letters to nature
7. Herbers, J. M. On caste ratios in ant colonies: population responses to changing environments.
Evolution 14, 575±585 (1980).
8. Hasegawa, E. The optimal caste ratio in polymorphic ants: estimation and empirical evidence. Am.
Nat. 149, 706±711 (1997).
9. Passera, L., Roncin, E., Kaufmann, B. & Keller, L. Increased soldier production in ant colonies exposed
to intraspeci®c competition. Nature 379, 630±631 (1996).
10. Johnston, A. B. & Wilson, E. O. Correlates of variation in the major/minor ratio of the ant, Pheidole
dentata (Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 78, 8±11 (1985).
11. Hasegawa, E. Nest defense and early production of the major workers in the dimorphic ant Colobopsis
nipponicus (Wheeler) (Hymenoptera: Formicidae). Behav. Ecol. Sociobiol. 33, 73±77 (1993).
12. Wilson, E. O. The relation between caste ratios and division of labor in the ant genus Pheidole
(Hymenoptera: Formicidae). Behav. Ecol. Sociobiol. 16, 89±98 (1984).
13. Beshers, S. N & Traniello, J. F. The adaptiveness of worker demography in the attine ant Trachymyrmex
septentrionalis. Ecology 75, 763±775 (1994).
14. Gordon, D. M. Group-level dynamics in harvester ants: young colonies and the role of patrolling.
Anim. Behav. 35, 833±843 (1987).
15. Gordon, D. M. Dynamics of task switching in harvester ants. Anim. Behav. 38, 194±204 (1989).
16. Strand, M. R. & Grbic, M. The development and evolution of polyembryonic insects. Curr. Top. Dev.
Biol. 35, 121±159 (1997).
17. Strand, M. R. Oviposition behavior and progeny allocation of the polyembryonic wasp Copidosoma
¯oridanum (Hymenoptera: Encyrtidae). J. Insect Behav. 2, 355±368 (1989).
18. Ode, P. J. & Strand, M. R. Progeny and sex allocation decisions of the polyembryonic wasp,
Copidosoma ¯oridanum. J. Anim. Ecol. 64, 213±224 (1995).
19. Browning, H. W. & Oatman, E. R. Intra- and interspeci®c relationships among some parasites of
Trichoplusia ni. Environ. Entomol. 13, 551±556 (1984).
20. Strand, M. R., Johnson, J. A. & Culin, J. D. Intrinsic, interspeci®c competition between the
polyembryonic parasitoid, Copidosoma ¯oridanum and solitary endoparasitoid Microplitis demolitor
in Pseudoplusia includens. Entomol. Exp. Appl. 55, 275±284 (1990).
21. Godfray, H. C. J. Parasitoids: Behavioral and Evolutionary Ecology (Princeton Univ. Press, Princeton,
1994).
22. Wheeler, D. E. The developmental basis of worker caste polymorphism in ants. Am. Nat. 138, 1218±
1238 (1991).
23. Grbic, M, Nagy, L. M. & Strand, M. R. Development of polyembryonic insects: a major departure from
typical insect embryogenesis. Dev. Genes Evol. 208, 69±81 (1998).
24. Baehrecke, E. H., Aiken, J. M., Dover, B. A. & Strand, M. R. Ecdysteroid induction of embryonic
morphogenesis in a parasitic wasp. Dev. Biol. 158, 275±287 (1993).
25. Grbic, M., Rivers, D. & Strand, M. R. Caste formation in the polyembryonic wasp Copidosoma
¯oridanum (Hymenoptera: Encyrtidae): in vivo and in vitro analysis. J. Insect Physiol. 43, 553±565
(1997).
26. Strand, M. R. & Pech, L. L. Immunological basis for compatibility in parasitoid-host relationships.
Annu. Rev. Entomol. 40, 31±56 (1995).
27. Strand, M. R., Witherell, R. A. & Trudeau, D. Two Microplitis demolitor polyDNAvirus mRNAs
expressed in hemocytes of Pseudoplusia includens contain a common cysteine-rich domain. J. Virol.
