| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Behavioral Ecology Vol. 11 No. 4: 421-428
© 2000 International Society for Behavioral Ecology
Sex-specific dispersal in spatially varying environments leads to habitat-dependent evolutionary stable offspring sex ratios
Department of Biology, Division of Zoology, University of Oslo, POB 1050 Blindern, N-0316 Oslo, Norway, and C.R.B.P.O., Muséum National d'Histoire Naturelle, 55 rue Buffon, F-75005 Paris, France
Address correspondence to R. Julliard at C.R.B.P.O., Muséum National d'Histoire Naturelle, 55 rue Buffon, F-75005 Paris, France. Email : julliard{at}mnhn.fr .
Received 14 December 1998; revised 8 November 1999; accepted 30 November 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONThe modelRESULTSDISCUSSIONREFERENCES |
|---|
When the environment varies spatially, so that some habitats are more favorable to reproduction than others, an individual should attempt to increase the number of offspring establishing in high-quality habitats. Hence, if male and female dispersal behavior differ, it may be adaptive to produce more offspring of the more dispersing sex in low-quality habitats, since these offspring are likely to disperse to another patch, and more offspring of the most philopatric sex in high-quality habitats, since these offspring are likely to remain in that patch. Such a strategy is shown to be evolutionarily stable provided that male and female dispersal rates are different and that reproductive success varies between habitats (lack of ideal free distribution). Highly biased sex ratios are predicted (1) in rare habitats, (2) in poor habitats, (3) when difference between habitat quality is large, (4) when at least one sex disperses at a rate close to random with respect to habitat availability, (5) when both sexes disperse at a high rate, (6) when individuals are unable to select their reproducing habitat, and, presumably, (7) with moderate temporal variation of habitat quality. The model appears to be a good candidate to explain the pattern of sex ratio variation in a variety of species : phytophagous arthropods, species with environmental sex determination, and territorial passerines.
Key words: environmental sex determination, ideal free distribution model, phytophagous arthropods, reaction norm, territorial passerines..
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONThe modelRESULTSDISCUSSIONREFERENCES |
|---|
In the words of Emlen (1997
291), "No one can
fault evolutionary theorists for a lack of creativity when it comes to
devising hypotheses of adaptive sex ratio variation." Many of these
hypotheses have received strong empirical support (review in
Ellegren and Sheldon, 1996 ;
Godfray and Werren, 1996). The
most successful ones in terms of empirical support are those focusing on
interactions between kin. For example, when offspring of one sex are likely to
compete with each other (or with their parents) more than offspring of the
other sex, it may be advantageous to produce more offspring of the other sex
less likely to compete locally [local resource competition, LRC
(Clark, 1978 ;
Taylor, 1994), which is a
generalization of local mate competition, LMC
(Hamilton, 1967)]. Conversely,
when offspring of one sex are likely to cooperate with each other (or with
their parents) more than offspring of the other sex, it may be advantageous to
produce more offspring of the cooperative sex [local resource enhancement, LRE
(Emlen et al., 1986 ;
Johnson, 1988 ;
Schwartz, 1988)]. Another
successful hypothesis is that parents should take advantage of the high
phenotypic or genetic quality of one of them to produce more of the sex that
would most benefit from this high quality
(Trivers and Willard, 1973).
For example, females in good condition are expected to produce more of the sex
likely to gain a high social position from inherited high phenotypic quality
(Leimar, 1996). Similarly,
females mated to a high-quality male are supposed to increase the number of
male offspring they produce, which may inherit the high quality of their
father (Burley, 1981 ;
Ellegren et al., 1996).
Parents may also take into consideration the time in the season when offspring
are produced : when territory acquisition depends on dominance in one sex more
than in the other, and when dominance is related to age, parents are expected
to produce more of that sex early in the season than later in the season
(Daan et al., 1996 ;
Lambin, 1994).
All these hypotheses consider a simple and small-scale environment, often
dealing only with what happens within the family group or in the birth
territory. However, natural environments are often spatially variable, with
relatively high-quality habitats allowing higher reproductive success than
relatively low-quality habitats (e.g.,
Dhondt et al., 1990). The main
explanation for this pattern is the inability of individuals to distribute
themselves in an ideal manner (see
Fretwell and Lucas, 1970), so
that density dependence does not ensure equal fitness in all habitat patches
(Holt, 1996). Furthermore, in
most species, dispersal rates vary between sexes, for reasons that are
generally independent of environment variability (e.g., sex-specific
territorial behavior, Greenwood,
1980 ; inbreeding avoidance,
Pusey, 1987). Hence, male and
female offspring may disperse from one habitat to another with different
probabilities. What is the evolutionarily stable sex ratio in such spatially
variable environments when dispersal probabilities vary between sexes ?
Nordborg (1991 ; 1289)
provided a rather abrupt answer : "Fisher's result does hold,"
that is, the sex ratio of offspring should be balanced among offspring as
established by Fisher (1930).
However, in his model, Nordborg searched for an unconditional evolutionarily
stable sex ratio irrespective of environmental variation ; in other words, he
did not allow any plasticity of sex ratio, as do most other hypotheses where
sex ratio is a flexible parameter.
