ARTICLE IN PRESS

Journal of Theoretical Biology 250 (2008) 524–531

www.elsevier.com/locate/yjtbi

A model for the evolutionary maintenance of monogyny in spiders

Lutz Fromhagea,, John M. McNamarab, Alasdair I. Houstona
a
School of Biological Sciences, University of Bristol, Woodland Road, Bristol BS8 1UG, UK
b
Department of Mathematics, University of Bristol, University Walk, Bristol BS8 1TW, UK
Received 10 July 2007; received in revised form 8 October 2007; accepted 8 October 2007
Available online 11 October 2007

Abstract

Sexual selection theory predicts that males should attempt to mate with several females, unless the benefits of male promiscuity are
trumped by alternative benefits associated with male monogamy (monogyny). Here we use a game theory model to address the adaptive
value of a monogynous strategy, which has the sole benefit of enhancing a male’s paternity share in the context of competition with other
males. We consider two ways in which monogynists might enhance their paternity: by outcompeting rival ejaculates in sperm competition,
and by reducing the probability that a female remates with rival males. The model is based on the biology of some particularly well-studied
spider species, in which males are morphologically restricted to mate with either one or at most two females in their lifetime. Our results
suggest that, regardless of the mechanism of paternity enhancement involved, a male-biased sex ratio is generally required for the evolution
and maintenance of monogyny. Moreover, we show that there is a large region of parameter space where monogyny and bigyny can coexist
as alternative mating strategies under negative frequency dependent selection. There is also a narrow range of conditions where either
monogyny or bigyny can be evolutionarily stable. Our results are in qualitative agreement with empirical findings in spiders.
r 2007 Elsevier Ltd. All rights reserved.

Keywords: Mating strategy; Monogamy; Paternity; Game theory; Sex ratio

1. Introduction repeated copulations (Wickler and Seibt, 1981). A second

Classical sexual selection theory predicts that males
should typically maximize their fitness by mating with
several females (Bateman, 1948; Trivers, 1972). Exceptions
to this rule, however, can evolve in species where the
benefits of male promiscuity are trumped by benefits to
males that focus their efforts on a single female. The
benefits associated with male monogamy (monogyny) can
be broadly divided in two classes: first, monogynous males
may increase their reproductive success by increasing the
number of surviving offspring of their mate. They may
achieve this by providing a parental investment (i.e., by
supplying the female and/or her offspring with care and
resources; Clutton-Brock, 1991; Fromhage et al., 2007b;
Trivers, 1972), and by ensuring female fertility through
Corresponding author. Tel.: +44 1179545945; fax: +44 1179257374.
E-mail addresses: lutzfromhage@web.de (L. Fromhage),
John.McNamara@bristol.ac.uk (J.M. McNamara),
A.I.Houston@bristol.ac.uk (A.I. Houston).
possibility, on which we focus in the present study, is that
monogynous males may enhance their paternity share in
the face of competition by other males. Although parental
investment and paternity enhancement are not mutually
exclusive activities (Kvarnemo, 2006), a promising ap-
proach to studying their significance is to focus on
relatively simple systems in which only one of these
activities is relevant in the absence of the other.
Here we use a game theory model to address the adaptive
value of a monogynous strategy which has the sole benefit
of enhancing a male’s paternity share in the context of
competition with other males. Because such behavior has
been described in several particularly well-studied species of
spiders (see Andrade and Kasumovic, 2005 for an overview),
we couch our model in spider terms. Spider males have paired
copulatory organs, the pedipalps. Although females in
entelegyne spiders have two separate genital openings, a
single pedipalp insertion into one of these openings can
fertilize a female’s lifetime production of eggs (Andrade and
Banta, 2002; Schneider et al., 2005). In several species where

