MARINE ECOLOGY PROGRESS SERIES
Mar Ecol Prog Ser
Vol. 293: 89?97, 2005 Published June 2
INTRODUCTION
Sex change is widespread among vertebrates and
invertebrates. Most species show environmentally
mediated lability in size or age at sex change. How-
ever, the number of factors influencing sex change
make it difficult to make precise predictions of the size
at which an individual changes sex. Interactions with
conspecifics have been shown to influence the initia-
tion or delay of sex change in sex-changing fishes (e.g.
Robertson 1972, Alonzo & Warner 2000a,b, Mackie
2003, Perry & Grober 2003), shrimp (see e.g. Charnov
1982, Koeller et al. 2000, Chiba et al. 2003), and mol-
lusks (e.g. Coe 1953, Hoagland 1978, Charnov 1982,
Soong & Chen 1991, 2003). Studies of protogynous
(females first) fish species have revealed a surprising
diversity of strategies to increase reproductive output,
including alternate mating strategies for males, accel-
erated and delayed sex change, and reversed sex
change (e.g. Warner & Swearer 1991, St. Mary 1993,
Lutnesky, 1994, Munday 2002). Some fishes can make
very fine adjustments in sex allocation in response to
local social interactions (e.g. Ross et al. 1983, Petersen
et al. 2001, Alonzo & Warner 2000b).
In contrast to the wide range of behavioral strategies
displayed by sex-changing fishes, most sex-changing
mollusks seem to employ simple strategies for sex
change (Warner et al. 1996). This simplicity may be
due to differences between the evolution of reproduc-
tive strategies in protandrous (males first) and protog-
ynous species, differences in the complexity of verte-
brate versus invertebrate behaviors, or may reflect the
relatively large research effort focused on fishes com-
pared to invertebrates. Strategies of sex change in
? Inter-Research 2005 ? www.int-res.com*Email: collinr@naos.si.edu
Effects of conspecific associations on size at sex
change in three species of calyptraeid gastropods
Rachel Collin1,*, Michelle McLellan1, Karl Gruber2, Catherine Bailey-Jourdain1
1Smithsonian Tropical Research Institute, Apartado Postal, 0843-03092, Balboa, Ancon, Panam?
2Bell Museum of Natural History and Department of Ecology, Evolution, and Behavior, University of Minnesota, 
1987 Upper Buford Circle, St. Paul, Minnesota 55108-6097, USA
ABSTRACT: In marine molluscs, sex change is often labile and is thought to be largely influenced by
interactions with conspecifics. Previous studies of calyptraeid gastropods concluded that their social
environment influences the timing of protandrous sex change. We conducted field surveys and labo-
ratory experiments to examine the effects of conspecifics on size at sex change in 3 Panamanian
calyptraeids. Crepidula cf. onyx, C. incurva Broderip, 1834 and Bostrycapulus calyptraeformis (Des-
hayes, 1830) vary in densities, sex ratio and mode of development, which suggests that they might
respond to associations with conspecifics in different ways. However, our laboratory experiments
showed that the response to interactions with conspecifics is generally similar. In all 3 species, indi-
viduals raised in isolation pass through a male phase and males raised alone change sex at the same
size as males raised with another male. Both C. cf. onyx and C. incurva change sex at a larger size
when kept with a female than when kept alone or with another male. The differences in size at sex
change between the treatments is small and the treatment effect explains more of the variation in size
at sex change in C. cf. onyx, the more solitary species, than in C. incurva, a species that is usually
found in pairs. In all species, individuals with high initial growth rates change sex sooner and at a
smaller size than those with slower initial growth rates. Growth rates increase significantly during sex
change in C. cf. onyx and C. incurva but not in B. calyptraeformis.
KEY WORDS:  Protandry ? Crepidula spp. ? Size advantage ? Bostrycapulus calyptreaformis
Resale or republication not permitted without written consent of the publisher
Mar Ecol Prog Ser 293: 89?97, 2005
protandrous shrimp, which have been more inten-
sively studied than mollusks, appear to be more vari-
able, with mixed strategies as well as pure protandry
occurring in several groups (e.g. Bauer & VanHoy
1996, Baldwin & Bauer 2003). 
It is difficult to assess the variability of sex change
strategies among mollusks because few groups have
been examined in detail and the conclusions of studies
on the same species often differ. Calyptraeids, protan-
drous, filter-feeding limpets, are the most well-studied
sex-changing mollusks (see e.g. Gould 1917, 1919,
1952, Coe 1936, 1938, 1953, Hoagland 1978, Collin
1995, Warner et al. 1996, Chen & Soong 2000). The
conclusions of these studies are often contradictory,
and interpretation of many older studies is difficult
because they may have employed experimental condi-
tions that starved the limpets (e.g. Coe?s [1948] criti-
cisms of Gould [1947], but see Gould & Hsiao's rebuttal
[1948]), the sample sizes were often low or not re-
ported (e.g. only a single replicate of each density
treatment in Hoagland 1978), the experimental design
was not described (e.g. Coe 1938), or/and the size of
the limpets in the different treatments were not re-
ported (e.g. Coe 1953). Differing results from studies
using the same species are common. For example, Coe
(1936) stated that there are some ?true? males that
never change, but Hoagland (1978), studying the same
species, found no evidence to support this in C. con-
vexa and thought it unlikely in Crepidula fornicata.
