The selective advantage of host feminization: A case study of the green crab Carcinus maenas and the parasitic barnacle Sacculina carcini

Abstract

Male crabs infected by parasitic barnacles (Rhizocephala) are known to be morphologically feminized. Here, we investigate morphological chances in green crabs, Carcinus maenas, induced by the parasitic barnacle Sacculina carcini. Infected males acquire a broader, longer and segmented abdomen, fringed with marginal setae. Copulatory appendages and pereopods are reduced in length, and the chelae become smaller. The feminization show great individual variation. Males with scars from lost externae, the parasites reproductive organ situated under the abdomen, are less modified than males carrying an externa, and the feminization is more pronounced in smaller than in larger males. No super-feminization is evident in female crabs that remain morphologically unaffected by infection. The protective value of a parasitically induced enlargement of the male abdomen may constitute an adaptation that increases parasite longevity. The additional effects on male morphology are viewed as pleiotropic side effects of the main adaptive value of enlarging the abdomen.

Full text

ORIGINAL PAPER
The selective advantage of host feminization: a case study
of the green crab Carcinus maenas and the parasitic
barnacle Sacculina carcini
Tommy Kristensen •Anders Isberg Nielsen •Anders Isak Jørgensen •
Kim N. Mouritsen •Henrik Glenner •Jens T. Christensen •Jørgen Lu
¨tzen •
Jens T. Høeg
Received: 13 April 2012 / Accepted: 19 June 2012 / Published online: 5 July 2012
ÓSpringer-Verlag 2012
Abstract Male crabs infected by parasitic barnacles
(Rhizocephala) are known to be morphologically femi-
nized. Here, we investigate morphological chances in green
crabs, Carcinus maenas, induced by the parasitic barnacle
Sacculina carcini. Infected males acquire a broader, longer
and segmented abdomen, fringed with marginal setae.
Copulatory appendages and pereopods are reduced in
length, and the chelae become smaller. The feminization
show great individual variation. Males with scars from lost
externae, the parasites reproductive organ situated under
the abdomen, are less modified than males carrying an
externa, and the feminization is more pronounced in
smaller than in larger males. No super-feminization is
evident in female crabs that remain morphologically
unaffected by infection. The protective value of a parasit-
ically induced enlargement of the male abdomen may
constitute an adaptation that increases parasite longevity.
The additional effects on male morphology are viewed as
pleiotropic side effects of the main adaptive value of
enlarging the abdomen.
Introduction
The way rhizocephalan barnacles influence their crusta-
cean hosts has elicited considerable interest, since Giard
(1886) discovered that males of the spider crab Macro-
podia rostrata L., 1758, parasitized by the sacculinid
Drepanorchis neglecta Fraisse, 1877, adopted a number of
feminine traits, in particular a broadened abdomen. In
female crabs, the abdomen covers and protects the egg
mass of the gravid crab and, if parasitized by D. neglecta,
the external reproductive apparatus, the externa, of the
parasite. Male crabs have narrow abdomens that fit in a
groove of the thoracic sternum of the crab like a key in a
lock. Infected male crabs develop abdomens as broad as
females that provide the same degree of protection to the
parasite as is normally offered to the brood. Since Giard’s
discovery, it has become clear that all studied brachyuran
hosts parasitized by species of the family Sacculinidae
become morphologically modified, sterilized and sub-
jected to physiological and behavioural changes (Høeg
and Lu
¨tzen 1996; Høeg 1996). As a rule, males are much
more heavily influenced than females and may be femi-
nized to such a degree that on a superficial view they
become almost indistinguishable from the opposite sex. In
most cases, sacculinized females undergo no morpholog-
ical changes.
Sacculina carcini Thompson, 1836, parasitizes a great
number of portunid crab species (Høeg and Lu
¨tzen 1985;
Øksnebjerg, 2000). There are several reports on how it
influences its host morphologically, but some of them
pertain to another species of Sacculina,S. benedeni
Kossmann, 1872 (Bulgurkov 1938; Vernet-Cornubert
1958). Day (1935) studied the way S. carcini influences
Liocarcinus holsatus (Fabricius, 1798) while Veillet
(1945) examined the modifications inflicted by S. carcini
Communicated by S. A. Poulet.
T. Kristensen J. Lu
¨tzen J. T. Høeg
Department of Biology, Section for Marine Biology,
Copenhagen University, Universitetsparken 4,
2100 Copenhagen, Denmark
A. I. Nielsen A. I. Jørgensen K. N. Mouritsen (&)
J. T. Christensen
Department of Bioscience, Marine Ecology, Aarhus University,
Ole Worms Alle
´1, 8000 Aarhus C, Denmark
e-mail: kim.mouritsen@biology.au.dk
H. Glenner
Department of Biology, Biological Institute, University
of Bergen, Thormøhlsgae 55, 5020 Bergen, Norway
123
Mar Biol (2012) 159:2015–2023
DOI 10.1007/s00227-012-1988-4