0022-5193/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jtbi.2007.10.008
 ARTICLE IN PRESS
L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531
monogynous mating behavior occurs, the pedipalps regularly
become damaged (or depleted of sperm) as a result of
copulation, so that each pedipalp can be used for one
copulation only (Andrade and Banta, 2002; Foellmer and
Fairbairn, 2003; Fromhage and Schneider, 2005b; Herber-
stein et al., 2005b; but see Breene and Sweet, 1985). Given this
constraint, males have only two options: either to use both
pedipalps to copulate with the same female (monogyny), or to
use each pedipalp to copulate with a different female
(bigyny). Although a constraint limiting males to a maximum
of two copulations may not necessarily apply in situations
where monogyny evolves de novo (because this constraint
may itself be a consequence of selection for monogyny;
cf. Fromhage et al.’s (2005) approach of comparing mono-
gyny with a polygynous strategy of multiple mating), we here
take this constraint as given, thus focussing on a comparison
between the adaptive value of monogyny versus bigyny. By
doing so, we aim to improve the understanding of mating
systems in which this constraint currently applies. Another
difference between the present study and that of Fromhage
et al. (2005) is that we explicitly consider two ways in which
monogynists might enhance their paternity: by outcompeting
rival ejaculates in sperm competition, and by reducing the
probability that a female remates with rival males. In spiders,
the latter goal is achieved through mechanisms such as mate-
guarding (Christenson and Goist, 1979), self-sacrifice to a
cannibalistic female (Andrade, 1996), and mating plugs
formed either of broken copulatory organs (Fromhage and
Schneider, 2006) or a male’s dead body (Foellmer and
Fairbairn, 2003). Another possible means of preventing
female remating would be the transfer of ejaculate compo-
nents that manipulate female behavior, as has been
documented in insects (Chapman et al., 2003). We take into
account that males in web-building spiders are the mate-
searching sex, and are hence typically exposed to an increased
risk of mortality compared to the sedentary females
(Andrade, 2003; Kasumovic et al., 2007; Vollrath and Parker,
1992). Our aim is to address the conditions under which
monogyny is likely to evolve and to be maintained by
selection.
2. The model
We envisage a large population of constant size, in which
a steady influx of newly maturing individuals is offset by a
matching rate of death. For simplicity, we assume that all
characteristics of the population are stationary in time, i.e.,
there is no element of seasonality in our model. There are
two kinds of males: monogynists, who mate with one
female only, and bigynists, who attempt to mate with two
females. Males sequentially search for females, attempting
to mate with every female they encounter. A male may die
either during the search, or after achieving his maximum
number of matings. We define m (0omo1) as the
mortality risk encountered by a male each time he searches
for a female. Implicit in this parameter are all factors that
may affect male travel mortality, such as search efficiency
525
and population density. Each female has a fixed lifespan,
during which she is available for encounters with males,
and at the end of which she lays eggs and dies. Note that
because female lifespan is a determinant of female density,
it too is implicit in the parameter m. We assume that the
probability that a female experiences a mating attempt at
any given time is independent of previous attempts that she
has experienced; in other words, mating attempts are
randomly distributed across females. We further assume
that each female is initially receptive when she becomes
mature, so that mating attempts with her are successful
(i.e., result in mating) with probability 1. Monogynists,
however, reduce the probability of remating in their mate,
such that further mating attempts with her are successful
with a reduced probability y (where 0pyp1).
Many quantities used to describe our model are scaled
by the number of females. It will be convenient to refer to
such relative quantities as if they were absolute quantities.
For example, to express the fact that a fraction F of all
females experience a given mating history, we might say
that there are F such females. Similarly, if there are c
mating attempts per female, we might say that there are c
attempts.
2.1. Monogynist mating attempts
In this section we derive the number of monogynist
mating attempts per female. By definition, monogynists
can mate with one female only. Although in spiders this
may involve two separate pedipalp insertions, we refer to
this as a single mating. Because we have assumed that
bigynist mating attempts do not interfere with other
(monogynist or bigynist) mating attempts, we can dis-
regard them for the time being and focus exclusively on
monogynist attempts. If mating attempts were always
successful, the expected number of mating attempts per
monogynist would equal (1m), the probability of surviv-
ing until finding a female. Defining p (0opo1) as the
proportion of monogynists among all males, and TSR as
the tertiary sex ratio of mature males to females that enter
the population, there would then be TSR p(1m) mono-
gynist mating attempts per female. The number of attempts
will be greater, however, if not all attempts are successful.
Let G be the success probability per attempt, i.e., the
probability that an attempt results in mating. Then, if a
monogynist makes an unsuccessful attempt (with prob-
ability 1G) and if he survives another search (with
probability 1m), he will make another attempt. This
second attempt may be followed by a third one and so on,
until the male is either successful or dies. The resulting
number of monogynist mating attempts is given by
X
1
cm ¼ TSR pð1  mÞðð1  GÞð1  mÞÞi
i¼0
TSR pð1  mÞ
¼ . ð1Þ
1  ð1  GÞð1  mÞ
 ARTICLE IN PRESS
526 L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531
Assuming that these cm attempts are randomly distrib-
uted across females, the fraction of females that experience
exactly g of these attempts is given by a Poisson
distribution with mean cm as poiðcm ; gÞ ¼ ðcgm =g!Þecm . Here
and throughout the following analysis, fractions of females
that experience some given mating history can also be
interpreted as probabilities that a randomly chosen female
will experience this mating history. If a monogynist
attempts to mate with a randomly chosen female, he faces
a probability poi(cm, g1) that this female will experience
exactly g monogynist mating attempts in total. This is
equivalent to the probability that the focal female
experiences g1 additional attempts apart from the focal
male’s. If a monogynist attempts to mate with a female that
experiences g such attempts in total, he has a chance 1/g of
being the first monogynist that attempts to mate with this
female. Given our assumptions, his attempt is then
successful with probability 1. On the other hand, he
has a chance 1  ð1=gÞ of not being first, in which case
he is successful with probability y. His average probability
of mating with this female is therefore ð1=gÞ þ yð1
ð1=gÞÞ ¼ y þ ðð1  yÞ=gÞ. Averaged across females that
experience any number of attempts, monogynist mating
attempts are hence successful with probability
X1  
1y ð1  yÞð1  ecm Þ
G¼ poiðcm ; g  1Þ y þ ¼ þ y.
g¼1
g cm
(2)
G and cm can be found numerically by iterating Eqs. (1)
and (2).
2.2. Bigynist mating attempts
Consider the behavior of bigynists up to their first
mating. By the same reasoning used to obtain Eq. (1), we
infer that there are
TSRð1  pÞð1  mÞ
cb1 ¼
1  ð1  GÞð1  mÞ
bigynist attempts at achieving a first mating. Note that,
because the success of a given mating attempt in our model
depends only on female mating history (but not on which
type of male makes the attempt), attempts of both male
types are successful with the same probability G. It follows
that cb1G bigynists are successful at achieving a first
mating. These bigynists, if they survive another mate
search (with probability 1m), can then go on to attempt
to mate with a second female. By the same reasoning used
to obtain Eq. (1), we infer that there are
cb1 Gð1  mÞ
cb2 ¼
1  ð1  GÞð1  mÞ
bigynist attempts at mating with a second female.
Combining first and second females, there are cb ¼ cb1+cb2
bigynist mating attempts in total.
2.3. Distribution of matings
We can now calculate the fractions of females that mate
with any given number of males of either type. Because
mating attempts are randomly distributed across females,
females that mate with at least one monogynist will
experience a total of h(X0) attempts by bigynists, as well as
g(X1) attempts by monogynists, with independent probabil-
ities poi(cb, h) and poiðcm ; gÞ=ð1  poiðcm ; 0ÞÞ (the latter being
a female’s conditional probability of experiencing g attempts,
given that she experiences at least one attempt). If a female
that mates with at least one monogynist experiences a total of
h attempts by bigynists, as well as g attempts by monogynists,
then the probability that she experiences exactly n(ph)
attempts by bigynists before experiencing the first attempt by
a monogynist is given by
!
Y
n1
hz g
Aðg; h; nÞ ¼ ,
g þ ðh  zÞ g þ ðh  nÞ
z¼0
where the bracketed term represents the probability that the
first n attempts are by bigynists, and the subsequent term
represents the probability that the (n+1)th attempt is by a
monogynist (given the first n attempts were by bigynists). If a
female experiences h bigynist mating attempts, n of which
occur before her first mating with a monogynist, then the
remaining hn attempts must occur at later times. Because
each of these hn attempts has a probability y of being
successful, the probability that exactly h^ of them are
successful is given by a binomial distribution as
 