Although Gould (1919) reports that small C. plana
must be associated with a female to develop into a fully
functional male, Coe (1936, 1938) stated that this asso-
ciation is not necessary for male development in either
C. plana or 6 other species of Crepidula. Some of these
conflicting conclusions could be due to differences or
vagaries in interpretation of similar results. For exam-
ple, several of these studies report a ?delayed? sex
change, which could refer to a delay with respect to
size or to a temporal delay (a distinction that is often
not made clear).
Sex change strategies may vary in concert with
mode of development, social behavior, and body size.
Coe (1938) suggested that larger species with plank-
tonic development (Crepidula fornicata, C. onyx and C.
plana) are gregarious and show large differences in
size at sex change between solitary and ?mated? indi-
viduals while smaller species with direct development
(C. convexa and C. adunca) are less gregarious and do
not show a difference in size at sex change between
solitary and mated individuals. The results of Hoag-
land (1978) for C. fornicata, C. plana and C. convexa, a
subset of the species examined by Coe (1938), support
the latter?s conclusions. However, C. nivea, Crucibu-
lum spinosum, and Crepidula norissiarum do not fit the
pattern (Coe 1938, Warner et al. 1996) and there is no
theoretical expectation that body size or mode of de-
velopment should be associated with strategies of sex
change.
We examined the lability of sex change in calyp-
traeid gastropods with the aim of understanding varia-
tion in sex allocation strategies among a wider range of
species and to obtain parameters for subsequent use in
quantitative models of sex change. We address the
following questions for 3 species (Crepidula cf. onyx,
C. incurva and Bostrycapulus calyptraeformis) that
display different life histories (Table 1) and gregarious-
ness in the field: (1) Do individuals raised alone bypass
the male stage and develop directly into females? (2)
Do individuals kept alone, or in association with males
or, in association with females change sex at different
sizes? (3) Does growth rate increase during sex
change? 
MATERIALS AND METHODS
Study species and field survey. The 3 calyptraeid
species Crepidula cf. onyx, C. incurva Broderip, 1834,
and Bostrycapulus calyptraeformis (Deshayes, 1830)
examined in this study co-occur commonly in
muddy?rocky habitats in the intertidal along the Pacific
coast of Panama. We used field-collected samples to
survey the natural size frequencies of males and fe-
males for each species from Playa Venado and Playa
Chumical (8? 30? N, 79? 40? W). The method of sampling
differed for each species, based on their natural history. 
Crepidula cf. onyx occurs at Playa Venado in densi-
ties of 1 or 2 individuals m?2 in the low intertidal (?0.5
90
Table 1. Crepidula cf. onyx, C. incurva and Bostrycapulus calyptraeformis. Life history differences between the 3 panamanian
calyptraeids. Sex ratio: breeding ratio males:males + females
Species Development Stacking Max. size (mm) Abundance Sex ratio
C. cf. onyx Direct Solitary 34 Uncommon 0.21
C. incurva Planktotrophic 53% in pairs 18 Patchy 0.40
B. calyptraeformis Planktotrophic Solitary or paired 25 Very abundant 0.55
(up to 1000 m?2)
Collin et al.: Calyptraeid sex change
to ?1.0 m). We collected them by carefully examining
the rocks within an area of about 200 m2 and collecting
all the C. cf. onyx during low tide on February 2, 2003.
Crepidula incurva occurs patchily in the intertidal
between ?0.5 and ?1.0 m. They are usually found
attached to the shells of other snails, especially Ceri-
thium sp., which congregate in large numbers in shal-
low tide pools and below large rocks. We collected C.
incurva from tide pools within an area of 100 m2 by
examining every snail in the pool and collecting all
that hosted a Crepidula incurva. All individuals were
collected from January 19 to 23, 2003.
Because Bostrycapulus calyptraeformis can reach
densities of more than 1000 m?2 in the low intertidal,
we surveyed this species using 0.25 m quadrats: 3
quadrats were placed haphazardly within 200 m of
each other in the intertidal where B. calyptraeformis
are known to occur, and all individuals were collected
from each quadrat from September 7 to 9, 2002.
Quadrat sampling ensured that we obtained a repre-
sentative sample including all the small individuals,
which are cryptic. 
Determination of sex. The shell lengths of all gas-
tropods were measured with calipers to the nearest
0.1 mm and their sex was determined by presence or
absence of the penis and the female genital papilla
(fgp). For both Crepidula cf. onyx and C. incurva this
was performed on living individuals which had been
removed from the substrate. Individuals with both a
penis and an fgp or with neither were classified as
transitional if they had previously been observed as
males, and as juveniles if they were small and had not
been observed with male morphology. 