By appropriately manipulating the sex ratio, it may be possible to increase the number of offspring establishing in highquality habitats : producing more offspring of the most dispersing sex in low-quality habitats would enhance dispersal into higher quality habitats, whereas producing more offspring of the most philopatric sex in high-quality habitats would increase the number of offspring likely to remain there. Here, I present a model with sex-dependent dispersal in heterogeneous environment (spatial variation of reproductive success not compensated by ideal free distribution) that predicts a unique evolutionarily stable habitat-dependent sex ratio reaction norm : a strongly biased sex ratio is expected to occur in rare habitats (either good or poor) when probabilities of dispersing between habitats differs between sexes. Hence, the model makes the simple prediction that parents should adjust the sex ratio of their offspring to the local habitat quality. The model is tested with published data of sex ratio variation in various animal taxa.
|
The model |
|---|
|
TOPABSTRACTINTRODUCTIONThe modelRESULTSDISCUSSIONREFERENCES |
|---|
Environment heterogeneity is defined as spatial variability of reproductive success. The environment consists of good habitat and poor habitat patches. Furthermore, a given habitat may be relatively rare or relatively common. The proportion of good habitats is defined as G (and 1 - G for poor habitat ; 0 < G < 1) ; reproducing individuals are distributed accordingly between good and poor habitats. Habitat quality determines the reproductive success, here defined as the number of offspring reaching the age of dispersal per reproducing individual. Reproductive success in the good habitats is assumed to be F times as high as the reproductive success in the poor habitats (F > 1). Note that the poor habitats are not necessarily demographic sinks (i.e., unsustainable populations without immigration ; Pulliam, 1988
). At each time step, new offspring may disperse and may thus
change habitat with a certain probability. This probability depends on two
factors. The first factor is the relative availability of the two habitats. In
the extreme case where individuals distribute themselves at random between the
two habitats, a proportion, G, would disperse into good habitats,
while a proportion, 1 - G, would disperse in poor habitats,
independently of birth habitat. Second, individuals may show some philopatry
toward their birth habitat ; the probability of changing habitat may therefore
be lower than predicted from the availability of the two types of habitat. Let
us define a dispersal coefficient independent of birth habitat type D
(D 
[0,1]), so that dispersal rate from good to poor habitat is
D(1 - G) and dispersal rate from poor to good habitat is
D(G). Consequently, a proportion, (1 - D) (1 -
G), of individuals born in a good habitat attempt to reproduce in
good habitats ; 1 - DG individuals born in a poor habitat attempt to
reproduce in poor habitats. If D = 0, individuals are unable to
change habitat. A high dispersal coefficient reflects a high probability of
changing habitat taking into account the relative availability of the two
habitats. This can be understood as large dispersal distance relative to the
mean distance between habitats. This may result either from long-distance
dispersal or from short distance between habitats (i.e., good and poor habitat
constitute a fine-grained mosaic). Difference in dispersal tendency between
sexes is taken into account with different dispersal coefficients :
Dmal and Dfem for male and female
offspring, respectively. The model is thus defined by only four parameters :
G, F, Dmal, and Dfem.
The aim of the model is to find evolutionarily stable habitat-dependent sex
ratios when these four parameters vary. A strategy is defined by the two sex
ratios an individual produces in good and poor habitat
(SRG and SRP, respectively ; sex ratio
is defined as the proportion of females among offspring). The evolutionarity
stable strategy (ESS) should simultaneously maximize the fitness (W)
of an individual reproducing in good habitat
[WG(SRG)] and in poor habitat
[WP(SRP)] ; that is (following the method
recommended by Taylor and Frank, 1996) :
 | (1) |
W is the sum of reproductive values of the different kinds of
offspring produced by an individual (males and females in good and poor
habitats). Within habitat, the reproductive value of individuals of a given
sex is inversely proportional to the proportion of individuals of that sex in
that habitat (i.e., secondary sex ratio, SSR). In addition, because
both males and females produce F times more individuals in good than
in poor habitat, the reproductive value of individuals settling in good
habitats is F times higher than in poor habitats. Hence,
 | (2) |
 | (3) |
SSRG and SSRP are secondary sex
ratios (i.e., the proportion of females among individuals after dispersal) in
good and poor habitat, respectively. Thus,
 | (4) |
 | (5) |
 | (6) |
 | (7) |
 | (8) |
, for a
justification of the method).
Except for particular values of G, F, Dmal, and
Dfem, there is no simple solution for Equation 1. The
calculation of the numeric solution was done using Mathematica 3.0
(Wolfram, 1997), and the
relationship between ESS sex ratio and the four parameters are presented
graphically.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONThe modelRESULTSDISCUSSIONREFERENCES |
|---|
In the following discussion, females are assumed to disperse less than or as much as males (i.e., Dfem
Dmal). The results equivalent if males are the less
dispersing sex, except that ESS sex ratios are biased in the opposite way.
