OPEN ACCESS Freely available online
The Cost of Sex: Quantifying Energetic Investment in
Gamete Production by Males and Females
April Hayward*, James F. Gillooly
Department of Biology, University of Florida, Gainesville, Florida, United States of America
The relative energetic investment in reproduction between the sexes forms the basis of sexual selection and life history
theories in evolutionary biology. It is often assumed that males invest considerably less in gametes than females, but
quantifying the energetic cost of gamete production in both sexes has remained a difficult challenge. For a broad diversity
of species (invertebrates, reptiles, amphibians, fishes, birds, and mammals), we compared the cost of gamete production
between the sexes in terms of the investment in gonad tissue and the rate of gamete biomass production. Investment in
gonad biomass was nearly proportional to body mass in both sexes, but gamete biomass production rate was
approximately two to four orders of magnitude higher in females. In both males and females, gamete biomass production
rate increased with organism mass as a power law, much like individual metabolic rate. This suggests that whole-organism
energetic may act as a primary constraint on gamete production among species. Residual variation in sperm production
rate was positively correlated with relative testes size. Together, these results suggest that understanding the heterogeneity
in rates of gamete production among species requires joint consideration of the effects of gonad mass and metabolism.
Citation: Hayward A, Gillooly JF (2011) The Cost of Sex: Quantifying Energetic Investment in Gamete Production by Males and Females. PLoS ONE 6(1): el6557.
Editor: Christian Voolstra, King Abdullah University of Science and Technology, Saudi Arabia
Received September 17, 2010; Accepted January 4, 2011; Published January 24, 2011
Copyright: � 2011 Hayward, Gillooly. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded by the University of Florida. The founder had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: email@example.com
Much of sexual selection and life history theory is based on
assumptions about the relative investment of males and females in
sexual reproduction. The predominance of female choice in sexual
selection, and parental care by females, is typically attributed to
the greater investment by females per ,tTi.lini [1,2]. In many
species, eggs are much larger than sperm (i.e. anisogamy) and
thus sperm are often considered relatively "cheap" to produce
[2,3,4,5]. However, quantifying the energetic cost of gamete
production both within and between the sexes has remained a
difficult challenge since it includes both the investment in gonads
and the rate of production of gamete biomass [4,6,7,8]. A better
understanding of these costs is important for understanding how
differences in animal mating systems, and the evolutionary forces
that shape them, are related to whole-organism physiology.
Efforts to understand the considerable heterogeneity in gamete
production within and between the sexes has generally taken place
in the context of life history theory. Among males, differences
among species in the size of gonads and/or sperm, and rates of
sperm production, are typically attributed to the intensity of
postcopulatory sexual selection in the form of sperm competition:
Males experiencing greater sperm competition are expected to
invest relatively more biomass in gonads and produce sperm at a
relatively higher rate [3,7,8,9,10]. In females, no similar theory has
been proposed to explain differences in the energy expended for
gonad biomass and gamete production, but the expectation from
life history theory is that females produce the optimal size and
number of gametes at a rate that maximizes lifetime reproductive
. ' PLoS ONE | www.plosone.org
success [2,3,11]. Between the sexes, the amount of energy invested
in gametes by males and females is often assumed to differ, since
females aim to maximize tTi'l int survival, whereas males aim to
inseminate as many females as possible [2,3]. However, little
consideration has been given to energetic limitations on the
production of gamete biomass that may be imposed li.._ 1
constraints on whole-organism metabolism. Moreover, broad-scale
interspecific comparisons of the energetic investment in gametes
are rare for females and almost non-existent for males (but see
Here we present a broad-scale comparative study that quantifies
two key features of energetic investment in gametes by males and
females for diverse species (i.e. invertebrates, reptiles, amphibians,
fishes, birds, and mammals) that vary tremendously in their life
histories. First, we compare the biomass allocation to gonads in
males and females across a broad range of body sizes and assess
differences in allocation among taxonomic groups. Like other
organs, we expect gonad mass to scale approximately linearly with
body mass [13,14] and for variation about the relationship to be
explained by differences in sperm competition among males and
differences in clutch size among females. Second, we compare
rates of production of gamete biomass in males and females across
a broad range of body masses. We hypothesize that, like other
rates of biomass production (e.g. ._.. d1. rate, [15,16]), the
production of gamete biomass should occur at a rate proportional
to whole-organism metabolic rate. This presumes that the
production of gamete biomass is a function of both gonad mass
and gonad metabolic rate and that gonad metabolic rate is
proportional to whole-organism metabolic rate. Thus, we expect
January 2011 1 Volume 6 1 Issue 1 | e16557
The Cost of Sex
gamete biomass production rates to scale as a power law with body
mass with an exponent of about /4, as is often observed for whole-
organism metabolic rate [16,17,18,19], but see [20,21,22]. Since
males must produce seminal fluid in addition to sperm, we also
assess male investment in ejaculate biomass production (i.e.