Because female Bostrycapulus calyptraeformis do not
possess an fgp, the sex of field samples was deter-
mined on the basis of the presence or absence of a
penis and the presence or absence of a capsule gland
(part of the female reproductive tract which can be
seen when the gastropod is removed from the shell). In
544 field-collected individuals there were 5 individu-
als, presumably transitionals, for which the presence of
a penis was not a perfect predictor of the absence of
the capsule gland. Therefore, we used the presence or
absence of the penis as an adequate indicator of sex for
experimental individuals.
Isolation experiments. Previous literature on sex
change in Crepidula species has suggested that juve-
niles raised in isolation bypass the male stage (Gould
1919). In order to test this hypothesis, juveniles of each
species were raised in isolation. Larvae of C. incurva
and Bostrycapulus calyptraeformis were raised follow-
ing the methods of Collin (2000). After settlement they
were raised in groups until they reached 2.0 to 2.5 mm
and could be easily handled. Juvenile C. cf. onyx were
held in groups for several days after hatching, until
they also reached a length of 2.0 to 2.5 mm. Juveniles
of each species were transferred to individual 355 ml
cups which were maintained in the laboratory at room
temperature. They were fed 10 ml of a mix of algal
cultures (Nannochloropsis sp., Rhodomonas sp., Duna-
liella tertiolecta, Isochrysis galbana and Chaetocerus
gracilis) daily; the proportion of each alga in the mix
varied from day to day, but all gastropods were raised
simultaneously and all were fed from the same mix on
each day. The water was replaced with fresh unfiltered
seawater every 2 d. Initially this diet was satiating;
after several months, however, the gastropods were
large enough to clear the containers between feedings
and therefore food probably became limiting. The size
and sex of each individual was recorded every 3 wk
until female characteristics were observed. During the
first few weeks there was some (10 to 15%) mortality
when the tiny juveniles crawled out of the water and
dried out. However, after reaching a size of 5 mm,
mortality was negligible.
Sex change experiments. To examine the effects of
association with conspecifics on size at sex change we
raised individuals of each species in 3 different social
environments. For both Crepidula incurva and Bostry-
capulus calyptraeformis we collected individuals from
the same locations as the field surveys. Males of sizes
below the size of sex change (C. cf. onyx <2.5 mm;
C. incurva <8.5 mm; B. calyptraeformis <10.7 mm)
were randomly assigned to a 355 ml cup with one of
the following treatments (1) alone, (2) 2 males together,
or (3) male with a female. Because male C. cf. onyx
were rare in the field, we arranged the same 3 treat-
ments starting with laboratory-hatched juveniles and
field-collected females. The average initial difference
in size between the 2 males in the paired-male
treatment were 0.38, 1.4, and 2.47 mm for C. cf. onyx,
C. incurva and B. calyptraeformis respectively. There
were initially 43 to 50 replicates of each treatment for
each species. All treatments were fed 10 ml of a mix of
algal cultures daily, which was satiating for small indi-
viduals. After the gastropods were large enough to
clear the containers between feedings, we increased
the ration to 20 ml algae d?1 for solitary individuals and
40 ml algae d?1 for paired individuals. We recorded the
size and sex of each individual every 3 wk for the first
4 mo of the experiment and every 4 wk thereafter. The
experiment was continued for more than 1 yr. After
reaching a size of 4 to 5 mm, mortality was negligible,
the gastropods grew considerably, and most female-
phase individuals that had the chance to mate pro-
duced eggs.
Analyses. Size at sex change in the field was calcu-
lated using logistic regression (standard method de-
scribed by Sokal & Rohlf 1995) to predict the size at
which 50% of the individuals were male. We follow the
91
Mar Ecol Prog Ser 293: 89?97, 2005
convention of Allsop & West (2003) in using ?L50? to rep-
resent to the size at which 50% of the individuals were
male.
The effect of each treatment on size at sex change
was analyzed using 1-way ANOVA with 3 treatments
and Fisher?s post-hoc least-significant difference (PLSD)
tests, in which the treatments were (1) solitary individ-
uals, (2) first male to change sex in a pair of males, (3)
male paired with a female. The second male to change
sex in the paired male treatment was not included,
since as soon as the first male changed sex, the treat-
ment changed from being a male paired with a male to
being a male paired with a female. This means that
only the first male to change sex experienced the con-
sistent treatment of being paired with a male. These
analyses were conducted using the largest size that
each individual was observed to be male and repeated
using the smallest size it was observed to be female.
Analyses were not conducted on the size of transition-
als because many individuals that changed sex were
not observed in the transitional phase and the transi-
tional phase could not be identified in Bostrycapulus
calyptraeformis without sacrificing them.
The time until sex change in each treatment was
examined using survival analysis. We used Kaplan-
Meier survival statistics with sex change recorded as
the event; those individuals that died or had not
changed sex by the end of the experiment were
treated as censored.
Quantitative predictions of sex change and size at
sex change often depend on the assumption that
growth rates are the same for males, females and tran-
sitionals. Studies of Crepidula spp. often state that
growth increases during sex change, although it is sel-
dom measured explicitly. To test the hypothesis that
growth rates increase during sex change we quantified
growth rates for the solitary individuals. Individuals in
the other treatments were not used, as food availability
may have been affected by competition with the other
individual in the same container. The growth rates
were binned into each time period between measure-
ments and significant differences between the growth
rates before, during and after the change were exam-
ined using paired t-tests, and ANCOVA using shell
length as a covariate.
