservative comparison, a fifth treatment included
females that were mated twice to a single non-sibling (N; see
Methods). Each mated female was monitored until she gave birth
to a first brood of nymphs, spontaneously aborted her first brood, or
failed to produce a brood in 45 d. All nymphs were counted at birth,
and DNA profiling19 was used to assign paternity for NS and SN
nymphs. A second experiment assessed the effect of relatedness on
number of sperm males allocated to spermatophores. Once-mated
males were paired with a full-sibling (S female) or a non-relative
(N female; n ? 12 per treatment), the sperm packet was collected,
and sperm were counted17.
A randomized block analysis of covariance (ANCOVA) showed a
significant effect of mating treatment (F4,187 ? 8.45, P , 0.001) and
female body size (F1,192 ? 7.55, P ? 0.007) on reproductive success,
that is, number of nymphs born. Comparison of treatment means
(Fig. 1) showed that female reproductive success sorted into three
levels, SS , NS ? SN ? N , NN, a grouping that was highly sig-
nificant (contrast analysis20, F 1,187 ? 22.12, P , 0.001), and
accounted for 65% of the mating treatment effect. Underlying this
pattern was a significant treatment effect on the frequency of
spontaneous brood abortion (Fisher exact test, P ? 0.029), with SS
females suffering the highest abortion rate (40%; see Fig. 1). Mating
treatment also significantly influenced the number of nymphs born
to females that successfully carried broods to term (F4,144 ? 7.10,
P , 0.001); the SS , NS ? SN ? N , NN pattern was again sig-
nificant (F 1,144 ? 13.93, P , 0.001), accounting for 49% of the
treatment effect. Thus, mating with a non-relative and a brother
diminished cost of inbreeding by both reducing the risk of spon-
taneous brood abortion and increasing embryonic survival in suc-
cessful broods. In the second experiment, relatedness had no effect on
number of sperm allocated to spermatophores (S-female,
LETTERS
Figure 1 | Effect of mating treatment on brood production. Rates of whole-
brood spontaneous abortion (filled bars) and number of nymphs
(mean ^ s.e.m.) born to females carrying broods to term (hatched bars) for
treatments in which females were mated to two full-siblings (sib ? sib), to a
non-relative followed by a full-sibling (non ? sib), to a full-sibling followed
by a non-relative (sib ? non), to a single non-relative (non), or to two non-
relatives (non ? non).
1Department of Biology and Program in Ecology, Evolution and Conservation Biology, University of Nevada, Reno, Nevada 89557, USA.
Vol 439|12 January 2006|doi:10.1038/nature04260
201
?  2006 Nature Publishing Group 
 
1,674 ^ 147; N-female, 1,637 ^ 151 (mean ^ s.e.m.); t22 ? 0.18,
P ? 0.86).
The SS , NS ? SN ? N , NN pattern of female reproductive
success is consistent with the hypothesis that polyandry enables
C. scorpioides females to bias paternity against sibling males. How-
ever, paternity assignment for NS and SN nymphs showed that this
was not the case. Regardless of mating order, females gave birth to
more inbred than outbred offspring. The proportion of NS and SN
inbred nymphs was 0.66 ^ 0.17 and 0.60 ^ 0.24 (mean ^ 95%
confidence interval), respectively, and did not differ significantly
between treatments (F1,31 ? 0.19, P ? 0.668). The pooled mean of
63% (^14%) suggests a sperm-competition advantage for sibling
males. Because of higher mortality of inbred embryos, this at-birth
value of 63% underestimates the actual fertilization success of sibling
males. After correcting for embryonic survival differences (Sup-
plementary Information), sibling-male fertilization success was
more than twice that of non-relative males (69% ^ 14% versus
31% ^ 14%, P ? 0.011). Although the mechanism underlying the
sperm-competition advantage is unknown, genetic similarity could
feasibly lower the risk for sibling-male sperm of being targeted by
female immune defences.
This sibling-male, sperm-competition advantage can have dele-
terious, postfertilization consequences for female reproductive suc-
cess. In 12 cases, females produced only inbred nymphs (Fig. 2).
These females gave birth to significantly fewer nymphs
(35.25 ^ 4.96; mean ^ s.e.m.) than females whose broods included
at least one outbred nymph (54.90 ^ 3.84; t30 ? 3.13, P ? 0.004).
The high rate of spontaneous brood abortion and small brood size
associated with inbreeding might result from males donating less
sperm and/or accessory gland proteins (Acps) to related females21;
however, the evidence available does not support this hypothesis.