ESS sex ratio for limit values of the parameters
Multiple ESSs were found when both sexes dispersed at random irrespective
of their birth habitat (i.e., Dfem =
Dmal = 1). All strategies with unbiased sex ratio at the
scale of the population, given by
 | (9) |
Effect of habitat heterogeneity
Habitat heterogeneity is measured both by the commonness of both habitats
(G and 1 - G) and the relative reproductive success
(F) in good habitats. Absolute bias of the ESS sex ratio decreases
with the commonness of the habitat (Figure
1) and is usually higher in the rarest habitat (either good or
poor). The effect of commonness is not symmetric for good and poor habitats :
sex ratio in the rarest habitat is more biased if the rare habitat is the poor
one than if it is the good one (Figures
1 and
2). Furthermore, biases usually
increase with increasing difference of reproductive success between good and
poor habitats (Figure 1).
|
|
Effect of sex-specific dispersal on ESS sex ratio
The bias of the ESS sex ratio is strongly dependent on the dispersal rate
of the most dispersing sex (Figure
2) and becomes important only if dispersal of one sex with respect
to habitat is close to random. Maximum biases are found when female dispersal
rate is high but lower than male dispersal rate
(Figure 2).
Effect of habitat selection
There is no habitat selection included in the model : individuals are
distributed according to habitat availability, not according to habitat
quality. The distribution of individuals is thus not ideal free
(Fretwell and Lucas, 1970). If
it were, relatively more individuals would go into good habitats until
density-dependent processes equalized reproductive success across habitats. In
such a case, there would be no reason to have a habitat-dependent biased sex
ratio. In the model, exchanges between habitats are unbalanced in the opposite
way : because relatively more offspring are produced in good habitats, more
individuals disperse from good to poor than from poor to good habitats (after
controlling for habitat availability). Theory predicts that when habitat
quality varies spatially, ESS dispersal should lead to balanced exchanges of
dispersing individuals between subpopulations
(McPeek and Holt, 1992). If
such a feature is included in the model (i.e., dispersal rate from poor to
good habitats is higher than dispersal rate from good to poor habitats), ESS
sex ratios inevitably tend to 0.5 in both habitats. Hence, biasing sex ratio
can be interpreted as a way to decrease the imbalance of exchanges between
habitats. More generally, biased sex ratios are obtained in situations when
individuals are unable to distribute themselves according to the ideal free
distribution or to realize balanced exchanges. Possibility of habitat
selection thus counterselects biased sex ratio, and, consequently, imperfect
habitat selection will favor the evolution of biased sex ratios.
Temporal variability
No temporal variability was included in the model. Intuitively, biasing sex
ratios is only advantageous if habitat quality is sufficiently stable so that
dispersing individuals leave habitats that will remain poor and philopatric
individuals stay in habitats that will remain good. Hence, the temporal
predictability of habitat quality is a key feature of the model. However,
temporal variability may be modeled as a fraction of good habitat becoming
poor habitat and, reciprocally, at each time step. In the algebra of the
model, this would be equivalent to a global increase of dispersal rate of both
males and females. Moderate temporal variability is thus likely to increase
the bias of the ESS sex ratio (see Figure
2).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONThe modelRESULTSDISCUSSIONREFERENCES |
|---|
Synopsis of the predictions and comparison with other models
The model supports predictions of previous models of sex ratio evolution with similar assumptions : in a spatially varying environment, with the same dispersal behavior between sexes, the ESS sex ratio is not biased (Nordborg, 1991
). The model
makes a number of other predictions : when reproductive success varies
spatially, and when offspring of both sexes disperse, but one sex disperses
less than the other, there is an ESS sex ratio reaction norm : the sex ratio
should be biased toward the less dispersing sex in good habitats and toward
the most dispersing sex in poor habitats. Several nonexclusive features tend
to increase the bias of the ESS sex ratio : biases (departure from 0.5) tend
to be large (1) in rare habitats, (2) in poor habitats, (3) when difference
between habitat quality is large, (4) when at least one sex disperses at a
rate close to random with respect to habitat availability, (5) when both sexes
disperse at a high rate (but dispersal probability has to be different between
sexes), (6) when individuals are unable to select their reproducing habitat,
and, presumably, (7) with moderate temporal variation of habitat quality.
The present model can be seen as a special case of Triverse and Willard's
(1973) general hypothesis :
parents should invest more in the sex that benefits the most from the local
environment. However, most studies referring to the Trivers-Willard hypothesis
have focused on how the local environment (often limited to the mother's body
condition) affects the phenotype of male and female offspring differently
(review in Leimar, 1996).
Here, it is sex-biased dispersal that drives different habitat-dependent
fitness between male and female offspring. Habitat quality does not directly
affect the phenotype of offspring. Hence, this model identifies some
conditions leading to biased habitat-dependent ESS sex ratio, which have been
previously overlooked. It is clear, however, that there are other evolutionary
causes leading to habitat-dependent ESS sex ratio. Identifying whether this
model or the classical interpretation of the Trivers-Willard hypothesis is the
better evolutionary explanation for an observed habitat-dependent sex ratio
will thus require a detailed knowledge of the life history of the species
under study.
Sex-biased dispersal in a heterogeneous environment leads to a secondary sex ratio different from 50:50. Even in the absence of a biased primary sex ratio, relatively more of the most dispersing sex are found in poor habitats, whereas more of the most philopatric sex are found in good habitats. Interestingly, at the ESS, primary sex ratios are thus biased in the same way as secondary sex ratios, and, as a consequence, secondary sex ratios are more biased than they would be in the absence of bias in the primary sex ratio. Hence, the model makes the paradoxical prediction that primary sex ratio should be biased in the same way as the secondary sex ratio.