To determine if initial growth rates affect the timing
of sex change, we divided the gastropods into those
that changed sex before or after the census date by
which about half of the individuals in the treatment
had changed. Logistic regression was used to test for
association between growth rates and early sex
change, with initial size as a covariate.
RESULTS
Despite pronounced differences in field sex ratio,
size distribution and population densities, the 3 species
responded to the experimental conditions in similar
ways. Juveniles raised in isolation passed through a
male phase and males raised in isolation changed sex
at the same size as a male raised with another male.
Both Crepidula cf. onyx and C. incurva showed sex
change at a larger size when males were paired with a
female. Individuals with high initial growth rates
changed sex sooner than those with slower initial
growth rates.
Crepidula cf. onyx
Crepidula cf. onyx from the field survey had a pro-
nounced female-biased sex ratio with 66% female, 16%
transitionals and only 18% males (Table 2, Fig. 1).
Transitional individuals of this species
show both a penis and an fgp. Transi-
tionals ranged in size from 11.7 to
25.4 mm, but size overlap between
males and females suggests that a sex
change usually occurred around 13 to
14 mm. The size at sex change esti-
mated by the logistic regression (L50)
was 13.8 mm. 
Juveniles raised in isolation usually
developed into males; 20 d after initia-
tion of the experiment, 29 individuals
had developed a penis and 6 had a
penis rudiment; 20 d later, most indi-
viduals were already transitional.
Despite the fact that most (82%) of
these were observed with a penis,
those were noticeably shorter and
92
Table 2. Crepidula cf. onyx, C. incurva and Bostrycapulus calyptraeformis. Data
from field surveys and sex change experiments showing size (mm) at sex change
and size overlap. L50: size at which 50% of individuals were male. ?: not applicable
Treatment Size overlap Size of L50 Max.
(min. female transitional individual
? max. male) individuals size
C. cf. onyx
Field 13.0?13.9 a11.7?19.0a 13.80 29.6
Expt 10.5?14.9 11.1?15.7 ? 26.1
C. incurva
Field 6.3?9.8 6.8?12.2 08.05 18.1
Expt 10.4?13.7 8.1?12.7 ? 20.5
B. calyptraeformis
Field 12.7?19.7 9.3?14.6 14.60 23.1
Expt 12.2?17.2 0? ? 19.2
aOutlier at 25.4 mm
Collin et al.: Calyptraeid sex change
thinner than the penes developed by the small
males kept in association with conspecifics. Those
individuals not observed with a penis may have
had one for so short a time that it was not detected
by our infrequent sampling.
There was a significant effect of treatment on
size at sex change (ANOVA p < 0.0001. Fig. 2,
Table 3). Solitary males and the first of the
paired males changed sex change at a signifi-
cantly smaller size than did males placed with
females (9.6 vs. 12.6 mm; Fisher?s PLSD test, p <
0.05) Among-group variance accounted for 34%
of the total variance. The second of the paired
males to change showed the same size at sex
change as males paired with females. The results
were qualitatively the same when the analyses
were repeated using the smallest recorded
female size for each individual. Using
the relationship between shell length
and dry weight (ln[wt] = ?6.883 +
0.127 ? length; r2 = 0.89; p < 0.001)
showed that solitary and first males in a
pair changed sex at 70% of the dry
weight (0.0035 g) at which males paired
with females changed sex (0.0051 g).
Sex change could be very rapid and
took less than 3 wk. Many of the males
that changed sex were not observed in a
transitional phase; 50% of males in the
solitary treatment and the first male in the
paired males treatment had changed sex
by 60 d (Fig. 3), while males paired with
females and the second in a pair of males
had changed by about 300 d. In the soli-
tary and first male treatments, all individ-
uals became female.
93
0
2
4
6
8
10
4 8 12 16 20 24 28
Shell length (mm)
N
um
be
r  
of
 in
di
vi
du
al
s
Male 
Transitional
Female
Crepidula cf. onyx
Fig. 1. Crepidula cf. onyx. Size?sex distribution for field-collected
samples. Arrow indicates logistic regression estimate of size at 
which 50% were male (L50). n = 110
Fig. 2. Crepidula cf. onyx, C. incurva and Bostrycapulus calyptraeformis.