on the estuary crab, Carcinus aestuarii Nardo, 1847. From
these studies, it is apparent that the degree each individual
S. carcini parasite exerts on its host is much more variable
than the non-or-all effect observed in D. neglecta.This
applies in particularly to how S. carcini modifies its
classical host, the green crab Carcinus maenas, where the
abdominal morphology of an infected male crab ranges
from almost no modifications to an appearance approxi-
mating that of an adult female (present study). This var-
iation of a parasite-induced host character of potential
importance for the survival of the parasite provides an
excellent chance for studying a morphological trait under
possible Darwinian selection. Hence, as part of a 3-year
study of the biology and parasites of the green crab, we
embarked on a large-scale field study on how C. maenas
is morphologically affected by S. carcini. In particular, it
was decided to investigate to what extent a feminized host
abdomen is of selective advantage for the parasite.
Materials and methods
Field sampling and measurements
In total, 366 uninfected and 249 sacculinized green crabs
were included in the study, all collected at four localities
in Limfjorden, Denmark. Limfjorden is a major but
shallow waterway (depth typically 4–6 m) that traverses
the northern part of Jutland connecting the North Sea
with the Kattegat/Baltic Sea estuary. The salinity varies
from c. 31 psu in the west to c. 23 psu in the east, with
bottom temperatures between 18 °C in July–August and
1–2 °C in February–March. The green crabs occur at all
depths and all over the fjord, while S. carcini in some
years may be absent from some of the eastern less saline
areas.
Crabs were caught in baited traps at three localities:
Venø Bay, off Bjørndrup and Lovns Broad (May–August
2009; depths: 1–6 m; salinity: 28–31 psu, 22–29 psu and
22–25 psu, respectively). Crabs from a fourth locality, off
Rønbjerg harbour, were collected in fish traps (26 June–2
July 2006; 2 m; 26 psu). The majority of the crabs was
above 30 mm in carapace width and therefore considered
mature. Six types of crabs were distinguished: apparently
non-parasitized males and females, parasitized males and
females with externae of Sacculina carcini, and parasitized
male and female crabs with scars from externae that had
been lost. Only few sacculinized female crabs were caught
at Rønbjerg, and thus, the analyses of these data focus
solely on males and non-infected females. The abdomens
of all crabs were lifted from the thorax to inspect them for
infestation (presence of externa or scar). To verify that all
apparently non-infected crabs were indeed uninfected, the
crabs from Lovns Broad, Venø Bay and Bjørndrup were
dissected for presence of an internal root system. The non-
infected crabs used for comparison were chosen randomly,
and damaged specimens were discarded. The number and
infection status of crabs used in the study are found in
Table 1. Besides the externally sacculinized crabs, a
number of males with broader abdomens than usual were
found. When dissected, those crabs were found to contain
rhizocephalan rootlets. Such crabs were also noted by Day
(1935), who explained them as sacculinized males that had
lost their externae and moulted again, and by Werner
(2001), who considered them parasitized males, the externa
of which had not yet emerged. In the present study, these
males were excluded from the analysis.
The following characters were measured on the col-
lected green crabs: (1) carapace width (CW) defined as the
distance between the fifth pair of lateral spines (Fig. 1a);
(2) maximum width of abdominal segment 3, 5 and 6
(Fig. 1b–d); (3) maximum length of abdominal segment 6
(Fig. 1b–d); (4) distance from base to tip of the left first
pair of pleopod of males (copulatory appendage) (Fig. 1f);
(5) height of the right claw from the dorsal notch near the
base of movable finger to the ventral margin (Fig. 1g); (6)
length of left and right pereopod 3 (second walking leg)
from base to tip (Fig. 1h). Data from solely the longer of
the two measured pereopods were used in the analysis.
Carapace width and pereopod length were measured by
ruler to the nearest mm, whereas the remaining characters
were measured to the nearest 0.1 mm using a dissection
Table 1 The total material studied distributed on sampling location and gender of the host, green crab Carcinus maenas
Location Males Females
Uninfected With externa With scar Uninfected With externa With scar
Lovns Broad (56°380400N,9°130300E) 49 0 0 0 0 0
Bjørndrup (56°4902500N,8°5104100E) 0 17 16 2 13 0
Venø Bay (56°3402800N,8°4003900E)0542988342
Rønbjerg (56°5302800N,9°904900E) 118 84 0 109 0 0
Uninfected: non-sacculinized crabs; with externa: sacculinized crabs possessing an externa; with scar: sacculinized or previously sacculinized
crabs with scar after lost externa
2016 Mar Biol (2012) 159:2015–2023
123