Males did not discriminate against sisters, either in propensity to
produce a spermatophore or in number of sperm allocated to sperm
packets. Adjustment of Acp quality and/or quantity independently of
sperm number would therefore be required for Acp-mediated treat-
ment effects on female reproductive success. Moreover, although
seminal products stimulate ovulation in some live-bearing species,
mating treatment had no effect on whether C. scorpioides females
produced a brood (Fisher exact test, P ? 0.2150). Finally, with a
mean ejaculate mass estimated at 1 mg and a mean brood mass of
8,800 mg just before birth, the material contribution made by males
seems negligible.
Our findings show that polyandry can reduce the high rate of
spontaneous brood abortion and the low reproductive success
associated with inbreeding, if inbred embryos develop with outbred
half-siblings. This rescuing effect can be quantified by considering
reproductive success from a sibling-male perspective (Supplemen-
tary Information). A male that mates with his sister sires nearly twice
as many inbred offspring that survive to birth if she also mates with
an unrelated male (NS and SN treatments, 21.52 ^ 2.28; mean ^
s.e.m.), as compared with a second brother (SS treatment,
12.70 ^ 2.55; Mann?Whitney test, U ? 1097.5, P ? 0.021). Indeed,
despite losing about a third of fertilizations to the unrelatedmale, the
reproductive success of the single, sibling male in NS and SN
replications is statistically indistinguishable from the combined
reproductive success (23.12 ^ 3.86) of the two brothers in SS
replications (Mann?Whitney test, U ? 1478, P ? 0.990), owing to
the reduced rate of spontaneous brood abortion.
How might rescuing of inbred embryos occur? With inbreeding,
homozygosity for a deleterious, recessive allele would result in
intrinsically weak offspring that are unable to sequester adequate
maternal resources. By theirmore vigorous activity, outbred embryos
might draw sufficient nutrients into the communal brood sac to
ensure development of the brood to birth. Alternatively, genetic
similarity of inbred embryos to their mother may blur the self/
non-self distinction essential for innate immune responses. Innate
immunity, common to invertebrates and vertebrates22, has an
important role in fetal loss, with perturbations of immune responses
triggering spontaneous abortion23. By establishing a non-self
presence in mixed inbred/outbred broods and activating the normal
cascade of feto-maternal interactions, outbred embryos might com-
pensate for attenuation of interactions between mother and inbred
concepti. Notably, both these hypotheses account for the variation in
inbreeding cost apparent between sibling pairs in Figs 1 and 2. At any
locus, whether it influences embryonic viability or is involved in
recognition, mendelian genetics dictate that full-siblings may share
one, both or no alleles. The fitness consequences of inbreeding will
thus vary, depending on the multilocus genotypes of the particular
brother and sister involved.
Our results are consistent with the view that parent?offspring
conflict over maternal resource allocation drives the evolution of
feto-maternal interactions, with resistance by the mother to manipu-
lation of her reproductive physiology by the embryo creating tension
that is crucial for normal embryonic development24. Relaxation of
this tension, through increased homozygosity or genetic similarity,
could explain both the high rate of spontaneous inbred-brood
abortion and the rescuing effect of outbred, half-siblings detected
in our study.
So far, evidence that polyandry reduces inbreeding costs has
been mixed, leading some to question the importance of genetic
incompatibility avoidance in the evolution of polyandry25. However,
most counterevidence comes from egg-laying species, with females
limited to pre-fertilizationmechanisms for inbreeding avoidance.We
have shown in a viviparous species that mixed paternity rather
than paternity biasing reduces inbreeding cost. If mixed-paternity
rescuing effects occur in other live-bearing species, reproductive
mode may well be an important factor influencing not only the
potential sources of incompatibility between paternal and maternal
genomes6, but also the range of postcopulatory defences available to
polyandrous females.
METHODS
Pseudoscorpions. Pseudoscorpions were laboratory-reared F2 and F3 descen-
dents of 82 field-inseminatedC. scorpioides females collected in Panama; nymphs
Figure 2 |Number of nymphs born versus proportion outbred in successful
NS and SN broods. Cubic spline27, standard and Gentleman?Givens
regressions26 all yielded a significant quadratic relationship between nymph
number and proportion sired by the non-relative male in non?sib
(triangles) and sib ? non (circles) treatments. The unbroken line indicates
the best-fit cubic spline solution. Broken lines indicate ^1 s.e.m., based on
1,000 bootstrap replicates. For standard, quadratic regression, R2 ? 0.239,
P ? 0.019. Neither linear nor quadratic regression was significant when
exclusively inbred broods were excluded (R2 ? 0.009, P ? 0.688 and
R2 ? 0.013, P ? 0.897, respectively).