Differences in last recorded male size in 3 experimental treatments; error
bars = 95% CI. Average sizes at sex change were C. cf. onyx: solitary males
9.8 mm, male with male 9.4 mm, male with female 12.7 mm; C. incurva: soli-
tary males 9.2 mm, male with male 9.2 mm, male with female 10.6 mm;
B. calyptraeformis: solitary males 14.0 mm, male with male 13.6 mm
C
um
ul
at
iv
e 
se
x 
ch
an
ge
(p
ro
p
or
tio
n 
m
al
es
)
Crepidula  cf. onyx
0.2
0.6
1.0
100 200 300 400 5000
Number of days
0 100 200 300 400 500
Bostrycapulus calyptraeformis
0 100 200 300 400 500
Crepidula incurva
Solitary Male with a male Second male Male with a female
Number of daysNumber of days
Fig. 3. Crepidula cf. onyx, C. incurva and Bostrycapulus calyptraeformis. Results of Kaplan-Meier survival analysis of sex change for
each species. Estimated mean (?SE) number of days until 50% had changed sex were C. cf. onyx: solitary males 61 ? 8 d, male with
male 61 ? 11 d, second male 318 ? 91 d, male with female 289 ? 28 d. C. incurva: solitary males 80 ? 17 d, male with male 60 ? 5.9 d,
second male in a pair 370 ? 137 d, male with female >400 d. B. calyptraeformis: solitary males 221 ? 31 d, male with male 286 ? 48 d
Mar Ecol Prog Ser 293: 89?97, 2005
Daily growth rates during the transitional period
were significantly higher than prior to the sex change
(paired t-test; df = 30; p < 0.05) or after the change
(df = 22; p < 0.0001) (Fig. 4; meanmale = 0.14 mm d?1;
meantransitional = 0.19 mm d?1; meanfemale 0.12 mm d?1).
An ANCOVA of growth rates using average size over
the observation period as a covariate found a signifi-
cant effect of sex (p < 0.05) but not of size (p > 0.05).
Solitary males were placed into 1 of 2 categories:
those that had changed sex before and those that had
changed sex after Day 61 of the experiment (the date
by which half had changed). Those that changed be-
fore day 61 were larger on that day than those that had
not yet changed (13.3 vs. 9.6 mm; t-test p < 0.0001).
There was a significant effect of growth rate on the
probability of changing sex (p < 0.01, r2 = 0.62), but
there was no significant effect of initial size (p > 0.1).
Despite having faster growth rates, those individuals
that changed sex before Day 61 changed at a smaller
size (8.46 mm) than those that changed sex later
(10.59 mm) (t-test; p < 0.005). 
Among 23 paired males, the larger individuals
always changed sex first.
Crepidula incurva
Field samples of Crepidula incurva had a slightly
female-biased sex ratio, with 56% females (Fig. 5).
Very few (3%) transitional individuals were collected
and sex change occurred at 6 to 12 mm (Table 2). The
L50 size at sex change was estimated as 8.05 mm
(Table 2); 50% of these snails occurred in stacks, and
those that were stacked had a larger L50 (8.47 mm)
than solitary individuals (L50 = 7.14 mm). 
Of 38 individuals raised alone, 37 passed through a
male phase with a distinct, well-developed penis; 1 of
the individuals grew quickly, and was not observed in
the normal male size range. In most cases, the penis
was well-developed by the first census after 3 wk into
the experiment and was observed for 1 or 2 subsequent
intervals before sex change. 
There was a significant effect of social treatment on
size at sex change (ANOVA p < 0.05. Fig. 2, Table 3).
Solitary males and males paired with males changed at
a significantly smaller size (9.2 mm) than males paired
with females (10.6 mm) (Fisher?s PLSD test, p < 0.05).
The among-group variance accounted for 11% of the
total variance and the difference between treatments
(1 mm) was surprisingly small. Shell lengths of 9.2 and
10.6 mm represent dry weights of 0.0038 and 0.0052 g
respectively (ln[wt] = ?10.479 + 2.209 ? ln[length]). The
second male of paired males changed sex at the same
size as males paired with females. Sex change occurred
at slightly larger sizes in laboratory-reared individuals
than in field-collected samples (Table 2). This is proba-
bly because Crepidula incurva lives attached to other
small snails and therefore growth was likely to be sub-
strate-limited in the field but not in our containers.
After 60 to 80 d, 50% of solitary males and the first
male of a pair to change had changed sex (Fig. 3),
94
Table 3. Crepidula cf. onyx and C. incurva. Results of ANOVA
testing effect of teatment on size at sex change in snails 
exposed to 3 experimental social treatments
df SS MS F p
C. cf. onyx
Treatment 02 184.9 92.40 15.86 <0.0001
Residual 82 477.8 5.8
C. incurva
Treatment 02 014.2 7.1 3.53 0.035
Residual 66 133.2 2.0
Transitional FemaleMale
0.05
0.15
0.25
0.35
G
ro
w
th
 (m
m
 d
-1
)
4 8 12 16 20
Crepidula cf. onyx
0
0.05
0.15
0.25
6 10 14 18
Crepidula incurva Bostrycapulus calyptraeformis
0.02
0.08
0.14
11 13 15 17
Average size (mm)Average size (mm) Average size (mm)












































































Fig. 4. Crepidula cf. onyx, C. incurva and Bostrycapulus calyptraeformis. Relationship between daily growth rate and average
size during observation period for individuals during male, transitional and female phases. Average growth rates for all 3 cate-
gories were significantly different for C. cf. onyx, growth rates for transitionals was higher than for both male and females in C. 
incurva, and there was no difference between the categories in B. calyptraeformis
Collin et al.: Calyptraeid sex change
while the second male of a pair to change and males
paired with females changed after more than 300 d. In
the solitary treatment, all males became female.
