Sexual Conflict in the Dung Fly Sepsis cynipsea

Author(s): P. I. Ward, J. Hemmi and T. Roosli

Source: Functional Ecology, Vol. 6, No. 6 (1992), pp. 649-653
Published by: British Ecological Society
Stable URL: https://www.jstor.org/stable/2389959
Accessed: 20-03-2019 03:12 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

British Ecological Society is collaborating with JSTOR to digitize, preserve and extend access
to Functional Ecology

This content downloaded from 152.118.231.148 on Wed, 20 Mar 2019 03:12:46 UTC
All use subject to https://about.jstor.org/terms
Functional
Ecology 1992
Sexual conflict in the dung fly Sepsis cynipsea
6, 649-653
P. I. WARD, J. HEMMI and T. ROOSLI
Zoologisches Museum der Universitat Zurich, Winterthurerstr. 190, 8057 Zurich, Switz

Summary

1. The interests of a male and a female over whether mating should occur are often in
conflict and in insects conflict resolution usually favours the female.
2. A male dung fly Sepsis cynipsea guards a female while she oviposits in fresh cow
dung.
3. The amount of time a female spends laying depends only on the number of eggs
laid.
4. When she leaves the dung, there is a conflict over whether or not mating should
occur.
5. A discriminant analysis shows that features of both the female and male behaviour
influence the outcome of the conflict. The proportion of time the female spends
shaking to dislodge the male and how long she spends laying were more important
than the male behaviour apparently aimed at establishing genital contact.
6. If a pair went on to copulate, only the duration of the conflict influenced the copula
duration.
7. We conclude that the female largely determines if copula occurs, possibly on the
basis of her sperm reserves, but if copula does take place then the male determines
the duration.
8. A female may in general avoid copula because of the risk of internal injury by the
male's armoured genitalia.

Key-words: Copula, genitalia, mating system, sexual selection

Functional Ecology (1992) 6, 649-653

Introduction mating by females. The evidence was weakly in

favour: where there was 'female control' (e.g.
There is often conflict between a male and a female as because the female emits a pheromone to attract
to whether mating should occur (Parker 1979). The males) multiple mating was less frequent than where
outcome of the conflict should depend on the selec- there was 'male control' (e.g. because males control
tion intensities operating on each participant and access to resources required by the females). A
asymmetries in their aims. Parker (1979) argues that number of other factors may also influence female
in the conflict over whether mating should occur, the mating frequency (see references in Ridley 1990) and
objectives of the participants are normally different; non-adaptive explanations are also possible (Halli-
usually males will be selected to mate and females to day & Arnold 1987).
resist. The costs of resistance for females, e.g. wasted We have examined the mating system of the dung
time or the risk of injury, may affect the outcome. fly Sepsis cynipsea (L.) (Parker 1972a,b). Males
Female insects may be expected to 'win' most mating arrive very quickly after the deposition of a cow pat.
conflicts, partly because they are generally larger When a female arrives to lay eggs in the dung, she will
than the males (Parker 1979; Eberhard 1985). None- normally be mounted by a male, which does not then
theless, females may sometimes be selected to mate attempt to establish genital contact. The male
to avoid continuous harassment by a male; a condi- remains on the female, defending her from other
tion Thornhill & Alcock (1983) call 'convenience males, while she oviposits. When the female leaves
polyandry'. the dung, there is a behavioural interaction which
In species where a reproductive female encounters appears to be, and we shall treat as, a mating conflict;
a number of males, as in many insects, the resolution the female shakes, often vigorously, in an attempt to
of mating conflicts may be reflected in the frequency dislodge the male and often extrudes her ovipositor
of multiple mating by females. Ridley (1990) tested while the male attempts to establish prolonged
the hypothesis that the sex which has 'control' over genital contact and copula. About 40% of pairs
649 mating should influence the frequency of multiple formed eventually copulate. Parker (1972a) found

This content downloaded from 152.118.231.148 on Wed, 20 Mar 2019 03:12:46 UTC
All use subject to https://about.jstor.org/terms
650 differences between pairs which did and did not 30