Although the formulation is already complicated, the model described a
simplistic environment. In particular, kin interactions were ignored, yet they
are generally recognized to be of central importance for the evolution of sex
ratio (see Introduction). Although introducing kin interaction in the model
would be technically difficult, it would be interesting because kin
competition or cost of inbreeding should affect both the ESS dispersal rate
(Hamilton and May, 1977 ;
Pusey, 1987) and the ESS sex
ratio (Bulmer, 1986 ;
Taylor, 1994). Similarly,
temporal variability is likely to increase the bias of the ESS sex ratio (see
above), but the combination of temporal and spatial variability would then
select for non-zero ESS dispersal rate (e.g.,
McPeek and Holt, 1992),
whereas zero-dispersal rate is the ESS with only spatial heterogeneity. How
male and female dispersal rates are related to the ESS dispersal may then
affect the ESS sex ratio independently of spatial variability (see an attempt
to take this into account in Reinhold,
1996). It would be interesting to see the interplay of these
selection pressures (kin interaction, temporal variability) with spatial
variability in shaping the evolution of sex ratio. In particular, coevolution
between sex ratio and dispersal rate may be evident. This is, however, beyond
the scope of this paper.
Illustrations from empirical case studies
The model predicts that in good environments parents should produce more of
the most philopatric sex, while in poor environments parents should produce
more of the most dispersing sex. In the following sections, I review some
empirical examples documenting habitat-dependent offspring sex ratios and
evaluate whether the model offers a plausible explantation of the observed
patterns.
Phytophagous arthropods
Most sex ratio variation in arthropods has been attributed (evolutionarily
speaking) to local mate competition
(Hamilton, 1967 ; review in
Wrensch and Ebbert, 1993).
This theory predicts that when males are likely to mate mostly with sisters or
other closely related females, the ESS sex ratio should be female biased.
However, with a small amount of outbreeding, ESS sex ratio shifts toward 0.5
(Hamilton 1967). The model has
received extraordinary qualitative and quantitative support, mainly from
parasitoid hymenopterans (Heimpel,
1997). There are, however, several cases of offspring sex ratio
variation, particularly in phytophagous arthropods, that do not fit with the
LMC hypothesis and which show consistent variation of sex ratio with some
environmental measurements.
In the sawfly (phytophagous hymenopterans of the families Diprionidae and
Tenthredinidae), consistent sex ration variation of the progeny are found with
respect to host plant quality (Craig et
al., 1992 ; Mopper and
Whitham, 1992) : the offspring sex ratio is biased toward females
on high-quality plants and either male-biased or unbiased on lower quality
plants.
Ruohomäki et
al. (1993 : 420) have shown that sawfly sex ratio varied
"among sites but not among host trees within sites," but they were
unable to find the environmental determinant of sex ratio. They concluded that
LMC could not explain biased sawfly sex ratio
(Walter et al., 1994). Craig
et al. (1992) demonstrated that
females raised on high-quality plants gain relatively more fitness advantages
than males. This gives strong support to the classical Trivers-Willard
(1973) hypothesis (see above).
This hypothesis was, however, rejected by
Ruohomäki and coworkers
(1993 : 420 ; see also
Walter et al., 1994), who
concluded that they were "unable to explain sex ratios of sawflies with
existing general models." Both male and female sawflies are winged and
show different dispersal behavior (Stein
et al., 1994), but it is difficult to tell from the available
published data which sex is the most likely to move between habitats. Hence,
the model cannot currently be tested due to insufficient knowledge of the
sawfly life history. However, sawflies are good candidates for such a
test.
The relationship between plant quality and offspring sex ratio in spider
mites (Acari : Tetranychidae) is the opposite of the one found in sawflies :
females raised on low-quality plants tend to produce relatively more female
offspring than females raised on better quality plants
(Stiefel and Margolies, 1992 ;
Wrensch and Young, 1978 ;
Young et al., 1986 ; these
three studies were based on three different species of spider mites). In
spider mites, females are the dispersing sex (e.g.,
Li and Margolies, 1994), and
it has been explicitly argued that "a higher percentage of females in
response to deteriorating conditions would result in more potential
dispersers" (Stiefel and Margolies,
1992 : 162).
Although density is an ambiguous determinant of sex ratio in spider mites
(see references above), crowding appears to affect in the same way the sex
ratio of two species of otherwise dramatically different insects : both circus
mealybug (Planococcus citri) and gypsy moth (Lymantria
dispar) females raised under high-density conditions produce relatively
more male offspring (mealybug : Varndell
and Godfray, 1996 ; gypsy moth :
Myers et al., 1998). In these
two species, males are winged and females are wingless, so that only males are
able to disperse. It is thus tempting to put forward the same kind of
hypothesis as for spider mites : more offspring of the dispersing sex are
produced to escape future poor environmental conditions. These species thus
meet the basic assumptions of the model : sex-biased dispersal, spatially
varying environment, and ability to adjust the sex ratio (yet to be
demonstrated for the gypsy moth). However, the model does not predict large
biases in sex ratios when one sex is unable to disperse, as is the case in
these examples. However, this prediction may no longer be true when habitat
quality varies temporally (see above), particularly when the variation in
predictable, as seems to be the case in all these species. Hence, the model
may provide a good theoretical framework for future studies on the evolution
of the sex ratio of these species.