Daily growth rates were significantly affected by size
(ANCOVA p < 0.05; Fig. 4), and marginally by sex
(ANCOVA; p = 0.06). Daily growth rates for transition-
als were significantly higher than for periods before
and after the transition (paired t-test meanmale =
0.06 mm d?1; meantransitional = 0.11mm d?1; meanfemale =
0.07 mm d?1; p < 0.01). When standardized for size
(ANCOVA) all 3 categories were significantly different
and growth (mm d?1) was still greater for transitionals
than for males and females (p < 0.05).
Snails that had changed sex before Day 77 (the date
by which half had changed) were significantly larger
(11.8 mm) on that day than those that had not changed
(9.2 mm) (t-test p < 0.0001). Those that changed
first were of smaller size at the sex change
(8.2 mm) than those that had not changed by this
date (10.2 mm) (t-test, p < 0.005). There was a
significant effect of growth rate on the probabil-
ity of changing sex (p < 0.05, r2 = 0.68), and initial
size did not have a significant effect (p > 0.2).
Among paired males, the larger individual
always changed sex first (n = 31).
Bostrycapulus calyptraeformis
Samples of Bostrycapulus calyptraeformis had
a male-biased sex ratio, with only 37% females
in the population and very few (4%) transitional
individuals (Fig. 6). The estimated L50 was
14.6 mm and fell between the smallest observed
female (12.7 mm) and the largest male (19.7 mm).
Transitionals were of smaller size (9.3 to
14.6 mm). 
When individuals were raised alone, 24 of the
27 surviving animals passed through a male phase
with a distinct, well-developed penis; 3 individu-
als appeared to bypass the male phase (or this was
too brief to be observed within our 3 wk sampling
periods). Development of the penis was slower
than in Crepidula incurva. The penis was gener-
ally visible after 60 d and was often retained for
more than 90 d. Growth rates were generally
slower in this species (Fig. 4).
Because growth and sex change were so slow,
we were not able to obtain sample sizes large
enough to examine males paired with females.
However, there was no significant difference in
the size at sex change for solitary males and those
paired with males (t-test; p > 0.1; n = 50). Solitary
males and those paired with males changed sex
after a similar number of days (Fig. 3). 
For solitary males there were no significant differ-
ences in growth rates before, during or after sex change
(p > 0.2), but the sample sizes were small (n = 10 to 20).
Individuals that changed sex before Day 194 (the day
by which half had changed) were significantly larger
(15.3 mm) than those that had not yet changed (14.1 mm)
on that day. The last male size was smaller (12.9 mm)
for males that changed sex by Day 194 than for those
that had not changed by this data (15.5 mm) (t-test,
p < 0.0005). There was a significant effect of growth
rate on the probability of having changed sex by this
date (p < 0.05 r2 = 0.30), and initial size was not a sig-
nificant covariate (p > 0.5), although growth rate was
correlated with initial size.
The larger individual in the paired males was most
likely to change sex first (19 of 23 pairs; ?2, p < 0.05). 
95
0
10
20
30
40
50
2 4 6 8 10 12 14 16 18
Shell length (mm)
N
um
be
r o
f i
nd
iv
id
ua
ls
Crepidula incurva
Male 
Transitional
Female
Fig. 5. Crepidula incurva. Size-sex distribution for field-collected
individuals. Arrow indicates logistic regression estimate of size at 
which 50% were male (L50). n = 347
0
10
20
30
5 7 9 11 13 15 17 19 21 23
Shell length (mm)
N
um
be
r 
of
 in
di
vi
du
al
s
Bostrycapulus calyptraeformis
Male 
Transitional
Female
Fig. 6. Bostrycapulus calyptraeformis. Size-sex distribution for
field-collected individuals. Arrow indicates logistic regression esti-
mate of size at which 50% were male (L50). n = 544.
Mar Ecol Prog Ser 293: 89?97, 2005
DISCUSSION
Despite the fact that these 3 species of calyptraeids
show different development, density and sex ratio, we
found that the general patterns of sex change were
similar. In all 3 species, (1) isolated juveniles passed
through a male phase (although this was somewhat
poorly developed and abbreviated in Crepidula cf.
onyx), (2) sex change occurred at the same size for
males raised alone and males raised with another
male, (3) males that changed sex earlier had higher
growth rates and changed at smaller sizes than males
that changed later, and (4) the largest male in a pair of
males almost always changed before the smaller male.
In addition, in 2 out of 3 species, transitionals had
higher growth rates than males or females.
Life history theory predicts that there should be an
optimal size at sex change (Charnov 1982). For species
with labile sex change, it is expected that optimal size
at sex change should show less variation among indi-
viduals under constant conditions (e.g. laboratory con-
ditions) than in a variable environment (e.g. the field).