P. 1. Ward et al. copulate in both female and male behaviour but did
not attempt to assess the relative importance of each.
j20 *
At the end of copula the male releases the female
with his legs and twists around the still-united Eb .0. 0
genitalia two or three times before he disengages
> 10 0
from the female (Parker 1972a). Ward (1983) showed
-J 0
that a male's body size influences his behaviour and 0~~~~

success in male-male competition in this species:

larger males are more successful in competition than 0 10 20 30 40 50 60 70

No. of eggs laid

smaller ones and the mate-searching behaviour of a
male is related to his size.
Fig. 1. The relationship between the number of eggs laid
The main aim of this paper is to assess the relative
and the time taken to lay them.
importance for the mating conflict resolution of male
and female body size and behaviour, and of the
events occurring before and during the conflict.
Natural variation was observed in behaviour, but we
recognize that the conclusions may be strengthened variations in duration. The proportions were arcsine
by experimental manipulations. square root transformed before analysis, as recom-
mended by Sokal & Rohlf (1981). The statistical
analyses were conducted using the SPSS package
Materials and methods
(SPSS 1990). The approach was to use stepwise
Pairs were observed as they arrived on fresh dung in multivariate procedures to identify variables from
various meadows near Zurich, Switzerland during previous stages in the sequence of behaviour with a
the summer of 1990. The time the female was on the significant relationship to subsequent variables.
dung was recorded. A female raises her abdomen in a Slopes in regression analyses (b) are given ? their
characteristic way each time she lays an egg and so SE.
the number of eggs laid could also be counted. The
frequency per minute of attacks by unpaired males on
Results
the pair (though no takeovers of the female were
observed) while the female oviposited was recorded, The time the female spent laying was used as the
and later used to indicate the degree of male-male dependent variable in a stepwise multiple regression
competition. When the female left the dung, the pair using the male and female sizes, the difference in size
was collected in a glass tube containing some grass between the two and the number of eggs laid as
stems. Pairs do not usually move far from the dung independent variables. Only the number of eggs laid
before either copulating or separating without was entered into the regression model and the
copulating and always crawl on grass stems rather regression was positive (b=0.30?0-02; model
than on the ground. The behaviour of males and F1,44=242-56, P<0.0001) (Fig. 1).
females did not differ noticeably from unconstrained The variables which may explain the variation in
pairs. The time taken until the pair either separated female shaking behaviour and male behaviour were
or began to copulate was recorded. During this time examined. These were: the sizes of the male and
the following behavioural traits were recorded: the female, the difference in size between the two, the
number of times the male 'stroked' the female with time the female had spent laying, the number of eggs
his abdomen and the number of occasions the male laid and the frequency per minute of attacks by
established 'temporary genital contact' [see Parker unpaired males during oviposition. Only the number
(1972a) for fuller descriptions of these male traits; it of eggs laid was related to the female shaking
is unlikely that sperm is transferred during temporary behaviour and the regression was negative
genital contact]; both these traits appear aimed at (b=-0.007+0.003, t=2*76, P<0-01; model
establishing extended genital contact and so copula. F1,44=7-74, P<0.01) (Fig. 2); i.e. a female shoo
The cumulative time the female spent shaking was the more eggs she had laid. The number of strokes
also recorded. The pair were then taken back to the made by the males was only related to female size and
laboratory where the hind tibia length of each animal the regression was positive (b=8-79?3-33, t=2-64,
was measured under a microscope. This was the P<0-02; model F1, 44=6-96, P<0-02) (Fig. 3). This
indicates that males attempt to attain copula more
index of body size previously used by Ward (1983).
The number of male strokes and temporary genital vigorously as female size increases. The number of
contacts per minute until the conflict was resolved, temporary genital contacts per minute was not signi-
the number of attacks by unpaired males per minute ficantly related to any of the other variables, prob-
during laying and the proportion of time the female ably because this factor interacts strongly with the
spent shaking were calculated to correct for the female shaking behaviour.

This content downloaded from 152.118.231.148 on Wed, 20 Mar 2019 03:12:46 UTC
All use subject to https://about.jstor.org/terms
 651 d 2
was wrongly predicted. This is a correct classification
Dungfly sexual ' a rate of 97*9% and should be contrasted with the
conflict ,, 0 *
random expectation of 41%, the percentage of cases
E Ad08
E~R0_8 .0_00 in the smaller group.
Copula duration was only related to the duration
E~.4 006
o0 of the conflict and the regression was positive
(b=032?008, t=4*04, P<0O001; model
02
0 0 F1 17=16*34; P<0.001) (Fig. 4).
cL 0-0 _ __
&00
0 10 20 30 40 50 60 70
No. of eggs laid

Fig. 2. The relationship between the number of eggs laid Discussion

and the female's subsequent shaking behaviour.
It is mainly female behaviour that decides the
outcome of the mating conflict in Sepsis cynipsea.
The proportion of the time the female spends
Forty-six pairs were observed, of which 19 went on shaking, behaviour clearly attempting to dislodge the
to copulate. Table 1 shows the means and standard male, was the main factor which differentiated pairs

deviations for pairs which did and did not copulate. which would go on to copulate from those which
The variables were used in a stepwise discriminant would not. The time the female had spent laying her
analysis to identify the combination which best eggs was the next factor in importance. These factors

discriminated pairs which copulated from those had much greater influence than the two male traits
which did not, i.e. copula or no copula was the aimed at attaining genital contact which entered into

grouping variable. The selection criterion for entry

into the model was the minimization of Wilks' X; i.e.
the variable which resulted in the largest difference
14i
between the group means was entered at each step.
2.
Four variables were entered into the model (Table E