Environmental sex determination
In several species of fish and reptiles, as well as in various
invertebrates, offspring gender is, at least partly, determined by some
environmental parameters (e.g., temperature, humidity ; see
Bull, 1983). Charnov and Bull
(1977) listed a set of four
conditions under which environmental sex determination (ESD) may evolve. One
of these conditions looks like the classical formulation of the
Trivers-Willard hypothesis : environmental heterogeneity should affect male
and female offspring fitness differently. Indeed, this condition has been
demonstrated in several species with ESD (review in
Janzen, 1996). However, ESD
remains an evolutionary enigma in the majority of cases. As pointed out
recently by Reinhold (1998),
sex-biased dispersal and spatial heterogeneity offer the condition for the
evolution of ESD. If habitat quality varies spatially and determines the local
reproductive success, and if male and female offspring disperse between
habitats with different probabilities, then ESD is favored if it allows the
production of more of the dispersing sex in poor habitats and/or more of the
philopatric sex in good habitats
(Reinhold, 1998).
One promising group of species for testing the model, as argued by Reinhold
(1998), is sea turtles, in
which females show natal site philopatry, while males mate in the open sea
relatively far from nesting areas. The homing ability of female sea turtles is
indeed excellent ; sea turtles are able to find their natal island after
traveling several thousand kilometers in the sea (e.g.,
Lohmann and Lohmann, 1996).
However, a recent study has shown that males are also highly philopatric to
courtship areas specific to a breeding area
(FitzSimmons et al., 1997).
When both sexes have low between-habitat dispersal rates, the model predicts
only slightly biased ESS sex ratios (Figure
2). This seems to contradict the highly biased sex ratio usually
reported in ESD species, including sea turtles
(Ewert et al., 1994). However,
the geographical scale relevant to the model may be at a much finer grained
scale than the between nesting areas implicitly assumed by Reinhold
(1998). Indeed, within the
main reproductive areas, breeding sea turtles even show a greater site
fidelity. For example, within the 35 km of homogenous-looking beach used for
nesting by the Tortuguero population of the green turtle Chelonia
mydas, individual females appear to breed within the same 1-2 km year
after year (Carr and Carr,
1972). If consistent environmental heterogeneity linked in one way
or another to temperature variation exists at the scale of a few kilometers
within the beach, then female sea turtles are likely to be only slightly
philopatric to their natal site environment. If we further assume that there
is a single courtship area for the Tortuguero population, then males would
mate at random with respect to the breeding habitat quality of females. In the
model specification, this corresponds to Dmal = 1
and Dfem slightly less than 1. These values
maximize the bias of ESS sex ratio, which may then reach 0% or 100% of females
in poor and good habitats, respectively
(Figure 2).
The excellent homing ability of female sea turtles may constrain females to breed on relatively fixed sites and may thus prevent habitat selection on their nesting site. ESD may, however, reduce this potential cost by allowing the production of more of the dispersing sex in poor-quality habitats and more of the philopatric sex in good-quality habitats. Although there are a lot of unanswered questions that remain to be properly documented before the model can be validated, the assumptions of the model do not appear unrealistic with respect to sea turtle's life history. This example further points out the importance of choosing the appropriate geographical scale for testing the model and how biased sex ratio may compensate the cost of reduced habitat selection.
Territorial passerines
An extreme case of sex ratio variation has been recently documented in the
Seychelles warbler Acrocephalus sechellensis
(Komdeur, 1996a ;
Komdeur et al., 1997) : birds
breeding in poor-quality territories (71% of territories), where the
reproductive success is low (0.2 offspring reaching 1 year of age per pair per
year ; Komdeur, 1992), produce
an excess of males (77% ; males are the most dispersing sex in this species),
while birds breeding in high-quality territories (12% of territories) where
reproductive success is high (1.2 offspring per pair per year) produce an
excess of females (65%). Birds changing territory quality change their sex
ratio accordingly, demonstrating that there is a habitat-dependent sex ratio
reaction norm (Komdeur et al.,
1997). Little information on dispersal in the Seychelles warbler
has been published so far : males disperse more than females, while females
often stay and eventually breed on their birth territory (and thus in their
birth habitat), although not all females do so
(Komdeur, 1992). There is also
a strong competitive advantage for locally born individuals
(Komdeur, 1992). Habitat
quality and proportion of good and poor habitats have been reported as
constant. Hence, the environment and life history of the Seychelles warbler
meet the assumptions of the model. Most interestingly, strongly biased
habitat-dependent sex ratio variation is observed, as predicted.
Komdeur et al. (1997)
argued that the bias in sex ratio resulted from the helping behavior of female
offspring : in good-quality territories female offspring may become helpers
and increase their parent's reproductive success ; hence, parents should
produce an excess of females (LRE hypothesis,
Komdeur et al., 1997). In
poor-quality territories, the presence of helpers is claimed to decrease
parental reproductive success (Komdeur,
1996a, Komdeur et al.,
1997), although it was once found to increase parental
reproductive success by 60% (Komdeur,
1994). Nevertheless, the first affirmation is put forward as an
explanation for the excess of males produced in low-quality territory (LRC,
Komdeur et al., 1997). The
present model shows that biased sex ratios in the Seychelles warbler may be
due to the spatial variation in territory quality and does not necessarily
require competition between kin or helping (and consequently does not rely on
the controversial effect of helping in poor-quality territories).