gametes + seminal fluid). We then quantify the amount of energy
devoted to egg, sperm, and ejaculate biomass production relative
to basal metabolic rate. Finally, we consider whether residual
sperm or egg biomass production rates are related to residual
gonad mass. In the case of sperm, a positive relationship between
residual sperm biomass production rates and residual testes mass
would be consistent with sperm competition theory.
Results and Discussion
Investment in Gonads
Across all species, gonad mass I scaled somewhat less than
linearly with soma mass \I for both males and females (Fig. 1;
logMg=-2.16+0.89 logMi, 95%CIb: 0.86-0.91, r =0.89,
p<0.0001), with no significant difference in either the slope or
intercept between the sexes i .......... ,.; of slopes ANCOVA:
F3,639 = 1768.50; Fi = 4670.07, Fex = 0.04, FMi*sex = 0.64). Thus,
body mass alone explained 89% of the variation in gonad mass
across all taxonomic groups. Including taxonomic group as a
covariate increased the amount of variation explained to 91% and
revealed statistically significant differences in slopes and intercepts
among groups (Table 1). However, in most groups, the slope of the
gonad mass-body mass relationship did not differ _;. ,1
from 1, 1 .... 1. amphibians showed a 1; i11 steeper relationship
(b=1.27, 95%CIb: 1.08-1.44) and birds showed a 1;_i1l
shallower relationship (b=0.71, '. .1 1, 0.59-0.83). Thus, the
scaling of gonad mass with body mass was similar across groups
and between the sexes, ,.1 ii. i;_ I ,, .. . . and females of diverse
species invest similarly in reproductive tissue biomass. This scaling
is quite similar to scaling relationships previously observed for
other organs, with the possible exception of the brain [13,14]. This
suggests that any previously hypothesized effect of sperm
competition on allocation to gonad mass, which would be reflected
in the variance about the gonad-soma relationship, is of secondary
importance relative to the constraints imposed by whole-organism
log somatic mass (g)
Figure 1. Relationships between gonad and soma mass. The
logarithm of gonad mass (g) versus the logarithm of soma mass (g) for
males (diamonds) and females (circles).
. ' PLoS ONE I www.plosone.org
Table 1. Gonad mass vs. somatic mass by taxon.
effect n intercept (95%CL)
-2.12 (-2.30 - -1.94)
0.41 (0.18 - -0.63)
0.50 (0.18 - 0.83)
0.21 (-1.02 - 0.60)
-0.52 (-0.77 - -0.27)
-0.25 (-0.48 - -0.02)
0.76 (0.82 - 1.06)
0.71 (0.59 - 0.83)
1.27 (1.08 - 1.44)
0.94 (0.82 - 1.06)
0.93 (0.86 - 1.00)
Separate slopes ANCOVA statistics for the relationship between the logarithm
of gonad (g) and soma mass (g) by taxon. Significant effects in bold; whole
model: r =0.91, p<0.0001, F11,631 =559.82; Ftaxon= 17.78, Fi= 156.97.