h  n h^ ^
^ h  n; yÞ ¼
binðh; y ð1  yÞhnh .
h^
Similarly, if a female experiences (g1) additional
monogynist mating attempts after having mated with
one monogynist already, the probability that exactly g^ of
these attempts are successful is given by binðg; ^ g  1; yÞ.
Consider a female that experiences g(X1) attempts by
monogynists, as well as h(X0) attempts by bigynists. Further
suppose that this female mates with n bigynists before mating
with a first monogynist. Such a female will experience j ¼
n þ h^ bigynist matings with probability binðh; ^ h  n; yÞ;
she will also experience i ¼ 1 þ g^ monogynist matings with
independent probability binðg; ^ g  1; yÞ. Thus, the probability
that such a female mates with exactly j bigynists
and i monogynists is given by binðj  n; h  n; yÞbin
ði  1; g  1; yÞ. Now consider all of the 1  poiðcm ; 0Þ
females that mate with at least one monogynist. For each
of these females, the probability Fij of mating with exactly i
monogynists and j bigynists (which requires gXi and hXj) is
given by the sum of the probabilities of all possible sequences
as events that can lead to this outcome:
( "
X 1
poiðcm ; gÞ X 1 Xj
F ij ¼ poiðcb ; hÞ ðAðg; h; nÞ
g¼i
1  poiðcm ; 0Þ h¼j n¼0
#)
 binðj  n; h  n; yÞbinði  1; g  1; yÞÞ . ð3Þ
 ARTICLE IN PRESS
L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531 527

Although we cannot solve this equation exactly, we can do
so approximately by ignoring females that experience
exceptionally high numbers of mating attempts. There are
large regions of parameter space in which females very rarely
experience more than, say, 20 mating attempts by a given
male type. Thus, to compute Fig. 2, we limit our attention to
the 499.9% of females for which g and h in Eq. (3) take
values p20.
of p, we can check for the existence of an equilibrium
frequency where both male types have equal fitness
(Fig. 2(A)). To specify the conditions where a given pure
strategy is evolutionarily stable, we must consider the case
where its alternative strategy is extremely rare. Because
Eq. (3) becomes undefined for p-0, and Eq. (4) becomes
undefined for p-1, we address these important special
cases separately in the Appendix.