LETTERS NATURE|Vol 439|12 January 2006
202
?  2006 Nature Publishing Group 
 
were reared individually to ensure virginity17. To generate full-sibling families, F1
females were randomly mated to a single, unrelated male, and progeny were
reared to adults. These F2 individuals either were used in block 1 or were mated
to provide F3 individuals for block 2.
Polyandry and inbreeding experiment. Within blocks, a completely random-
ized mating design was used to maximize sample size. Before the experiment,
each male was mated once to a non-experimental female. Each mating pair was
placed in a 28-mm-diameter arena, interactions were videotaped for 30min17,
and pseudoscorpions were then returned to their vials. After 48 h, females were
presented with the same male (N treatment) or a different male (NN, NS,
SN and SS treatments). Only females that unambiguously accepted a sperm
packet in both matings were retained for subsequent monitoring (n ? 41 for
the N treatment; n ? 39 for NN; n ? 39 for NS; n ? 36 for SN, and n ? 44
for SS).
Monitoring female reproductive success. Females were maintained in trans-
parent vials containing Ficus frass, were fed nine Tribolium confusum larvae per
week, and were checked every 3?4 d, increasing to daily at late stages of gestation.
During gestation, females remain in a silken nest constructed on the vial wall,
enabling non-intrusive monitoring of brood development17. After a 14-d
developmental period, nymphs are born simultaneously and remain in the
nest for 2?3 d. Within 24 h of birth, nymphs were counted, and NS and SN
nymphs were frozen for paternity testing. Females were assigned a reproductive
success score of 0 if they aborted their entire first brood. Females not producing a
brood sac in 45 d (n ? 11) were excluded from the reproductive success analyses,
because such females do not become gravid17.
Statistical analyses were done with SAS26. Cephalothorax length, the trait
most strongly correlated with female lifetime reproductive success17, was used to
control for body size in the ANCOVA analysis of female reproductive success (see
text). For the analysis of the complete data set, a non-parametric ANCOVAwas
done on rank-transformed reproductive success data and residuals were nor-
mally distributed (Shapiro-Wilk test, W ? 0.989, P ? 0.157). For the analysis
excluding cases of spontaneous abortion, data transformation was not required
(W ? 0.990, P ? 0.403).
Minisatellite paternity assignment. Amplification by PCR of alleles at the
cCscMS23 minisatellite locus (heterozygosity ? 0.99)19 was used to assign
paternity for 465 nymphs from 16 replications of each of the NS and SN
treatments. PCR products from the mother, the two putative sires and 8?20
offspring were run on agarose gels stained with ethidium bromide to visualize
alleles. Paternity was assigned on the basis of the presence of unique paternal
alleles in offspring. Only replications in which the putative sires shared no alleles
were scored (32 out of 34 replications tested).
Several measures ensured amplification of both alleles from heterozygous
individuals (Supplementary Information). First, a strict consensus primer pair
was designed from a large sample of cCscMS23 alleles (n ? 22). Second, the high
primer-pair Tm enabled a robust, two-step PCR reaction. Finally, PCR was done
with a mixture of DNA polymerases capable of amplifying high molecular mass
alleles (to 22 kb) from femtogram quantities of DNA.
Received 14 April; accepted 21 September 2005.
1. Jennions, M. D. & Petrie, M. Why do females mate multiply? A review of the
genetic benefits. Biol. Rev. Camb. Phil. Soc. 75, 21?-64 (2000).
2. Hasselquist, D., Bensch, S. & von Schantz, T. Correlation between male song
repertoire, extra-pair paternity and offspring survival in the great reed warbler.
Nature 381, 229?-232 (1996).
3. Tregenza, T. & Wedell, N. Polyandrous females avoid the costs of inbreeding.
Nature 415, 71?-73 (2002).
4. Stockley, P. Female multiple mating behaviour, early reproductive failure and
litter size variation in mammals. Proc. R. Soc. Lond. B 270, 271?-278 (2003).
5. Birkhead, T. R., M?ller, A. P. & Sutherland, W. J. Why do females make it so
difficult for males to fertilize their eggs? J. Theor. Biol. 161, 51?-60 (1993).
6. Zeh, J. A. & Zeh, D. W. Reproductive mode and the genetic benefits of
polyandry. Anim. Behav. 61, 1051?-1063 (2001).
7. Tregenza, T. & Wedell, N. Genetic compatibility, mate choice and patterns of
parentage: invited review. Mol. Ecol. 9, 1013?-1027 (2000).