Gamete Biomass Production Rates
Across all species, rates of gamete biomass production in males
and females scaled similarly to each other and sub-linearly with
body mass (Fig. 2A), indicating that mass-specific rates of gamete
biomass production are greater in smaller-bodied species.
Specifically, sperm biomass production rates (W) scaled with body
mass raised to the 0.66 power and egg biomass production rates
(W) scaled to the 0.80 power (separate slopes ANCOVA:
F 3,119= 277.35, r2 = 0.87, p<0.0001; log(sperm biomass produc-
tion) =-6.09+0.66 1._1....1 mass (95%CIb: 0.55-0.77);
log(egg biomass production) =-2.66+0.80 1. I.....1 mass
i'. .1 1, 0.7311 I As hypothesized, the slopes of these
relationships did not differ _i,;;. i,1 from the '/4-power scaling
of metabolic rate with body mass. Thus, like other rates of biomass
production, a roughly constant fraction of a species' metabolism is
devoted to the production of gamete biomass. Notably, these
results indicate that, on a mass-specific basis, the biomass allocated
to sperm or egg production per lifetime is approximately invariant.
In other words, since mass-specific gamete biomass production
scaled approximately as M " and lifespan scales approximately as
M", lifetime gamete biomass production scales as Mo. Thus, on
average, species invest about the same fraction of their lifetime
energy budget in the production of gamete biomass, regardless of
the specific life-history strategy they might employ to enhance
In terms of the energetic cost of gamete biomass production,
differences between males and females were substantial. Specifi-
cally, our analysis indicates that males and females devoted about
0.1 and 300% of the energy used for basal metabolism to the
production of gamete biomass respectively. Thus, the cost of egg
production was roughly 3.5 orders of magnitude higher than the
cost of sperm production. This difference could not be attributed
to males' investment in ejaculate production. Total ejaculate
biomass production (W) scaled with body mass as log(ejaculate
biomass production) =-5.52+0.75 1... ....1 mass) (95%CIb:
0.60-0.90, r2 =0.75, p<0.0001), such that the cost of ejaculate
production (i.e. gametes + seminal fluid) constituted only 0.4% of
basal metabolism. Thus, while the cost of ejaculate production was
approximately 4-fold higher than that of sperm production alone,
egg production rates remained nearly 3 orders of magnitude
higher than ejaculate production rates. Still, like rates of gamete
biomass production, ejaculate biomass production rates scaled to
the /4-power of body mass and were therefore proportional to
For both males and females, our estimates of the energetic cost
of gamete biomass production across species are consistent with
January 2011 1 Volume 6 1 Issue 1 | e16557
The Cost of Sex
-8 -4 0 4
log body mass (g)
-2 -1 0 1 2
residual gamete mass
Figure 2. Relationships between gamete biomass production
rates, body size, and gonad mass. (A) The logarithm of
temperature-corrected daily sperm biomass production rates (W;
diamonds, dashed black line) and the logarithm of daily egg biomass
production rate (W; circles, solid line) versus the logarithm of body mass
(g). The relationship between metabolism and body mass for
ectotherms at 20�C  is plotted for comparative purposes (dashed
orange line). (B) Residual daily gamete biomass production rates (from
a log-log plot of daily gamete biomass production rates versus body
mass) versus residual gonad mass (from a log-log plot of gonad mass
versus body mass) males (diamonds, dashed line) and females (circles).
the limited amount of previous work on individual species or
species groups. For males, our estimate of the cost of ejaculate
production (0.4% of BMR) is similar to that previously reported
for Japanese macaques (0.8-6%; ). For females, our estimate
of the energetic cost of egg production (-300% of BMR) is in close
agreement with the range of estimates (-20% to -200%) that
have been presented for various species of birds [24,25,26]. If we
assume that eggs are produced 1.i. 1.....i the year, rather than
only ,1 ;,_ il,. breeding season, our estimate of the energetic cost
of egg biomass production decreases to about 50% of BMR. Our
findings for females also compliment those of a recent study that
found that mass-specific reproductive biomass production rates
scaled as body mass to the power of -0.37 in mammals, when
reproductive biomass was measured in terms of the mass of newly
weaned young .