2.4. Sperm competition and reproductive success 3. Results and discussion

Consider a female that has mated with i (X1) mono-
gynists as well as with j (X0) bigynists (as happens with
probability ð1  poiðcm ; 0ÞÞF ij ). If sperm competition fol-
lows a raffle principle, where a monogynous ejaculate
contains an x-fold amount of sperm compared to a
bigynous ejaculate, then a proportion xi=ðxi þ jÞ of the
female’s eggs are fertilized by monogynists. We can now
infer wm, the reproductive success per monogynist, by
dividing the collective success of monogynists by their
number, pTSR:
P P1
ð1  poiðcm ; 0ÞÞ 1
i¼1 j¼0 F ij ðxi=ðxi þ jÞÞ
wm ¼ . (4)
pTSR
If a female mates with no monogynist but with at least
one bigynist, as happens with probability poiðcm ; 0Þ
ð1  poiðcb ; 0ÞÞ, then all of her eggs are fertilized by
bigynists. If a female mates with i (X1) monogynists as
well as with j (X0) bigynists, then a proportion j=ðxi þ jÞ of
her eggs are fertilized by bigynists. The reproductive
success per bigynist can be obtained by dividing the
collective success of bigynists by their number, ð1  pÞTSR:
P1 P1
poiðcm ; 0Þð1  poiðcb ; 0ÞÞ þ ð1  poiðcm ; 0ÞÞ i¼1
wb ¼
ð1  pÞTSR
Monogyny is favored by selection if wm4wb; the reverse
is true if wmowb. By calculating wm/wb for different values
Our computations reveal regions of parameter space in
which monogyny and bigyny are evolutionarily stable
against invasion by each other. These regions are mostly
separated by a ‘mixed’ region, where neither pure strategy,
but a mixture of the two, is evolutionarily stable (Fig. 1).
On the other hand, there is also a narrow range of
conditions where either monogyny or bigyny can be
evolutionarily stable (Fig. 1(C)).
In the baseline case, where monogynists transfer twice as
much ejaculate to one female as do bigynists, so the total
amount of ejaculate is equal (i.e., x ¼ 2), a male-biased
TSR is required for monogyny to be maintained at a non-
zero frequency. The required male bias is extreme if
monogynists cannot prevent their mate from remating (i.e.,
y ¼ 1; Fig. 1(A)); if they can, a more moderate bias is
sufficient (Fig. 1(B) and (C)). The importance of the TSR
here lies in the fact that the degree of male bias determines
the degree of competition over paternity success. If this
competition is weak, males have a good chance of
achieving full paternity even without investing in paternity
protection at the cost of losing the opportunity for a
second mating.
j¼0 F ij ðj=ðxi þ jÞÞ
. Deviations from the baseline case (x ¼ 2) are biologi-
cally plausible because copulation duration in spiders is
(5)
often controlled by females, which limits the extent to
which males can empty their pedipalps (e.g., Elgar et al.,
2000; Schneider et al., 2006). This creates the possibility for

Fig. 1. ESS regions in parameter space of TSR and m. For each set of parameter values, there is a ‘mixed’ region (shaded) where neither monogyny nor
bigyny are evolutionarily stable. In each case, monogyny is stable above the mixed region (i.e., toward higher values of TSR), whereas bigyny is stable
below. Each panel (A–C) shows solutions for three sets of parameter values, where the competitiveness of monogynous ejaculates is moderate (x ¼ 2; light
shading), high (x ¼ 3; medium shading), or very high (x ¼ 10; dark shading). From panels (A) through (C), monogynists are increasingly able to reduce
the probability of remating in their mate. (A) y ¼ 1; (B) y ¼ 0.5; and (C) y ¼ 0.3. In (C), each mixed region is joined to the right by a corresponding
narrow area in which the ESS regions of monogyny and bigyny overlap (broken outline; magnified inset), meaning that either strategy can be stable here.
 ARTICLE IN PRESS
528 L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531