71, 2146±2156 (1997).
28. Webb, B. A. in The Insect Viruses (eds Miller, L. K. & Ball, A. L.) 105±139 (Plenum, New York, 1998).
Acknowledgements
We thank J. Johnson for assistance during the study, M. Grbic for discussions, and
M. Visser, F. Wackers and W. van der Putten for comments on an early draft of the
manuscript. This work was supported by the NSF.
Correspondence and requests for materials should be addressed to M.R.S.
(e-mail: [email protected]).
.................................................................
stress6±9. Here we show that genetic variation underlies the
response to environmental stress (variable food quality) of a
sexual ornament (male eye span) in the stalk-eyed ¯y
Cyrtodiopsis dalmanni. Some male genotypes develop large eye
span under all conditions, whereas other genotypes progressively
reduce eye span as conditions deteriorate. Several non-sexual
traits (female eye span, male and female wing length) also show
genetic variation in condition-dependent expression, but their
genetic response is entirely explained by scaling with body size. In
contrast, the male sexual ornament still reveals genetic variation
in the response to environmental stress after accounting for
differences in body size. These results strongly support the
hypothesis that female mate choice yields genetic bene®ts for
offspring.
In the Malaysian stalk-eyed ¯y C. dalmanni, males have exaggerated eye span and females show strong mating preferences for males
with greater eye span10,11. Here we investigated how genotype and
environment (and their interaction) affected the development of
sexual and non-sexual traits. Experimental males were mated to
virgin females to generate full and half-sibling families (see Methods).
Eggs and larvae from each family were cultured on corn, spinach or
damp cotton wool, which represent food environments of decreasing nutritional value. Emerging ¯ies were sexed, and the male sexual
trait (eye span), several comparable non-sexual traits (female eye
span, male and female wing length), and body size (thorax length)
were measured.
Food quality, as expected, had a large effect on absolute trait sizes
(Table 1). For all traits and both sexes, spinach-fed ¯ies were smaller
than corn-fed ¯ies (P , 0.001). However, the main change occurred
with the shift to the poorest environment, ¯ies that had fed on
cotton wool being dramatically smaller than ¯ies that had fed
on spinach (P , 0.001, Fig. 1). Despite this strong conditiondependent environmental effect, there was clear evidence for genetic
variation in all traits, in both sexes (Table 1). This was still the case
when each environment was analysed separately (all P , 0.05).
Of greater interest were the interactions between effects due to
food quality and effects due to genotypeÐthat is, food quality by
genotype interactions as these test whether there is a genetic basis
for environmental condition dependence. For absolute male eye
span, the interaction was signi®cant (Table 1). Comparisons of the
genetic variance in each environment showed that differences
between genotypes increased under food stress, especially in the
cotton-wool environment (Fig. 1; corn±wool F29,125 = 1.937, P =
0.007; spinach±wool F31,113 = 1.664, P = 0.028; corn±spinach F31,67 =
1.462, P = 0.099; tested using the Satterthwaite method, see
Methods). Nonetheless genotype ranks were maintained across
Condition-dependent signalling
of genetic variation in
stalk-eyed ¯ies
Patrice David*², Tracey Bjorksten*³, Kevin Fowler*
& Andrew Pomiankowski*
* The Galton Laboratory, Department of Biology, University College London,
4 Stephenson Way, London NW1 2HE, UK
² CEFE-CNRS, 1919 route de Mende, 34293 Montpellier Cedex 05, France
³ Department of Biology, Colorado State University, Fort Collins, Colorado 80523,
USA
..............................................................................................................................................
Handicap models of sexual selection predict that male sexual
ornaments have strong condition-dependent expression and this
allows females to evaluate male genetic quality1±5. A number of
previous experiments have demonstrated heightened conditiondependence of sexual ornaments in response to environmental
186
Figure 1 Genotypic responses for absolute male and female eye span in corn, spinach
and cotton-wool environments. Each line represents the mean eye span of a group of
full-siblings.