2). The first two variables entered had marked effects k 10.
on Wilks' X for the model as a whole and were
E 0
features of the female's behaviour; the proportion of 6
time she spent shaking and the time she had spent
2 4-
laying. The next two were the male traits apparently
0

2
aimed at establishing genital contact; the number of
strokes and temporary genital contacts per minute,
but these had much smaller effects on the model. The -2-
I 1-05 1. 1-15 1-2 1-25 1-3 1-35 1-4
overall canonical correlation was 0*881, equivalent to
Female hind tibia length (mm)
the Pearson correlation between the discriminant
Fig. 3. The relationship between the female's hind tibia
score and the grouping variable. The classification
length and the number of strokes, which are almost
using these discriminant results contained only one
certainly attempts to attain genital contact, made by the
error. All the pairs which did copulate were success- male. The sizes of the circles reflect the degree of overlap
fully identified and one pair which did not copulate amongst points.

Table 1. The means and SD for the variables used in the discriminant analysis of pairs which did or did not copulate. n= 19
copulated and n=27 did not copulate

Pairs which Pairs which did

copulated not copulate

Variable x SD x SD

No. of eggs laid 46-05 14-54 26-15 15-76

Laying time (min) 16-16 4-96 9-78 5-21
Conflict duration (min) 18-68 17-02 7-78 11-09
Proportion of time female 0-06 0 09 0-65 0*25
shook (transformed)
Male strokes min- 0 49 0-57 1-00 2-60
Temporary genital contacts min- 0-85 0-73 0-27 0-36
No. of attacks min- 0-60 0-37 0-91 0-93
Female size (mm) 1-21 0-09 1-19 0-09
Male size (mm) 1-16 0-08 1-15 0*08
Size difference (mm) 0-05 0-10 0-05 0-09

This content downloaded from 152.118.231.148 on Wed, 20 Mar 2019 03:12:46 UTC
All use subject to https://about.jstor.org/terms
652 40- genital contact, per minute of the conflict the larger
P. I. Ward et al. 0 the female with which he was paired. This makes
.E3*E 30 -* evolutionary sense as larger females produce larger
o20 -~~~ ' batches of eggs than smaller ones (Ward 1983). A
al20- * *0 male will therefore probably fertilize more eggs in the
female's next batch the larger the female. A male's
size was not found to be significantly related to any of
C)

~.0 0O~ the traits investigated here. This is in contrast to the

subtle ways mate-searching behaviour is related to
0 10 20 30 4050 60
Conflict duration (min)
size (Ward 1983), which suggests sexual selection,
i.e. gaining access to mates, has a greater influence on
Fig. 4. The relationship between the conflict duration and male size than does selection after access has been
subsequent copula duration.
achieved.
The twisting by the male around the engaged
genitalia involved at end of copula (Parker 1972a)
the stepwise discriminant analysis. Thus, although suggests that the male may in some way be 'anchored'
females could be said to usually 'win' the conflict, as inside the female during copula. This does seem
expected (Parker 1979; Eberhard 1985), a persistent plausible when the structure of the aedeagus is
male can sometimes be successful in overcoming a considered. Pont (1979) describes the Sepsis aedea-
female's resistance. This makes Sepsis cynipsea an gus as 'complex and usually flamboyant', but this
interesting case in terms of Ridley's (1990) compara- does not do that of Sepsis cynipsea full justice. The
tive classification of insect mating systems. Females distal end, where the penis is, has two apparently
must encounter males if they are to lay their eggs, immovable chitinous spikes which may well be used
suggesting 'male control', but they largely determine to anchor the male inside the female. Under these
whether mating takes place, i.e. there is 'female spikes is what appears to be an inflatable bulb
control' of mating frequency. We suspect that more equipped with stout backward-pointing hairs. More
detailed field examinations will reveal more such proximally, there are two scoops similar in morpho-
examples of what may be called 'mixed control'. logy to the structures used by Odonata males to
The number of eggs the female laid was the best remove stored sperm from the females' sperma-
predictor of her subsequent shaking behaviour; the thecae (e.g. Waage 1979). Between these scoops is a
more eggs she laid, the less she shook. As the females tight-meshed mass of stout hairs which appear to be
use stored sperm to fertilize their eggs, this suggests moveable. The main part of the aedeagus is also
that the main reason females remate is to replenish covered in backward-pointing scales. This morpho-
their sperm stores. However, almost all females logy has all the hallmarks of structures used by males
to remove sperm from females. The morphology
shake a little, suggesting males must always be able to
overcome minimal resistance at least. This may be could simply function to anchor the male in the
tested by counting sperm in females after having female but the complexity and apparent mobility
observed their shaking behaviour. The males have suggests more complex functions.
modified forelegs with which they hold the female The mechanism of sperm competition in this
(Parker 1972a; Pont 1979) and although the males species is at present unknown. We expect, however,
must also have a good hold to resist the attacks of that it could explain why copula duration was related
unpaired males, it seems likely that the main function only to the conflict duration. One possibility is that
of the modified forelegs is to resist being dislodged by males are more persistent and then copulate for
the female, as suggested by Thornhill & Alcock longer the greater their sperm reserves. Sperm
(1983; see also Spieth 1952; Thornhill 1984). reserves have been shown to have an important
A male made more strokes with his abdomen, influence on copula duration in another dung fly,
which are almost certainly attempts to establish Scathophaga stercoraria (Ward & Simmons 1991).