selection on male types to be modified by female
preferences—with potentially important consequences: if
females have a strong preference for monogynists (causing
xb2), monogyny can invade even under a balanced TSR
(Fig. 1(B) and (C)).
Because the present model, in contrast to earlier work
(Fromhage et al., 2005), is based on an explicit mechanism
of sperm competition (as characterized by the parameters x
and y), we can use it to calculate the fitness of each strategy
at any frequency. Under conditions where neither pure
strategy was evolutionarily stable, we always found a single
intermediate frequency of monogyny, henceforth called p*,
where both male types had equal fitness. This equilibrium
is stabilized by selection against whatever male type is more
frequent than under p* (Fig. 2(A)). Other things being
equal, p* generally increases with TSR, but decreases with
y (Fig. 2(B)). Less straightforward is the effect of male
travel mortality m in this context: depending on other
parameter values, p* either increases or decreases with m,
often reaching a maximum at some intermediate value
(Fig. 2(C)). The stability of p* is caused by the fact that a
high frequency of monogyny generally reduces the total
number of mating attempts that occur in the population.
This in turn reduces the risk of sperm competition—a
situation that can be best exploited by bigynists. Perhaps
less intuitive is the result that monogyny and bigyny can
also be alternative evolutionarily stable strategies (ESSs)
for given parameter values (Figs. 1(C) and 2(A)), which
means that, depending on initial conditions, either strategy
may become established. This result can be explained by a
disadvantage to bigynists when rare that comes into play at
low values of y (as in Fig. 1(C)), thus allowing monogyny
to thrive at high frequencies when, other things being
equal, monogyny would be disfavored at low frequencies.
This frequency-dependent disadvantage to bigynists arises
as follows: as the frequency of monogyny increases, mating
attempts often fail because the target female has already
mated with a monogynist. Because unsuccessful males can
make further attempts only after incurring a risk of
mortality, this results in fewer expected matings for males
of both types. The latter effect, however, is more
pronounced for bigynists, because bigynists’ second mat-
ings are doubly affected: they require not only that a
second receptive female is found, but also that a first
receptive female was found earlier. Thus, in effect, bigynists
are disproportionally exposed to travel mortality when y is
low and the frequency of monogyny is high. Although the
possibility of alternative ESSs for given parameter values is
an intriguing result, the very narrow range of conditions
that produces it indicates that it probably does not play a
major role in nature.
How do these results relate to empirical findings? Given
the degree of simplification used in our model, we cannot
quantitatively predict the behavior of any particular
species. For example, we have assumed that sperm
competition follows a raffle principle (an assumption
frequently violated in spiders; Uhl, 2002) and that mate
availability is not subject to seasonal change. Nevertheless,
we can identify qualitative trends that help to expose the
logic of this type of mating system.
The main prediction from our model is that monogyny
should be largely restricted to species in which there is a
male-biased TSR. This is supported by a correlation
between monogyny and ‘male accumulation’ on female
webs (a surrogate measure of sex ratio) reported in a recent
comparative study on araneoid spiders (Miller, 2007).
Some of the better-studied species in which monogyny
occurs in combination with a male-biased TSR (or a more
or less indirect surrogate measure thereof) include Latro-
dectus hasselti (Andrade, 1996), Argiope aurantia (Foellmer
and Fairbairn, 2005), Nephila clavipes (Christenson, 1989;
Christenson et al., 1985), and Nephila fenestrata (Fromh-
age et al., 2007a). There is tentative evidence, however, that
the TSR in the largely monogynous Argiope bruennichi may
be close to unity (J.M. Schneider, pers. comm. 2007).
Miller (2007) also found a positive correlation between
extreme sexual size dimorphism and ‘male accumulation,’
suggesting that the sex ratio bias is driven by sexual size
dimorphism. This appears plausible in the light of a
widespread trade-off between adult size and juvenile
survival (Stearns, 1992), whereby the smaller sex is
typically more likely to reach maturity (Clutton-Brock,
1991). In seasonal habitats, sexual size dimorphism in
spiders is often associated with protandry, i.e., males

Fig. 2. Equilibrium frequency of monogyny in cases where neither pure strategy is evolutionarily stable. (A) Examples of how relative fitness of monogyny
changes with p; the equilibrium frequency p* is located where wm/wb ¼ 1. Selection can be negatively (solid line; settings: TSR ¼ 2.75, m ¼ 0.5, x ¼ 2,
y ¼ 0.5) or positively (broken line; TSR ¼ 3, m ¼ 0.86, x ¼ 2, y ¼ 0.3) frequency dependent; (B) p* as a function of TSR under different values of y
(settings: m ¼ 0.5; x ¼ 2); and (C) p* as a function of m under different values of TSR (settings: x ¼ 2, y ¼ 0.5).
 ARTICLE IN PRESS
L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531 529