© 2000 Macmillan Magazines Ltd
NATURE | VOL 406 | 13 JULY 2000 | www.nature.com
letters to nature
Table 1 Absolute trait value ANOVA
Males²
Females
Eye span
Wing length
Thorax length
Eye span
Wing length
Thorax length
85.91***
3.62***
2.05***
251
83.20***
3.41***
2.20***
241
97.73***
2.87***
1.59**
254
68.41***
2.39***
1.43*
279
76.24***
3.09***
1.82**
286
72.78***
2.19**
1.91***
286
...................................................................................................................................................................................................................................................................................................................................................................
Food quality
Genotype
Food quality ´ genotype
N
...................................................................................................................................................................................................................................................................................................................................................................
Shown are the environmental effects due to variation in food quality, the genetic effects due to variation amongst full-sibling families, and interactions.
² F values are given for each effect; ***P , 0.001, **P , 0.01, *P , 0.05; the degrees of freedom were d.f. = 2 (food quality), d.f. = 24±27 (genotypes) and d.f. = 48±54 (interaction).
environments (genetic correlation coef®cients 6 s.e.: corn±spinach
ra = 0.660 6 0.106, P , 0.01; corn±wool ra = 0.576 6 0.148,
P , 0.001; spinach±wool ra = 0.407 6 0.207, P , 0.001), so a
genotype which performed well in one environment tended to do
well in all environments (Fig. 1). However, similar food quality
by genotype interactions were present for the non-sexual traits
(Table 1). Most of these (thorax and wing length in both sexes, but
not female eye span) also showed increased variance in stressful
environments and ranks maintained across environments (data
not shown). So absolute measures of both sexual and non-sexual
traits showed similar genetic components to their environmental
condition-dependence.
These patterns may simply re¯ect scaling with body size. To
discount this, we calculated relative trait values, dividing eye span
and wing length by thorax length (see Methods). Relative male eye
span showed the same strong environmental condition-dependence
as absolute male eye span (Tables 1 and 2, similar F values).
However, removing the effect of body size from the non-sexual
characters reduced environmental condition-dependence of male
and female wing length and female eye span (Tables 1 and 2, lower F
values). These results are consistent with previous work in stalkeyed ¯ies demonstrating that sexual ornaments are more sensitive to
environmental conditions than non-sexual traits9. Genetic variation
Figure 2 Genotypic responses for relative male and female eye span across the three
environments. Relative values are eye span divided by thorax length for each sex. Each
line represents the mean eye span of a group of full-siblings.
remained strong for relative trait values of both sexes, indicating
that traits have genetic variation independent from body size
(Table 2).
Food quality by genotype interactions showed similar changes.
The strength of the food quality by genotype interaction remained
unchanged for relative male eye span (Table 2). Once again
genetic variance increased under food stress (Fig. 2, corn±wool
F32,117 = 1.981, P = 0.004; spinach±wool F34,99 = 1.802, P = 0.013;
corn±spinach F41,65 = 1.140, P = 0.315), and genotype ranks were
maintained across environments (corn±spinach ra = 0.624 6 0.177,
P , 0.001; corn±wool ra = 0.877 6 0.060, P , 0.001; spinach±wool
ra = 0.206 6 0.262, P , 0.001). In contrast, removing body size
scaling from the non-sexual traits eliminated or greatly reduced any
food quality by genotype interaction (Table 2). In addition, despite
genetic variation being present in relative trait values of the nonsexual traits (Table 2), there was no increase in genetic variation in
more stressful environments (Fig. 2, all P . 0.05). So for relative
trait values, only male eye span showed a genetic basis to environmental condition-dependence.
Evidence that sexual ornaments have condition-dependent
expression comes largely from studies using experimental manipulations of environmental stress6±9. If females use sexual traits to
evaluate male genetic quality these ®ndings are paradoxical, as
sensitivity to environmental variation can potentially mask genetic
effects. Reduced or low heritability of sexual traits in the face of high
environmental variation has been reported in several semi-natural
studies of bird species12,13. However, in our experiment, variation
between environments, in terms of changes in food quality, magni®ed the differences between male genotypes (especially in the most
stressful environment).