While body mass explained much of the variation in rates of
gamete biomass production in both sexes (63 and 94% in males
and females, respectively), residual testes mass (from a log-log
regression of testes mass on body mass) explained a small amount
. ' PLoS ONE | www.plosone.org
of the residual variation in gamete biomass production rates in
males i-. I. ;.1,, . sperm production) = -0.09+1.23 log(residual
testes mass 1_ '. . i, 0.56-1.89, r = 0.24, p<0.001; Fig. 2B).
These results might be viewed as consistent with predictions from
sperm competition theory, 1 ... 1, the relationship was weak and
highly variable among groups. Based on limited data, a separate
slopes ANCOVA indicated that birds and mammals had
statistically significant, positive relationships, whereas fishes and
invertebrates did not show statistically significant relationships
(F7,37 = 11.78, r = 0.63, p<0.0001; birds: b = 1.98 (95%CI: 1.06
S'* p<0.001, n= 14; mammals: b= 1.02 (0.28-1.75 95%CI),
p<0.0001, n=21; fishes: n=5; invertebrates: n=5). Note,
however, that residual egg biomass production rates were not
related to residual ovary mass (Fig. 2B).
Our results provide insights regarding the investment by males
and females in gonad and gamete biomass. Both within and
between the sexes, investment in gonad biomass was quite similar
across species. This is reflected in the similarity in both the slopes
and intercepts of the scaling relationships of testes and ovary mass
with body mass. With respect to gamete biomass production, the
story appears to be quite different. Within each sex, the production
of gamete biomass scaled sub-linearly with body mass across
species in about the same way as whole-organism metabolic rate.
However, between the sexes, rates of gamete biomass production
were two to four orders of magnitude higher in females. This
presents an interesting question for future research as it suggests
that mass-specific rates of gamete biomass production, and
perhaps mass-specific rates of metabolism in general, were much
higher in ovaries than testes.
More generally, the results presented here raise questions
regarding hypotheses aimed at explaining differences in repro-
ductive investment between males and females based on sexual
selection. For example, sperm competition theory and models of
parental investment generally assume that gamete biomass
production is proportional to gonad mass . However, our
results indicate that rates of production of gamete biomass are
roughly proportional to whole-organism metabolic rate and that
gonad mass is of secondary importance. As such, these results
point to the need to integrate theory on sexual selection and life
history with fundamental principles of organism-level physiology
when addressing questions related to costs of sexual reproduction.
Qn ..ii ;,_ 11,. energetic cost of gamete production and relating it
to individual metabolic rate, as we have done here, may ultimately
be helpful in quantifying tradeoffs in parental investment given the
finite energy budget of individuals.
Materials and Methods
Gonad and somatic tissue mass data were obtained from the
literature for 656 species (Table Si). Estimates of daily rates of egg
(72 species) and sperm (51 species) biomass production during the
breeding season (g/d) were either obtained directly from the
literature, or calculated by multiplying estimates of gamete mass
by published estimates of the number of gametes produced per day
(Table S2). When production rates were presented as annual rates
of production, they were converted into daily rates of production
by dividing by breeding season 1.. , 11. (d/yr), as obtained from the
literature (see Table S2). Gamete mass was estimated from gamete
volume by assuming biomass density was equivalent to that of
water (1 g/mL; ). When not reported directly (in many cases),
total sperm volume was estimated using the volume of each of the
three primary sperm features (i.e. head, midpiece, and tail; J. F.