microscope fitted with an ocular micrometer. Pereopod
length, width of the fifth abdominal segment and length of
segment 6 were only obtained from crabs collected at
Rønbjerg, on which animals no measure of abdominal
segment 3 and claw height was achieved.
Morphological modification as a result of sacculiniza-
tion is expressed as the relative abdominal width (RAW),
defined as the ratio between the widths of segment 3
(which is comparatively unaffected) and segment 6 (which
increases following infection). This measure could not be
calculated for the Rønbjerg sample, and hence, the width of
segment 6 is expressed in relation to CW. Also, length of
the copulatory appendage, claw height and length of the
second walking leg are expressed in relation to CW.
Explanation of terms: externa—the bean-shaped repro-
ductive body of the rhizocephalan parasite attached to and
located under the ventral side of the crab’s abdomen.
Scar—a dark, circular spot arising from the point of
attachment when the externa had died and detached.
Scarred crabs—crabs with a scar. Sacculinized crabs—
crabs infected by S. carcini.Externally sacculinized
crabs—crabs having a visible externa or scar. Root sys-
tem—the sacculated trophic part of the parasite that
consists of numerous connected tubular and semitranspar-
ent rootlets embedded in the host tissue.
Data analyses
The data analysis was carried out using IBM SPSS 19.0
(one-way analyses) and SMATR 2.0 (regressions and
ANCOVA). In those instances where data did not meet the
prerequisites of parametric tests (normality, homogeneity
of variance), data were either transformed prior to analysis
or nonparametric tests were applied. All performed post
hoc tests are corrected for multiple comparisons.
Crabs from Rønbjerg were analysed separately by
expressing all measured characters as a function of CW in
analyses of covariance comparing the three main groups of
crabs in question: infected males, uninfected males and
uninfected females. This was done because the Rønbjerg
animals (1) were collected at a single site within a narrow
time span and depth (low variance), (2) included a rela-
tively large number of individuals from all of the three
main crab groups (see Table 1) (high statistical power) and
(3) were not measured for the width of segment 3 (inval-
idating RAW calculation, and hence, one-way analyses).
Fig. 1 Morphological characters and measures taken on uninfected
and sacculinized green crabs Carcinus maenas.aDorsal view of
carapace. Dorsal view of abdomen of non-infected female (b) infected
male with marginal setae (c) and non-infected male (d). eVentral
view of infected but unmodified male with scar from lost externa
(white arrow). fCopulatory appendage (pleopod 1). gClaw. hPereo-
pod 3 (second walking leg). The double-arrows show distances
measured. Not to scale; see text for details
Mar Biol (2012) 159:2015–2023 2017
123