In addition, our experiment revealed the unique nature of the
sexual trait, male eye span. We compared sexual and non-sexual
traits, whilst controlling for general body size effects. Male eye span
alone showed a genetic basis to environmental condition-dependence that was independent of body size scaling, and increased
genetic variance in stressful environments. This echoes previous
®ndings that the expression of male eye span in stalk-eyed ¯ies is
strongly informative about condition9,14. Taken overall, the evidence
points to a `good genes' interpretation of male sexual signalling in
stalk-eyed ¯ies, as condition-dependence is predicted by handicap
models of sexual selection1±5. In the future, it will be important to
establish whether there are similar genetic responses to environmental condition-dependence in the exaggerated sexual traits of
other species.
M
Table 2 Relative trait value ANOVA
Methods
1.2
males
females
Relative eye span
1.1
1
0.9
0.8
0.7
0.6
Corn
Spinach
Cotton wool
Food
Males²
Females
Relative
eye span
Relative wing
length
Relative
eye span
Relative wing
length
88.85***
3.82***
2.06***
243
10.55***
4.38***
0.73 NS
232
1.66 NS
2.57***
1.07 NS
266
29.05***
3.41***
1.454*
272
.............................................................................................................................................................................
Food quality
Genotype
Food quality ´ genotype
N
.............................................................................................................................................................................
Shown are the environmental, genetic and interaction effects.
² F values are given for each effect; ***P , 0.001, **P , 0.01, *P , 0.05, NS P . 0.05; the degrees of
freedom were d.f. = 2 (food quality), d.f. = 24±26 (genotypes) and d.f. = 48±52 (interaction).
NATURE | VOL 406 | 13 JULY 2000 | www.nature.com
Fly collection
Flies used in the experiment were taken from stock populations established in 1993 from
®eld collections near the Gombak river in Malaysia, and maintained in population cages of
at least 200 individuals at 25 8C, 12 h:12 h light:dark cycle, on ground-corn medium. Virgin
males (N = 18) were collected and mated sequentially to ®ve virgin females, for two days
each, to generate full-sibling and half-sibling families. Mated females were kept separately in
large cages (20 ´ 20 ´ 40 cm) and were alternately presented with small weigh boats
containing corn or spinach as a food source and egg lay medium. Weigh boats were replaced
every two days and kept in individual pots to allow larval growth. Excess food (30 g per boat)
was given to minimize larval competition. Females also laid eggs on the damp, cotton-wool /
®lter-paper cage lining. All emerging ¯ies from these three sources were collected and frozen.
© 2000 Macmillan Magazines Ltd
187
letters to nature
.................................................................
Measurements
The number of progeny was highly variable between families and environments. Twentynine families consisting of 12 sires mated to 2 or 3 dams provided suf®cient progeny (®ve
male and ®ve female) from all food treatments. Stored ¯ies were sexed before measurement of eye-stalk length (median eye-stalk bristle to the centre of the head)9 and wing
length using a video camera mounted on a monocular microscope and the image analysis
program NIH Image (Version 1.55). In addition thorax length was measured from the
centre of the head to the lower edge of the thorax. All measurements were made `blind' by a
single person (T.B.). Two replicates of each measurement were taken on different days.
Values were averaged over the left and right sides and the two replicates.