maturing earlier than females (Morbey and Ydenberg, Acknowledgment

2001), which is likely to make male–male competition
particularly intense early in the season.
Monogyny in association with male-biased sex ratios has
also been reported in taxa as diverse as insects (Monnin
and Peeters, 1998; Moritz and Southwick, 1992), flat
worms (Morand and Muller-Graf, 2000), fish, mollusks,
crustaceans, and representatives of the phyla Cycliophora
and Echiura (Vollrath, 1998), suggesting that similar
selective forces may be at work on a broader taxonomic
scale.
Another prediction from our model is that monogyny
and bigyny can coexist under negative frequency dependent
selection. Indeed, observed behavior in several spider
species is in line with this prediction. Males in the genus
Argiope are often cannibalized by the female after
copulating with one pedipalp only. If they survive,
however, they may or may not perform their second (and
last) copulation with the same female. In laboratory trials
with virgin males and females, the proportion of surviving
males that used both pedipalps to copulate with the same
female was 88% (A. aemula; Sasaki and Iwahashi, 1995),
78% (A. bruennichi; S. Nessler, pers. comm. 2007), 64%
(A. aurantia; Foellmer and Fairbairn, 2004), or 16%
(A. keyserlingi; Herberstein et al., 2005a). Similarly, in
the golden orb spider N. fenestrata, 70% of males used
both rather than one of their pedipalps to copulate with the
same female (Fromhage and Schneider, 2005a). These
observations are not sufficient, however, to establish that
alternative male behavior reflects a mixed evolutionarily
LF was supported by a research fellowship from the
Deutsche Forschungsgemeinschaft.
Appendix. Invasion case
Consider a bigynous mutant in a population where
almost all males are monogynists. The expected number of
mating attempts of this mutant, which we refer to as cb_mut ,
can be derived by the same reasoning used to derive cb in
the main text. Each time the mutant encounters a female,
there is a probability poi(cm, g) that this female experiences
a total of g mating attempts by monogynists. If g ¼ 0, the
mutant mates and receives full paternity. If gX1, the
mutant succeeds in mating with the female with probability
1 if he is the first male that attempts to do so (as happens
with probability 1=ð1 þ gÞ), and with probability y if he is
not first (as happens with probability 1  1=ð1 þ gÞ).
Independent of whether or not the mutant mates first,
the female always mates with the first of the g monogynists
that attempt to mate with her. The success probability for
each of the g1 remaining monogynist mating attempts
that are made with this female drops to y. The probability
that exactly g^ of these g1 attempts are successful
is therefore given by a binomial distribution as
^ g  1; yÞ. Thus, the mutant’s expected reproductive
binðg;
success is given by
"
X1