This does not appear to be the case for the calyptraeid
species examined here. The size range between the
smallest female and largest male were similar for sam-
ples from the field and in the laboratory experiment
(Table 2), as was the size range at which transitionals
were observed. In fact, the range of size overlap ob-
served for males and females of Crepidula cf. onyx was
actually larger in the laboratory than in field-collected
samples. This is surprising as the cue from con-
specifics, which is presumably chemical or tactile,
must have been highly concentrated in our treatments
relative to the field (with the possible exception of
Bostrycapulus calyptraeformis, which occurs in high
densities in the field).
Authors of previous comparisons of sex change
among calyptraeids have suggested that the level and
types of responses to conspecifics varies among spe-
cies, and that species with large body size, planktonic
larvae and gregarious behavior are more responsive to
conspecifics than are less social species with direct
development (Coe 1938, Hoagland 1978). Our results,
however, show there is little difference in the type of
effect of conspecific interactions between Crepidula
incurva, a somewhat gregarious species with plank-
tonic development, and C. cf. onyx a solitary species
with direct development. The available evidence also
suggests that Bostrycapulus calyptraeformis, a gregar-
ious species with planktonic development, shows simi-
lar patterns. Warner et al. (1996) also concluded that
the direct-developing C. norrisiarum (which is large,
has direct development, and is usually found in groups
of 2 to 13) also modulates sex change in response to
associations with conspecifics. It seems clear from this
and previous studies that Crepidula species do not fall
into 2 groups defined by a suite of life history charac-
ters, and that regardless of mode of development or
frequency of associations between individuals, they
generally have labile sex change which can be modu-
lated by individuals' associations with conspecifics. 
We did, however, find a difference between Crepi-
dula cf. onyx and C. incurva in the magnitude of the
treatment effect. Associations with conspecifics altered
the size at sex change by 3 mm and 32% dry body
weight in C. cf. onyx and accounted for 34% of the
variation in size at sex change. This 3 mm difference is
similar to the observed size range of transitionals
(4.5 mm) and is similar to the range of size overlap for
males and females observed in the experimental treat-
ment (Table 2). In C. incurva, however, association
with conspecifics altered the size at sex change by only
1 mm or 23% dry body weight and accounted for only
11% of the variation in size at sex change. This 1 mm
difference is considerably smaller than the 3.5 mm
range for transitionals and 4.5 mm range of size over-
lap for males and females observed among experimen-
tal individuals. This result suggests that while interac-
tions with conspecifics accounts for a reasonable
amount of the observed variation in C. cf. onyx, the
majority of the observed variation in size at sex change
is due to other factors in C. incurva. The greater re-
sponsiveness of the more solitary species to associa-
tions with conspecifics compared to the social species
could reflect the low predictability of encounters
between potential mates, which would select for rapid
and sensitive responses to nearby conspecifics.
We found that the male phase is prolonged in the
presence of a female both with respect to time and
with respect to size. However, contrary to theoretical
predictions and the conclusions of Coe (1938), males
housed with females neither delayed sex change
indefinitely, nor did they reach unusually large sizes
before changing sex. In all treatments, those males that
did not change sex also remained small. This suggests
that sex change cannot be easily uncoupled from body
size, and that those factors that increase growth
inevitably lead to sex change. A connection between
growth rate and sex change is supported by the result
that faster-growing individuals changed sex sooner
and at a smaller size than slower-growing individuals.
Individual growth histories have been predicted to
provide cues for other irreversible life history transi-
tions (e.g. metamorphosis, Wilbur & Collins 1973) and
in combination with current size, may provide cues
that initiate sex change. If this is the case, the slower
growth rates of those males kept with females in our
experiments may account for their larger size at sex
change. However, males in the solitary treatment had
higher growth rates than the paired males, yet
96
Collin et al.: Calyptraeid sex change
changed sex at the same size. Experiments designed
specifically to examine this hypothesis are necessary to
further elucidate the effects of growth rate on sex
change.
Acknowledgements. We thank the McGill-STRI Program for
organizing the internship program from which this project
originated and M. Salazar and M. Venagas for their help with
animal care. R. Robertson and J. Christy and 4 anonymous
reviewers commented on a previous version of the manu-
script.
LITERATURE CITED
Allsop DJ, West SA (2003) Changing sex at the same relative
body size. Nature 425:783?784
Alonzo SH, Warner RR (2000a) Female choice, conflict
between the sexes and the evolution of male alternative
reproductive behaviours. Evol Ecol Res 2:149?170
Alonzo SH, Warner RR (2000b) Allocation to mate guarding or
increased sperm production in a Mediterranean wrasse.
Am Nat 156:266?275
Baldwin AP, Bauer RT (2003) Growth, survivorship, life-span,
and sex change in the hermaphroditic shrimp Lysmata
wurdemanni (Decapoda: Caridea: Hippolytidae). Mar Biol
143:157?166
Bauer RT, VanHoy R (1996) Variation in sexual systems
(protandry, gonochorism) and reproductive biology among
three species of the shrimp genus Thor (Decapoda:
Caridea). Bull Mar Sci 59(1):53?73
Charnov EL (1982) The theory of sex allocation. Princeton
University Press, Princeton, NJ
Chen MH, Soong K (2000) Sex change in the hat snail, Calyp-
traea morbida (Reeve) (Gastropoda: Calyptraeidae): an
analysis of substratum, size, and reproductive characteris-
tics. Veliger 43:210?217
Chiba S, Goshima S, Shinomiya Y (2003) Male?male compe-
tition selects for delayed sex change in the protandrous
shrimp Pandalus latirostris. Mar Biol 142:1153?1157
Coe WR (1936) Sex phases in Crepidula. J Exp Zool 72:
455?477
Coe WR (1938) Influence of association on the sexual phases
of gastropods having protandric consecutive sexuality.