Statistical analysis
For each sex and each trait separately, a mixed-model ANOVA was evaluated with factors
food quality (®xed), genotype (random; a genotype was de®ned as a full-sibling family)
and the interaction15. As some families shared the same father, the analysis was repeated
using hierarchical ANOVA with effects sire and dam-within-sire. Qualitatively similar
relationships were found. The analysis was also carried out with pairs of environments to
compute differences between environments, genetic correlation coef®cients15,16 and their
standard errors17. Genetic correlation coef®cients between environments are de®ned as the
correlation between breeding values of the same genotype in two environments16. The
equality of genetic variances between two environments was tested using Satterthwaite's
k MS ‡k MS
method15 based on the ratio F ˆ k21 MSff ;1;2 ‡k12 MSe:2
, where subscripts f and e refer to family and
e;1
error mean squares, 1 and 2 to two environments, ki to the number of replicates per family
in environment i and MS to the mean squares. The approximate degrees of freedom for
this F test are given in ref. 15. This ratio is constructed so that the denominator and the
numerator have equal expectations when genetic variances are equal in both environments, even if error variance and sample sizes differ. For all traits and all environments,
hierarchical ANOVAs with effects sire and dam-within-sire revealed no evidence of
dominance or common environments as the dam component of variance was not
signi®cantly larger than the sire component15,16. So our estimates of ra are not in¯ated by
dominance or effects due to common environments. The analysis was repeated using
relative trait values, with thorax length used as a general measure of body size. We used
relative values rather than residuals. Analysis of residuals was more complex, as traits
displayed changes in their allometric slopes and intercepts across environments. A full
analysis using residuals reached the same conclusions (K.F. et al., manuscript in
preparation). Sample sizes were not the same for different traits because occasionally
specimens exhibited damage to heads or wings. Given the number of tests performed,
results with borderline signi®cance (0.01 , P , 0.05) are treated with caution in our
discussion.
Received 8 March; accepted 18 May 2000.
1. Andersson, M. Evolution of condition-dependent sex ornaments and mating preferences: sexual
selection based on viability differences. Evolution 40, 804±816 (1986).
2. Pomiankowski, A. Sexual selection: the handicap principle does workÐsometimes. Proc. R. Soc. Lond.
B 231, 123±145 (1987).
3. Grafen, A. Biological signals as handicaps. J. Theor. Biol. 144, 517±546 (1990).
4. Iwasa, Y. & Pomiankowski, A. The evolution of mate preferences for multiple handicaps. Evolution 48,
853±867 (1994).
5. Rowe, L. & Houle, D. The lek paradox and the capture of genetic variance by condition dependent
traits. Proc. R. Soc. Lond. B 263, 1415±1421 (1996).
6. Hill, G. E. Proximate basis of variation in carotenoid pigmentation in male house ®nches. Auk 109, 1±
12 (1992).
7. Gustaffson, L., QvarnstroÈm, A. & Sheldon, B. C. Trade-offs between life-history traits and a secondary
sexual character in male collared ¯ycatchers. Nature 375, 311±313 (1995).
8. Hunt, J. & Simmons, L. W. Patterns of ¯uctuating asymmetry in beetles' horns: an experimental
examination of the honest signalling hypothesis. Behav. Ecol. Sociobiol. 41, 109±114 (1997).
9. David, P. et al. Male sexual ornament size but not asymmetry re¯ects condition in stalk-eyed ¯ies.
Proc. R. Soc. Lond. B 265, 1±6 (1998).
10. Burkhardt, D. & de la Motte, I. Big `antlers' are favoured: female choice in stalk-eyed ¯ies (Diptera,
Insecta), ®eld collected harems and laboratory experiments. J. Comp. Physiol. A 162, 649±652 (1988).
11. Wilkinson, G. S. & Reillo, P. R. Female choice response to arti®cial selection on an exaggerated male
trait in a stalk-eyed ¯y. Proc. R. Soc. Lond. B 255, 1±6 (1994).
12. Grif®ths, S. C., Owens, I. P. F. & Burke, T. Environmental determination of a sexually selected trait.
Nature 400, 358±360 (1999).
13. Qvarnstrom, A. Genotype-by-environment interactions in the determination of the size of a
secondary sexual character in the collared ¯ycatcher (Ficedula albicollis). Evolution 53, 1564±1572
(1999).
14. Wilkinson, G. S. & Taper, M. Evolution of genetic variation for condition-dependent traits in stalkeyed ¯ies. Proc. R. Soc. Lond. B 266, 1685±1690 (1999).