stable strategy (ESS); an alternative possibility is that male wb_mut ¼ cb_mut poiðcm ; 0Þ þ poiðcm ; gÞ
g¼1
behavior may be conditional on factors yet to be identified !#
(cf. Binmore and Samuelson, 2001). X
g1
 ^ g  1; yÞB
binðg; ,
Our model makes no clear-cut predictions regarding the
^
g¼0
role of male travel mortality for the maintenance of
monogyny. This is because travel mortality has a dual where
effect on the relative success of male types: on one hand, it    
reduces the probability that a bigynist is going to find a 1 1 1 y
B¼ þ 1 .
second mate, thus cutting down the rewards for further ^
1 þ g 1 þ xð1 þ gÞ ^
1 þ g 1 þ xð1 þ gÞ
mate search. On the other hand, it also reduces the
probability that a given male will need to protect his This can be simplified to
paternity, thus undermining the incentive for being "
monogynous. We have treated male travel mortality and X
1
wb_mut ¼ cb_mut poiðcm ; 0Þ þ poiðcm ; gÞ
TSR as independent parameters in our model. It is g¼1
worth noting, however, that these parameters may not !#
X
g1
^ g  1; yÞð1 þ gyÞ
binðg;
necessarily be independent over evolutionary time: Voll-  . ðA:1Þ
rath and Parker (1992) suggested that high male travel ^
g¼0
^
ð1 þ gÞð1 þ x þ xgÞ
mortality selects for a life-history strategy where males
mature at a smaller size, but in greater numbers, than Reproductive success of the predominant male strategy
females. If this argument holds, then high male travel (in this case monogyny) can be calculated using the
mortality might favor the evolution of monogyny indirectly constraint that mean male and female reproductive
via its effect on the TSR. Note that such a dependency success are linked via the sex ratio. In a large population
between mortality and TSR would not affect our conclu- where almost all males are monogynists, mean female
sions regarding the predicted mating system under any reproductive success is equivalent to the probability
given combination of these parameters; it would merely 1poi(cm, 0) that a female mates at least once. Because
cause certain parameter combinations to be more common the contribution of rare bigynists to mean male reproduc-
in nature than others. tive success is negligible, we can equate reproductive
 ARTICLE IN PRESS
530 L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531
success of monogynists with mean male reproductive
success, and calculate it as
1  poiðcm ; 0Þ
wm_maj ¼ . 349–368.
TSR
Now consider the situation where a rare monogynous
mutant attempts to invade a population in which all other
males are bigynists. The mutant can expect (1m) matings,
and has a probability poi(cb, h) of mating with a female that
experiences a total of h mating attempts by bigynists. When
mating with a female that experiences h bigynist mating
attempts in total, the mutant faces a probability 1=ð1 þ hÞ
that n of these h bigynist attempts have already taken place
successfully. After the female has mated with the mono-
gynous mutant, the success probability for each of the hn
further bigynist mating attempts that are made with her
drops to y. The probability that exactly h^ of these hn
further attempts are successful is therefore given by a
binomial distribution as binðh; ^ h  n; yÞ. So if the mutant
is the (n+1)th male to attempt to mate with a given
female, the female receives not only the mutant’s ejaculate
(of value x), but also n ejaculates of earlier suitors
and, with probability binðh; ^ h  n; yÞ, h^ ejaculates of later
suitors. The mutant’s expected reproductive success is thus
given by
X
1 X
h 270, 183–185.
1 Foellmer, M.W., Fairbairn, D.J., 2004. Males under attack: sexual
wm_mut ¼ ð1  mÞ poiðcb ; hÞ
h¼0 n¼0
1 þ h
X
hn
x Foellmer, M.W., Fairbairn, D.J., 2005. Selection on male size, leg length
 ^ h  n; yÞ
binðh; . ðA:2Þ
^
h¼0
x þ n þ h^
Mean female reproductive success in this situation is
equivalent to the probability 1poi(cb, 0) that a female
mates at least once. Using the link between male and
female mean reproductive success as above, the reproduc-
tive success of bigynists is given by
1  poiðcb ; 0Þ
wb_maj ¼ .
TSR
Monogyny is an ESS if wm_maj 4wb_mut ; bigyny is an ESS
if wb_maj 4wm_mut . Although we cannot solve these condi-
tions exactly, we can do so approximately by ignoring
females that experience exceptionally high numbers of
mating attempts. To compute Fig. 1, we let g and h in
Eqs. (A.1) and (A.2) take values p50. Under the parameter
values shown, this takes into account 499.9% of all
females.
References
Andrade, M.C.B., 1996. Sexual selection for male sacrifice in the
Australian redback spider. Science 271, 70–72.
Andrade, M.C.B., 2003. Risky mate search and male self-sacrifice in
redback spiders. Behav. Ecol. 14, 531–538.
Andrade, M.C.B., Banta, E.M., 2002. Value of male remating and
functional sterility in redback spiders. Anim. Behav. 63, 857–870.
Andrade, M.C.B., Kasumovic, M.M., 2005. Terminal investment strate-
gies and male mate choice: extreme tests of Bateman. Integr. Comp.
Biol. 45, 838–847.
Bateman, A.J., 1948. Intra-sexual selection in Drosophila. Heredity 2,
Binmore, K., Samuelson, L., 2001. Can mixed strategies be stable in
asymmetric games? J. Theor. Biol. 210, 1–14.