In the Seychelles warbler (and in many other species with helpers ; e.g.,
Emlen, 1994), helping behavior
is clearly beneficial to the helper : a helper may gain breeding experience,
it may increase its inclusive fitness, and, eventually, it may inherit its
parents' territory (Komdeur,
1996b ; see also Komdeur,
1992). The benefit for the parents, although often significant in
terms of number of offspring produced
(Komdeur, 1994), may be
reduced by the many sources of competition between helpers and their parents :
competition for food, competition for breeding (e.g., helpers may lay an egg
in their parents' nest ; Komdeur,
1994). However, the model presented here highlights a potential
indirect benefit of helping for the parents : the insurance that good-quality
territories will remain in the family. In such a case, it is then philopatry,
and not helping per se, that is beneficial. It is interesting that models of
sex ratio evolution in the presence of helping behavior (e.g.,
Emlen et al., 1986) specify
that helpers only help their own parents. Otherwise, biased sex ratio may not
be evolutionarily stable : a mutant parent may invade the population by
producing offspring of the presumably rare nonhelping sex, yet still benefit
from helping by foreign offspring of the other presumably abundant sex. Hence,
biased sex ratio may not be a secondary adaptation in species with help at the
nest as suggested by Emlen et al.
(1986), but both may result
from spatial heterogeneity in reproductive success (see
Koening et al., 1992, for a
model of evolution of helping in relation to spatial environment variability).
Note, however, that by increasing the relative number of individuals of the
philopatric sex in good habitats, habitat-dependent biased sex ratio may
facilitate the evolution of helping.
A similar scenario may well occur in other territorial passerines (but
without helping behavior) in which offspring of one sex are more likely to
inherit their parents territory than offspring of the other sex. Females of
two such species (collared flycatcher Ficedula albicollis :
Ellegren et al., 1996 ; blue
tit Parus caeruleus : Svensson
and Nilsson, 1996) have been shown to bias their offspring sex
ratio in response to their mate's phenotypic quality : more young males (the
most philopatric sex) are produced when the father is of relatively good
quality, whereas more females are produced when the father is of lower
quality. However, male phenotypic quality may well be a measure of territory
quality (either used by the female, or by the biologist in the field). Indeed,
good males are likely to preempt good-quality territories
(Pulliam and Danielson, 1991).
Sex ratio was not biased when the father (of presumed good quality) was not
the territorial male (Sheldon and
Ellegren, 1996) : young resulting from extrapair copulation are
equally likely to be male or female. This suggests that territory rather than
male quality may indeed be the ultimate (and proximate ?) determinant of sex
ratio (although this was not the explanation proposed in these papers). Given
that in these two species territory quality is likely to vary in space, and
that male offspring are more likely than female offspring to inherit their
parents' territory (Greenwood et al.,
1979 ;
Pärt,
1990), such species may well fit into the assumptions of the
model. Hence, together with the case of the Seychelles warbler, the model is
able to challenge some of the most influential recent publications on bird sex
ratio (Sheldon, 1998).
Conclusion
The model presented here shows that, in species with sex-biased dispersal
and which reproduce in a spatially heterogeneous environment, an individual
may increase its fitness by biasing its offspring sex ratio toward the more
dispersing sex in poor-quality habitats and toward the more philopatric sex in
good-quality habitats. Although a similar arguments has been proposed
independently several times for particular cases [e.g., Young et al.
(1986) for spider mite ;
Reinhold (1998) for species
with environmental sex determination ; see also Robinson and O'Brien
(1991) for a group-living
monkey], I have shown the generality of this model and identified several
conditions under which highly biased sex ratios are likely to occur. Of these
conditions, the most important ones appear to be high between-habitat
dispersal rates for both sexes and a reduced possibility of habitat selection.
Temporal variability of habitat quality may also be critical, but further
studies are required. The geographic scale at which environmental
heterogeneity and between-habitat dispersal are measured has been shown to be
of critical importance for testing the model.
The model predictions have been tested with cases of habitat-dependent sex ratio in very different species (phytophagous arthropods, species with environmental sex determination, territorial passerines). Although none of these examples was conclusive, the model offers a plausible explanation for the evolution of biased sex ratios, sometimes alternative to the widely accepted ones. This model may thus highlight some biological cases where the common hypothesis does not match (or matches too well), and should be taken into account for further studies on the evolution of the sex ratio.
|
ACKNOWLEDGEMENTS |
|---|
G. Bertault, J. Blondel, T. Boulinier, J. Clobert, R. A. Ims, R. Nager, A. -C. Prévot-Julliard, H. Prévot, and N. C. Stenseth provided opportunities for discussion and/or valuable comments and previous versions of the manuscript. Two anonymous referee encouraged me to use the Taylor and Frank approach to build the model presented here. I was further helped in this task by D. Couvet, J. Fejoz, and J. Shykoff and her team. During most of the time spent preparing the manuscript, I was supported by grants from the Norwegian Science Council and the University of Oslo (both to N. C. Stenseth).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONThe modelRESULTSDISCUSSIONREFERENCES |
|---|
Bull JJ, 1983. Evolution of sex determining mechanisms. Menlo Park, California : Benjamin Cummings.