8. Olsson, M., Shine, R., Madsen, T., Gullberg, A. & Tegelstro?m, H. Sperm
selection by females. Nature 383, 585 (1996).
9. Foerster, K., Delhey, K., Johnsen, A., Lifjeld, J. T. & Kempenaers, B. Females
increase offspring heterozygosity and fitness through extra-pair matings.
Nature 425, 714?-717 (2003).
10. Bretman, A., Wedell, N. & Tregenza, T. Molecular evidence of post-copulatory
inbreeding avoidance in the field cricket Gryllus bimaculatus. Proc. R. Soc. Lond. B
271, 159?-164 (2004).
11. Ober, C. et al. Variation in the HLA-G promoter region influences miscarriage
rates. Am. J. Human Genet. 72, 1425?-1435 (2003).
12. Choudhury, S. R. & Knapp, L. A. Human reproductive failure II: immunogenetic
and interacting factors. Human Reprod. Update 7, 135?-160 (2001).
13. Davies, C. J. et al. Major histocompatibility antigen expression on the bovine
placenta: its relationship to abnormal pregnancies and retained placenta. Anim.
Reprod. Sci. 82?-83, 267?-280 (2004).
14. Keller, L. F. & Waller, D. M. Inbreeding effects in wild populations. Trends Ecol.
Evol. 17, 230?-241 (2002).
15. Hill, J. A. in Immunology of Human Reproduction (eds Kurpisz, M. & Fernandez,
N.) 401?-424 (Bios Scientific, Oxford, UK, 1995).
16. Weygoldt, P. The Biology of Pseudoscorpions (Harvard Univ. Press, Cambridge,
MA, 1969).
17. Newcomer, S. D., Zeh, J. A. & Zeh, D. W. Genetic benefits enhance the
reproductive success of polyandrous females. Proc. Natl Acad. Sci. USA 96,
10236?-10241 (1999).
18. Zeh, D. W., Zeh, J. A. & Bermingham, E. Polyandrous, sperm-storing females:
carriers of male genotypes through episodes of adverse selection. Proc. R. Soc.
Lond. B 264, 119?-125 (1997).
19. Zeh, D. W., Zeh, J. A. & May, C. A. Charomid cloning vectors meet the
pedipalpal chelae: single-locus minisatellite DNA probes for paternity
assignment in the harlequin beetle-riding pseudoscorpion. Mol. Ecol. 3,
517?-522 (1994).
20. Rosenthal, R. & Rosnow, R. L. Contrast Analysis. Focused Comparisons in the
Analysis of Variance (Cambridge Univ. Press, Cambridge, UK, 1985).
21. Simmons, L. W. The evolution of polyandry: an examination of the genetic
incompatibility and good-sperm hypotheses. J. Evol. Biol. 14, 585?-594 (2001).
22. Hoffmann, J. A. The immune response of Drosophila. Nature 426, 33?-38
(2003).
23. Caucheteux, S. M., Kanellopoulos-Langevin, C. & Ojcius, D. M. At the innate
frontiers between mother and fetus: linking abortion with complement
activation. Immunity 18, 169?-172 (2003).
24. Haig, D. Genetic conflicts in human pregnancy. Q. Rev. Biol. 68, 495?-532
(1993).
25. Jennions, M. D., Hunt, J., Graham, R. & Brooks, R. No evidence for inbreeding
avoidance through postcopulatory mechanisms in the black field cricket,
Teleogryllus commodus. Evolution 58, 2472?-2477 (2004).
26. SAS Institute, The SAS System for Windows, Release 8.02 (SAS Institute, Cary,
2001).
27. Schluter, D. & Nychka, D. Exploring fitness surfaces. Am. Nat. 143, 597?-616
(1994).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank M. Bonilla and S. Trimmer for assistance;
K. Panter for discussion; the Smithsonian Tropical Research Institute for
logistical support; and La Autoridad Nacional del Ambiente for permission to
collect in Panama. This research was supported by grants from the US National
Science Foundation and the National Geographic Society.
Author Contributions J.A.Z. designed the study, carried out the mating and
rearing experiments, and took the primary role in writing the paper. D.W.Z.
performed the molecular and statistical analyses.
Author Information Reprints and permissions information is available at
npg.nature.com/reprintsandpermissions. The authors declare no competing
financial interests. Correspondence and requests for materials should be
addressed to J.A.Z. (jaz@unr.edu).
NATURE|Vol 439|12 January 2006 LETTERS
203