Table 2. A summary table of the discriminant analysis results

Wilks'
Step Variable entered X Equivalent F df P

1 Proportion of time female 0-319 93-68 1,44 <0-0001

shook (transformed)
2 Laying time (min) 0-246 65-76 2,43 <0-0001
3 Male strokes min 0-230 46-79 3,42 <0-0001
4 Temporary genital contacts minx 0-223 35-63 4,41 <0.0001

This content downloaded from 152.118.231.148 on Wed, 20 Mar 2019 03:12:46 UTC
All use subject to https://about.jstor.org/terms
653 Acknowledgements Ridley, M. (1990) The control and frequency of mating in
insects. Functional Ecology 4, 75-84.
Dung fly sexual Thanks to R. Semlitsch, A. Porter and an anonymous
Sokal, R.R. & Rohlf, F.J. (1981) Biometry, 2nd edn. W.H.
conflict referee for comments on Freeman,
the San manuscript.
Fraficisco.
Spieth, H.T. (1952) Mating behaviour within the genus
Drosophila. Bulletin of the American Museum of Natural
References History 99, 395-474.
SPSS (1990) SPSS Reference Guide. SPSS, Chicago.
Eberhard, W.G. (1985) Sexual Selection and Animal Geni- Thornhill, R. (1984) Alternative hypotheses for traits
talia. Harvard University Press, Cambridge, Massa- believed to have evolved by sperm competition. Sperm
chusetts. Competition and the Evolution of Animal Mating Systems
Halliday, T. & Arnold, S.J. (1987) Multiple mating by (ed. R.L. Smith), pp. 151-178. Academic Press, New
females: a perspective from quantitative genetics. York.
Animal Behaviour 35, 939-941. Thornhill, R. & Alcock, J. (1983) The Evolution of Insect
Parker, G.A. (1972a) Reproductive behaviour of Sepsis Mating Systems. Harvard University Press, Cambridge,
cynipsea (L.) (Diptera: Sepsidae) I. A preliminary Massachusetts.
analysis of the reproductive strategy and its associated Waage, J.K. (1979) Dual function of the damselfly penis:
behaviour patterns. Behaviour 41, 172-206. sperm removal and transfer. Science 203, 916-918.
Parker, G.A. (1972b) Reproductive behaviour of Sepsis Ward, P.I. (1983) The effects of size on the mating
cynipsea (L.) (Diptera: Sepsidae) II. The significance of behaviour of the dung fly Sepsis cynipsea. Behavioral
the precopulatory passive phase and emigration. Behav- Ecology and Sociobiology 13, 75-80.
iour 41, 242-250. Ward, P.I. & Simmons, L. (1991) Copula duration and
Parker G.A. (1979) Sexual selection and sexual conflict. testes size in the yellow dung fly, Scathophaga stercoraria
Sexual Selection and Reproductive Competition in Insects (L.): the effects of diet, body size, and mating history.
(eds. M.S. Blum & N.A. Blum), pp. 123-166. Academic Behavioral Ecology and Sociobiology 29, 77-85.
Press, New York.
Pont, A.C. (1979) Sepsidae. Royal Entomological Society Received 13 January 1992; revised 26 May 1992; accepted I
of London, London. June 1992

This content downloaded from 152.118.231.148 on Wed, 20 Mar 2019 03:12:46 UTC
All use subject to https://about.jstor.org/terms