Breene, R.G., Sweet, M.H., 1985. Evidence of insemination of multiple
females by the male black widow spider, Latrodectus curacaviensis
(Muller). J. Arachnol. 13, 331–335.
Chapman, T., Bangham, J., Vinti, G., Seifried, B., Lung, O., Wolfner,
M.F., Smith, H.K., Partridge, L., 2003. The sex peptide of Drosophila
melanogaster: female post-mating responses analyzed by using RNA
interference. Proc. Natl Acad. Sci. USA 100, 9923–9928.
Christenson, T.E., 1989. Sperm depletion in the orb-weaving spider
Nephila clavipes (Araneae, Araneidae). J. Arachnol. 17, 115–118.
Christenson, T.E., Goist, K.C.J., 1979. Costs and benefits of male–male
competition in the orb weaving spider, Nephila clavipes. Behav. Ecol.
Sociobiol. 5, 87–92.
Christenson, T.E., Brown, S.G., Wenzl, P.A., Hill, E.M., Goist, K.C.,
1985. Mating behaviour of the golden-orb-weaving spider, Nephila
clavipes: I. Female receptivity and male courtship. J. Comp. Psychol.
99, 160–166.
Clutton-Brock, T.H., 1991. The Evolution of Parental Care. In:
Monographs in Behaviour and Ecology. Princeton University Press,
Princeton.
Elgar, M.A., Schneider, J.M., Herberstein, M.E., 2000. Female control of
paternity in the sexually cannibalistic spider Argiope keyserlingi. Proc.
R. Soc. London B 267, 2439–2443.
Foellmer, M.W., Fairbairn, D.J., 2003. Spontaneous male death during
copulation in an orb-weaving spider. Proc. R. Soc. London B Suppl.
cannibalism and its consequences for male morphology and behaviour
in an orb-weaving spider. Evol. Ecol. Res. 6, 163–181.
and condition during mate search in a sexually highly dimorphic orb-
weaving spider. Oecologia 142, 653–662.
Fromhage, L., Schneider, J.M., 2005a. Safer sex with feeding females:
sexual conflict in a cannibalistic spider. Behav. Ecol. 16, 377–382.
Fromhage, L., Schneider, J.M., 2005b. Virgin doves and mated hawks:
contest behaviour in a spider. Anim. Behav. 70, 1099–1104.
Fromhage, L., Schneider, J.M., 2006. Emasculation to plug up females:
the significance of pedipalp damage in Nephila fenestrata. Behav. Ecol.
17, 353–357.
Fromhage, L., Elgar, M.A., Schneider, J.M., 2005. Faithful without care:
the evolution of monogyny. Evolution 59, 1400–1405.
Fromhage, L., Jacobs, K., Schneider, J.M., 2007a. Monogynous mating
behaviour and its ecological basis in the golden orb spider Nephila
fenestrata. Ethology 113, 813–820.
Fromhage, L., McNamara, J.M., Houston, A.I., 2007b. Stability and
value of male care for offspring—is it worth only half the trouble? Biol.
Lett. 3, 234–236.
Herberstein, M.E., Barry, K.L., Turoczy, M.A., Wills, E., Youssef, C.,
Elgar, M.A., 2005a. Post-copulation mate guarding in the sexually
cannibalistic St Andrew’s Cross spider (Araneae Araneidae). Ethol.
Ecol. Evol. 17, 17–26.
Herberstein, M.E., Gaskett, A.C., Schneider, J.M., Vella, N.G.F., Elgar,
M.A., 2005b. Limits to male copulation frequency: sexual cannibalism
and sterility in St Andrew’s cross spiders (Araneae, Araneidae).
Ethology 111, 1050–1061.
Kasumovic, M.M., Bruce, M.J., Herberstein, M.E., Andrade, M.C.B.,
2007. Risky mate search and mate preference in the golden orb-web
spider (Nephila plumipes). Behav. Ecol. 18, 189–195.
Kvarnemo, C., 2006. Evolution and maintenance of male care: is increased
paternity a neglected benefit of care? Behav. Ecol. 17, 144–148.
Miller, J.A., 2007. Repeated evolution of male sacrifice behavior in spiders
correlated with genital mutilation. Evolution 61, 1301–1315.
 ARTICLE IN PRESS
L. Fromhage et al. / Journal of Theoretical Biology 250 (2008) 524–531
Monnin, T., Peeters, C., 1998. Monogyny and regulation of worker
mating in the queenless ant Dinoponera quadriceps. Anim. Behav. 55,
299–306.
Morand, S., Muller-Graf, C.D.M., 2000. Muscles or testes?
Comparative evidence for sexual competition among dioecious
blood parasites (Schistosomatidae) of vertebrates. Parasitology 120,
45–56.
Morbey, Y.E., Ydenberg, R.C., 2001. Protandrous arrival timing to
breeding areas: a review. Ecol. Lett. 4, 663–673.
Moritz, R.F.A., Southwick, E.E., 1992. Bees as Superorganisms: An
Evolutionary Reality. Springer, Heidelberg.
Sasaki, T., Iwahashi, O., 1995. Sexual cannibalism in an orb-weaving
spider Argiope aemula. Anim. Behav. 49, 1119–1121.
Schneider, J.M., Fromhage, L., Uhl, G., 2005. Extremely short copula-
tions do not affect hatching success in Argiope bruennichi (Araneae,
Araneidae). J. Arachnol. 33, 663–669.
531
Schneider, J.M., Gilberg, S., Fromhage, L., Uhl, G., 2006. Sexual conflict
over copulation duration in a cannibalistic spider. Anim. Behav. 71,
781–788.
Stearns, S.C., 1992. The Evolution of Life Histories. Oxford University
Press, Oxford.
Trivers, R.L., 1972. Parental investment and sexual selection. In:
Campbell, B. (Ed.), Sexual Selection and the Descent of Man. Aldine,
Chicago, pp. 1871–1971.
Uhl, G., 2002. Female genital morphology and sperm priority patterns in
spiders (Araneae). In: Toft, S., Scharff, N. (Eds.), European
Arachnology 2000. Aarhus University Press, Aarhus, pp. 145–156.
Vollrath, F., 1998. Dwarf males. Trends Ecol. Evol. 13, 159–163.
Vollrath, F., Parker, G.A., 1992. Sexual dimorphism and distorted sex
ratios in spiders. Nature 360, 156–159.
Wickler, W., Seibt, U., 1981. Monogamy in crustacea and man.
Z. Tierpsychol. 57, 215–234.