Biol Bull (Woods Hole) 75:274?285
Coe WR (1948) Nutrition and sexuality in protandric gas-
tropods of the genus Crepidula. J Exp Zool 108:155?169
Coe WR (1953) Influences of association, isolation and nutri-
tion on the sexuality of snails in the genus Crepidula.
J Exp Zool 122:5?19
Collin R (1995) Sex, size, and position: a test of models pre-
dicting the size at sex change in the protandrous gastro-
pod Crepidula fornicata. Am Nat 146:815?831
Collin R (2000) Sex change, reproduction and development of
Crepidula adunca and C. lingulata (Gastropoda: Calyp-
traeidae). Veliger 43:24?33
Gould HN (1917) Studies on sex in the hermaphrodite mol-
lusk Crepidula plana. 2. Influence of the environment on
sex. J Exp Zool 23:225?250.
Gould HN (1919) Studies on sex in the hermaphrodite mol-
lusk Crepidula plana . 3. Transference of the male produc-
ing stimulus through sea water. J Exp Zool 29:113?120
Gould HN (1947) Conditions affecting the development of the
male phase in Crepidula plana. Biol Bull (Woods Hole) 93:
194
Gould HN (1952) Studies on sex in the hermaphrodite mol-
lusk Crepidula plana. IV. Internal and external factors
influencing growth and sex development. J Exp Zool 119:
93?163
Gould HN, Hsiao SC (1948) New experiments and obser-
vations on sexual instability in Crepidula plana. Biol Bull
(Woods Hole) 255?256
Hoagland KE (1978) Protandry and the evolution of environ-
mentally mediated sex change: a study of the Mollusca.
Malacologia 17:365?391
Koeller P, Mohn R, Etter M (2000) Density dependant sex
change in northern shrimp, Pandalus borealis, on the Scot-
ian Shelf. J Northwest Atl Fish Sci 27:107?118
Lutnesky MMF (1994) Density-dependent protogynous sex
change in territorial-haremic fishes: models and evidence.
Behav Ecol 5:375?383
Mackie MC (2003) Socially controlled sex-change in the half-
moon grouper, Epinephelus rivulatus, at Ningaloo Reef,
Western Australia. Coral Reefs 22:133?142
Munday PL (2002) Bi-directional sex change: testing the
growth-rate advantage model. Behav Ecol Sociobiol 52:
247?254
Perry AN, Grober MS (2003) A model for social control of sex
change: interactions of behavior, neuropeptides, gluco-
corticoids, and sex steroids. Horm Behav 43:31?38
Petersen CW, Warner RR, Shapiro DY, Marconato A (2001)
Components of fertilization success in the bluehead
wrasse, Thalassoma bifasciatum. Behav Ecol 12:237?245
Robertson DR (1972) Social control of sex reversal in a coral
reef fish. Science 177:1007?1009
Ross RM, Losey GS, Diamond N (1983) Sex change in a coral
reef fish: dependence on stimulation and inhibition on
relative size. Science 221:574?575
Sokal RR, Rohlf FJ (1995) Biometry: the principles and prac-
tice of statistics in biological research, 3rd edn. WH Free-
man & Co, New York
Soong K, Chen JL (1991) Population structure and sex-change
in the coral-inhabiting snails Coralliophila violacea at
Hsiao-Liuchiu, Taiwan. Mar Biol 111:81?86
Soong K, Chen JL (2003) Sex expression of an immobile coral-
inhabiting snail, Quoyula monodonta. Mar Biol 143:
351?358
St. Mary CM (1993) Sex allocation in a simultaneous her-
maphrodite, the blue-banded goby (Lythrypnus dalli): the
effects of body size and behavioral gender and the conse-
quences for reproduction. Behav Ecol 5:304?313
Warner RR, Swearer S (1991) Social control of sex change in
the bluehead wrasse, Thalassoma bifasciatum (Pisces:
Labridae). Biol Bull (Woods Hole) 181:199?201
Warner RR, Fitch DL, Standish JD (1996) Social control of sex
change in the shelf limpet, Crepidula norrisiarum: size-
specific responses to local group composition. J Exp Mar
Biol Ecol 204:155?167
Wilbur HM, Collins JP (1973) Ecological aspects of amphibian
metamorphosis. Science 182:1305?1314
97
Editorial responsibility: Otto Kinne (Editor-in-Chief), 
Oldendorf/Luhe, Germany
Submitted: September 16, 2004; Accepted: November 30, 2004
Proofs received from author(s): May 9, 2005