Bulmer MG, 1986. Sex ratios in geographically structured populations. Trends Ecol Evol 1 : 35-38.
Burley N, 1981. Sex ratio manipulation and selection
for attractiveness. Science 211 :
721-722.
Carr A, Carr MH, 1972. Site fixity in the Caribbean green turtle. Ecology 53 : 425-429.[Web of Science]
Charnov EL, Bull JJ, 1977. When is sex environmentally determined ? Nature 266 : 828-830.[Medline]
Clark AB, 1978. Sex ratio and local resource competition in a prosimian primate. Science 208 : 1157-1159.
Craig TP, Price PW, Itami JK, 1992. Facultative sex ratio shifts by a herbivorous insect in response to variation in host plant quality. Oecologia 92 : 153-161.[Web of Science]
Daan S, Dijkstra C, Weissing FJ, 1996. An evolutionary
explanation for seasonal trends in avian sex ratios. Behav Ecol
7 : 426-430.
Dhondt A, Adriaensen F, Matthysen F, Kempenaers B, 1990. Nonadaptive clutch size in tits. Nature 348 : 723-725.
Ellegren H, Gustafsson L, Sheldon BC, 1996. Sex ratio
adjustment in relation to paternal attractiveness in a wild bird population.
Proc Natl Acad Sci USA 93 :
11723-11728.
Ellegren H, Sheldon BC, 1996. New tools for sex identification and the study of sex allocation in birds. Trends Ecol Evol 11 : 255-259.[Web of Science]
Emlen ST, 1994. Benefits, constraints and the evolution of the family. Trends Ecol Evol 9 : 282-285.
Emlen ST, 1997. When mothers prefer daughters over sons. Trends Ecol Evol 12 : 291-292.
Emlen ST, Emlen JM, Levin SA, 1986. Sex ratio selection in species with helpers-at-the-nest. Am Nat 127 : 1-8.[Web of Science]
Ewert MA, Kackson DR, Nelson CE, 1994. Patterns of temperature-dependent sex determination in turtles. J Exp Zool 270 : 3-15.[Web of Science]
FitzSimmons, NN, Limpus CJ, Norman JA, Goldizen AR, Miller JD,
Moritz C, 1997. Philopatry of male marine turtles inferred from
mitochondrial DNA markers. Proc Natl Acad Sci USA
94 : 8912-8917.
Fisher RA, 1930. The genetical theory of natural selection. Oxford : Clarendon Press.
Fretwell SD, Lucas HL Jr, 1970. On territorial behaviour and other factors influencing habitat distribution in birds. Acta Biotheor 19 : 16-36.
Godfray HCJ, Werren JH, 1996. Recent developments in sex ratio studies. Trends Ecol Evol 11 : 59-63.
Greenwood PJ, 1980. Mating systems, philopatry and optimal dispersal in birds and mammals. Anim Behav 28: 1140-1162.
Greenwood PJ, Harvey PH, Perrins CM, 1979. The role of dispersal in the great tit (Parus major), the causes, consequences and heritability of natal dispersal. J Anim Ecol 48: 123-142.
Hamilton WD, 1967. Extraordinary sex ratios.
Science 156:
477-488.
Hamilton WD, May RM, 1977. Dispersal in stable habitats. Nature 269 : 578-581.
Heimpel GE, 1997. Extraordinary sex ratios for extraordinary reasons. Trends Ecol Evol 12: 298-299.
Holt RD, 1996. Adaptive evolution in source-sink environments : direct and indirect effects of density-dependence on niche evolution. Oikos 75: 182-192.[Web of Science]
Janzen FJ, 1996. Is temperature-dependent sex determination in reptiles adaptive ? Trends Ecol Evol 11: 253.
Johnson CN, 1988. Dispersal and sex ratio at birth in primates. Nature 332: 726-728.[Medline]
Johnson ML, Gaines MS, 1990. Evolution of dispersal : theoretical models and empirical tests using bird and mammals. Annu Rev Ecol Syst 21: 449-480.
Koening WD, Pitelka FA, Carmen WJ, Mumme RL, Stanback MT, 1992. The evolution of delayed dispersal in cooperative breeders. Q Rev Biol 67: 111-150.[Medline]
Komdeur J, 1992. Importance of habitat saturation and territory quality for evolution of cooperative breeding in the Seychelles warbler. Nature 358: 493-495.
Komdeur J, 1994. Experimental-evidence for helping and hindering by previous offspring in the cooperative-breeding Seychelles warbler Acrocephalus sechellensis. Behav Ecol Sociobiol 34: 175-186.[Web of Science]
Komdeur J, 1996a. Facultative sex ratio bias in the
offspring of Seychelles warblers. Proc R Soc Lond B
263: 661-666.
Komdeur J, 1996b. Influence of helping and breeding
experience on reproductive performance in the Seychelles warbler : a
translocation experiment. Behav Ecol 7:
326-333.
Komdeur J, Daan S, Tinbergen J, Mateman C, 1997. Extreme adaptive modification in sex ratio of the Seychelles warbler's eggs. Nature 385: 522-525.
Lambin X, 1994. Sex ratio variation in relation to female philopatry in Townsend's voles. J Anim Ecol 63: 945-953.