January 2011 1 Volume 6 1 Issue 1 | e16557
The Cost of Sex
Gillooly, H. B. Vander Zanden, & A. Hayward, unpublished
data). The volume of a given sperm feature was determined using
linear dimensions and the approximate shape of each feature as
reported or inferred from the literature. When shape was not
reported, the equation for the volume of an ellipsoid or spheroid
was used to calculate head and midpiece volume (depending on
the number of dimensions available) and the equation for a
cylinder was used to calculate tail volume. When insufficient data
were available to calculate sperm volume for a given species,
volume was either approximated using sperm volumes) from
closely related taxa or estimated using observed allometric
relationships between total volume and head volume for mammals
(log total volume = 0.56+0.74 * log head volume, r = 0.80,
p<0.0001) or total volume and midpiece volume for birds (log
total sperm volume = 1.07+0.23 * midpiece volume, r = 0.45,
p<0.05) (see Table S2). Daily ejaculate production rates were
estimated from previously published data using ejaculate volume
and the frequency of ejaculation or using the concentration of
sperm per ejaculate and the number sperm produced per day (see
Gamete mass expressed in grams of carbon was converted to
dry weight (g) using a conversion factor of 0.4. Dry weight was
converted to wet weight (g) using a conversion factor of 0.25 g dry
weight/g wet weight . Biomass production data were
converted from g/d to J/d using a conversion factor of
7.0x106J/g and then into Watts by dividing by 86400 s/d in
order to compare the amount of energy devoted to gamete
biomass production with basal metabolic rate of ectotherms at
20 C (metabolic rate (W)=0.14 * mass I._- ' . Gamete
biomass production was assumed to have occurred at ambient
environmental temperature for ectotherms and at 38'C and 40�C
for mammals and birds, respectively. Rates of production were
corrected to 20'C by assuming production increased exponentially
with temperature using the Boltzmann-Arrhenius factor (e-E/kT),
where E is the .i 1 .i activation energy of the respiratory
complex (-1.04xlU -JJ (0.65 eV k is Boltzmann's constant
(1.381 10-23 JK-1 (8.62x10-5 eV*K-' and T is absolute
1. Kokko H, Jennions M (2003) It takes two to tango. Trends in Ecology &
Evolution 18: 103-104.
2. Trivers R (1972) Parental investment and sexual selection. In: Campbell B, ed.
Sexual Selection and the Descent o fMan 1871-1971: Aldine Press. pp 139-179.
3. Parker GA (1970) Sperm competition and its evolutionary consequences in the
insects. Biological Reviews 45: 525-567.
4. Parker GA (1982) Why are there so many tiny sperm? Sperm competition and
the maintenance of two sexes. Journal of Theoretical Biology 96: 281-294.
5. Bateman AJ (1948) Intra-sexual selection in Drosophila. Heredity 2: 349-368.
6. Parker GA (1993) Sperm competition games: sperm size and sperm number
under adult control. Proceedings of the Royal Society of London Series B:
Biological Sciences 253: 245-254.
7. Moller AP (1991) Sperm competition, sperm depletion, paternal care, and
relative testis size in birds. The American Naturalist 137: 882 906.
8. Ginsberg JR, Huck UW (1989) Sperm competition in mammals. Trends in
Ecology & Evolution 4: 74 79.
9. Moller AP (1989) Ejaculate quality, testes size and sperm production in
mammals. Functional Ecology 3: 91-96.
10. Hosken DJ, Ward PI (2001) Experimental evidence for testis size evolution via
sperm competition. Ecology Letters 4: 1013.
11. Smith CC, Fretwell SD (1974) The optimal balance between size and number of
offspring. The American Naturalist 108: 499-506.
12. Stockley P, Gage M, Parker G, Moller A (1997) Sperm competition in fishes: the
evolution of testis size and ejaculate characteristics. The American Naturalist
13. Peters RH (1983) The ecological implications of body size. Cambridge:
Cambridge University Press.