15. Lynch, M. & Walsh, B. Genetics and Analysis of Quantitative Traits (Sinauer Associates, Sunderland,
Massachussetts, 1998).
16. Roff, D. A. Evolutionary Quantitative Genetics (Chapman & Hall, London, 1997).
17. Falconer, D. S. & Mackay, T. F. C. Introduction to Quantitative Genetics 4th edn (Longman, Harlow,
1996).
Acknowledgements
We thank T. Chapman and G. Hurst for comments, and A. Hingle for help in rearing ¯y
stocks. This work was supported by Royal Society university research fellowships (K.F. and
A.P.) and the NERC.
Correspondence and requests for materials should be addressed to A.P.
(e-mail: [email protected]).
188
The segment polarity network is a
robust developmental module
George von Dassow, Eli Meir, Edwin M. Munro & Garrett M. Odell
University of Washington, Department of Zoology, Box 351800, Seattle,
Washington 98195-1800, USA
..............................................................................................................................................
All insects possess homologous segments, but segment speci®cation differs radically among insect orders. In Drosophila, maternal morphogens control the patterned activation of gap genes,
which encode transcriptional regulators that shape the patterned
expression of pair-rule genes. This patterning cascade takes place
before cellularization. Pair-rule gene products subsequently
`imprint' segment polarity genes with reiterated patterns, thus
de®ning the primordial segments. This mechanism must be
greatly modi®ed in insect groups in which many segments
emerge only after cellularization1. In beetles and parasitic
wasps, for instance, pair-rule homologues are expressed in patterns consistent with roles during segmentation, but these patterns emerge within cellular ®elds2±4. In contrast, although in
locusts pair-rule homologues may not control segmentation5,6,
some segment polarity genes and their interactions are
conserved3,7±10. Perhaps segmentation is modular, with each
module autonomously expressing a characteristic intrinsic behaviour in response to transient stimuli. If so, evolution could
rearrange inputs to modules without changing their intrinsic
behaviours. Here we suggest, using computer simulations, that
the Drosophila segment polarity genes constitute such a module,
and that this module is resistant to variations in the kinetic
constants that govern its behaviour.
Gap and pair-rule gene products are nuclear proteins that form
short-range gradients in the Drosophila syncytial blastoderm, locally
modulating each other's expression through direct transcriptional
control. In contrast, segment polarity genes re®ne and maintain
their expression state through a network of cross-regulatory interactions that require cell±cell communication1±15. Many segment
polarity genes, unlike gap and pair-rule genes, remain active
throughout development; the segment polarity network remembers
the pattern imprinted upon it, then provides positional read-outs
for subsequent developmental processes, including speci®cation of
neuroblasts, denticle patterns and appendage primordia. Thus, the
intrinsic behaviour of the putative segment polarity module consists
of stable, reiterated, asymmetric expression of its constituent genes,
especially of the principal outputs wingless (wg), hedgehog (hh) and
engrailed (en)15; in Drosophila the transient stimuli are pair-rule
genes.
We used computer simulations to investigate whether the known
interactions among segment polarity genes suf®ce to confer the
properties expected of a developmental module. Given its probable
conservation among diverse insects and perhaps beyond, we
expected such a module to exhibit buffering against quantitative
changes in gene function, and to be insensitive to the exact nature of
input stimuli. We summarized (Box 1) the interactions among those
segment polarity gene products that we reasoned might suf®ce to
mimic the wild-type expression patterns of segment polarity genes
(Fig. 1a). We abbreviated intermediate pathways and did not
explicitly represent `generic' components such as the transcriptional
machinery. To formulate a dynamical model based on Box 1a
required nearly 50 free parameters, including half-lives of messenger
RNAs and proteins, binding rates, and cooperativity coef®cients.
The real values of these are unknown and the biologically realistic
range for most parameters spans several orders of magnitude.
Therefore we asked: is there any set of parameter values for
© 2000 Macmillan Magazines Ltd
NATURE | VOL 406 | 13 JULY 2000 | www.nature.com