Leimar O, 1996. Life-history analysis of the Trivers
and Willard sex ratio problem. Behav Ecol
7: 316-325.
Li JB, Margolies DC, 1994. Responses to direct and indirect selection on aerial dispersal behavior in Tetranychus urticae. Evolution 72: 10-22.
Lohmann KJ, Lohmann CMF, 1996. Detection of magnetic field intensity by sea turtles. Nature 380: 59-61.
McPeek, MA, Holt RD, 1992. The evolution of dispersal in spatially and temporally varying environments. Am Nat 140: 1010-1027.[Web of Science]
Mopper S, Whitham TG, 1992. The plant sex paradox : effects on pinyon sawfly sex ratios and fecundity. Ecology 73: 515-525.[Web of Science]
Myers JH, Boettner G, Elkinton J, 1998. Maternal effect in gypsy moth : only sex ratio varies with population density. Ecology 79: 305-314.[Web of Science]
Nordborg M, 1991. Sex ratio selection with general migration schemes : Fisher's result does hold. Evolution 45: 1289-1293.[Web of Science]
Pärt T, 1990. Natal dispersal in the collared flycatcher Ficedula albicollis : possible causes and reproductive consequences. Ornis Scand 21: 83-88.
Pulliam HR, 1988. Sources, sinks, and population regulation. Am Nat 132: 652-661.[Web of Science]
Pulliam HR, Danielson BJ, 1991. Sources, sinks, and habitat selection : a landscape perspective on population dynamic. Am Nat 137: 550-566.
Pusey A, 1987. Sex-biased dispersal and inbreeding avoidance in birds and mammals. Trends Ecol Evol 2: 295-299.
Reinhold K, 1996. Sex-ratio selection with asymmetrical migration of the sexes can lead to an uneven sex ratio. Oikos 75: 15-19.[Web of Science]
Reinhold K, 1998. Nest-site philopatry and selection for environmental sex determination. Evol Ecol 12: 245-250.
Robinson JG, O'Brien TG, 1991. Adjustment in birth sex ratio in wedge-capped capuchin monkeys. Am Nat 138: 1173-1186.[Web of Science]
Ruohomäki K, Haukioja E, Walter GH, 1993. Resource quality and spatial variation in sex ratios of free-living solitary sawfly, Dineura virididorsata (Hym., Tenthredinidae). Oecologia 96: 420-425.[Web of Science]
Schwartz MP, 1988. Local resource enhancement and sex ratios in a primitively social bee. Nature 331: 346-348.
Sheldon BC, 1998. Recent studies of avian sex ratios. Heredity 80 : 397-402.[Web of Science]
Sheldon BC, Ellegren H, 1996. Offspring sex and
paternity in the collared flycatcher. Proc R Soc Lond B
263: 1017-1021.
Stein SJ, Price PW, Craig TP, Itami JK, 1994. Dispersal of a galling sawfly : implications for further studies of insect population dynamics. J Anim Ecol 63: 666-676.
Stiefel VL, Margolies DC, 1992. Do components of colonization—dispersal cycles affect the offspring sex ratios of Banks grass mites (Oligonychus pratensis) ? Entomol Exp Appl 64: 161-166.
Svensson E, Nilsson J-
,
1996. Mate quality affects offspring sex ratio in blue tits.
Proc R Soc Lond B 263:
357-361.
Taylor PD, 1994. Sex ratio in a stepping-stone population with sexspecific dispersal. Theor Popul Biol 45: 203-218.[Web of Science][Medline]
Taylor PD, Frank SA, 1996. How to make a kin selection model. J Theor Biol 180: 27-37.[Web of Science][Medline]
Trivers RL, Willard DE, 1973. Natural selection of
parental ability to vary the sex ratio of offspring. Science
179: 90-92.
Varndell NP, Godfray HCJ, 1996. Facultative adjustment of the sex ratio in an insect (Planococcus citri, Pseudococcidae) with paternal genome loss. Evolution 50: 2100-2105.[Web of Science]
Walter GH, Ruohomäki K, Haukioja E, Vainio E, 1994. Reproductivebehavior of mated and virgin females of a solitary sawfly Dineura virididorsata. Entomol Exp Appl 70: 83-90.
Wolfram S, 1997. The Mathematica book. Cambridge : Cambridge University Press.
Wrensch, DL, Ebbert MA, 1993. Evolution and diversity of sex ratio in insects and mites. New York : Chapman and Hall.
Wrensch DL, Young SSY, 1978. Effects of density and host quality on rate of development, survivorship, and sex ratio in carmine spider mite. Environ Entomol 7: 499-501.[Web of Science]
Young SSY, Wrensch DL, Kongchuensin M, 1986. Control of sex ratio by female spider mites. Entom Exp Appl 40: 53-60.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. B. Hjernquist, K. A. Thuman Hjernquist, J. T. Forsman, and L. Gustafsson Sex allocation in response to local resource competition over breeding territories Behav. Ecol., March 1, 2009; 20(2): 335 - 339. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Forsman, M. B. Hjernquist, J. Taipale, and L. Gustafsson Competitor density cues for habitat quality facilitating habitat selection and investment decisions Behav. Ecol., February 19, 2008; (2008) arn005v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||