14. Calder WA (1984) Size, function and life history. Harvard: Harvard University
15. Ernest SKM, Enquist BJ, Brown JH, Charnov EL, Gillooly JF, et al. (2003)
Thermodynamic and metabolic effects on the scaling of production and
population energy use. Ecology Letters 6: 990-995.
temperature in Kelvin . This is equivalent to assuming a Qo
of about 2.5.
Statistical analyses were carried out in Statistica 8.0. Homoge-
neity of slopes analysis was used to determine whether the
relationship between gonad and soma mass differed among sexes
or groups. Separate slopes ANCOVA was used to compare the
scaling of gonad mass with soma mass among groups, and to
evaluate differences in the scaling of gamete production rates and
body mass between the sexes. Bivariate OLS regressions were used
to evaluate the relationship between ejaculate biomass production
rate and body mass and between residual gamete biomass
production rates and residual testes mass. Differences in the
scaling of residual sperm biomass production rates with residual
testes mass among groups were evaluated using a separate slopes
We wish to thank Jennifer Parker and Hannah B. Vander
Zanden for their efforts in compiling and collating portions of the
data presented here.
Table S1 Gonad and soma mass data. Raw data and
sources for gonad and soma mass data used in this study.
Table S2 Gamete biomass production rate data. Raw
data and sources for egg and sperm biomass production rate data
used in this study.
Conceived and designed the experiments: AH JFG. Performed the
experiments: AH JFG. Analyzed the data: AH JFG. Contributed
reagents/materials/analysis tools: AH JFG. Wrote the paper: AH JFG.
16. Savage VM, GilloolyJF, Woodruff WH, West GB, Allen AP, et al. (2004) The
predominance of quarter-power scaling in biology. Functional Ecology 18:
17. GilloolyJF, BrownJH, West GB, Savage VM, Charnov EL (2001) Effects of size
and temperature on metabolic rate. Science 293: 2248-2251.
18. Kleiber M (1932) Body size and metabolism. Hilgardia 6: 315-351.
19. Hemmingsen AM (1960) Energy metabolism as related to body size and
respiratory surfaces, and its evolution. Reports of the Steno Memorial Hospital
Copenhagen 9: 1110.
20. Bokma F (2004) Evidence against universal metabolic allometry. Functional
Ecology 18: 184-187.
21. Dodds PS, Rothman DH, WeitzJS (2001) Re-examination of the "3/4-law" of
Metabolism. Journal of Theoretical Biology 209: 927.
22. McNab BK (1988) Complications inherent in scaling the basal rate of
metabolism in mammals. The Quarterly Review of Biology 63: 25-54.
23. Thomsen R, SoltisJ, Matsubara M, Matsubayashi K, Onuma M, et al. (2006)
How costly are ejaculates forJapanese macaques? Primates 47: 272274.
24. Salvante KG, Vezina F, Williams TD (2010) Evidence for within-individual
energy reallocation in cold-challenged, egg-producing birds. Journal of
Experimental Biology 213: 1991-2000.
25. Vezina F, Williams TD (2005) The metabolic cost of egg production is
repeatable. Journal of Experimental Biology 208: 25332538.
26. KingJR (1973) Energetics of reproduction in birds. In: Farner DS, ed. Breeding
biology of birds. Washington, DC: National Academy of Sciences. pp 78-107.
27. Hamilton MJ, Davidson AD, Sibly RM, BrownJH (2010) Universal scaling of
production rates across mammalian lineages. Proceedings of the Royal Society
B: Biological Sciences.
28. Allen AP, GilloolyJF (2007) The mechanistic basis of the metabolic theory of
ecology. Oikos 116: 1073-1077.
January 2011 1 Volume 6 1 Issue 1 | e16557
,., PLoS ONE I www.plosone.org