Results
Abdominal morphology
The relative abdominal width (RAW) of crabs from Venø
Bay, Bjørndrup and Lovns Broad combined, differed
according to infection status and gender (Fig. 2a). The
RAW of externally sacculinized males (externae present)
was significantly smaller (i.e. having broader abdomen; on
average 11.1 %) than that of uninfected males (Fig. 1c–d),
whereas scarred sacculinized males (Fig. 1e) attained
intermediate RAW values. In an analysis including also
female categories, scarred males were statistically similar
to the remaining male categories (Fig. 2a). Considering
solely males in a separate analysis, all three male catego-
ries could be separated statistically (Kruskal–Wallis test,
v
2
2
=65.73, P\0.0005; post hoc test: P[[\0.0005;
0.044]). Hence, scarred males have a slightly narrower
abdomen than males with externae, but also a broader
abdomen than non-infected males. Females on the other
hand attained a similar RAW regardless of infection status
and always had a broader abdomen than any male
(Fig. 2a). For all three male categories, there was no sig-
nificant relationship between RAW and carapace width
(Linear regression; r
2
\0.054, P[0.054). In contrast, the
RAW of non-infected females was negatively related to
carapace width (r
2
=0.325, P\0.0005) as was the case
also for infected females (scarred and with externae) within
the same size range as non-infected females (r
2
=0.107,
P=0.030). Hence, regardless of infection status, the rel-
ative abdominal width (RAW) increases with size/age in
female crabs, but not in males.
In the Rønbjerg sample, the width of the sixth abdom-
inal segment as a function of carapace width (Fig. 1a–d)
was used to analyse the parasite-induced abdominal
broadening in male crabs. Width of segment 6 increased
linearly with carapace width for non-infected females, non-
infected males and infected males, albeit at a significantly
higher rate in females than in males irrespective of the
infection status of the latter [Fig. 3a; equality of slope
among groups: Standardized Major Axis (SMA) analysis,
T=200.78, P\0.0005; post hoc test: females versus
male groups, P=0.0001]. But whereas the slope of
infected and non-infected males was similar (post hoc test,
P=0.687), suggesting similar metric modification
regardless of size/age, infected males generally attained a
broader abdomen than non-infected (by 12.4 % based on
grand means) (Fig. 3a; equality of grand means: SMA on
square-root transformed data, Wald test, W
1
=223.39,
P\0.0005). Similar results were obtained also for the
width of the fifth abdominal segment (data not shown),
suggesting that infection results in a broadening of the
entire abdomen. Not only the width but also the length of
Fig. 2 Abdominal and claw dimension of infected and uninfected
Carcinus maenas from Venø Bay, Bjørndrup and Lovns Broad
combined. aBox-plot (mean, SE, range) of relative abdominal width
(RAW) of infected females with externae or scar (FI,n=49), non-
infected females (FN,n=90), infected males with externae (ME,
n=71), infected males with scars (MS,n=45) and non-infected
males (MN,n=49). Median RAW differs significantly among
infection status categories (Kruskal–Wallis test, v
4
2
=245.59,
P\0.0005). Lines above box-plots connect statistically similar
categories according to post hoc test correcting for multiple compar-
isons (Pvalues given). All other pair-wise comparisons demonstrated
significantly different RAW (P\0.0005). bBox-plot (mean, SE,
range) of the claw:carapace ratio (claw height divided by carapace
width). Sample sizes and category abbreviations: see above. Median
claw:carapace ratio differs significantly among categories (Kruskal–
Wallis test, v
4
2
=151.239, P\0.0005). Lines above box-plots
connect statistically similar categories according to post hoc test
correcting for multiple comparisons (Pvalues given). All other pair-
wise comparisons demonstrated significantly different claw:carapace
ratios (P\0.022)
2018 Mar Biol (2012) 159:2015–2023
123

the abdomen of males increases following infection. As is
the case for the width, the length of the sixth abdominal
segment increased linearly with carapace width for all three
analysed groups of crabs, albeit at a significantly higher
rate in females than in the two male groups (Fig. 3b;
equality of slope among groups: SMA, T=113.20,
P=0.0001; post hoc test: females versus male groups,
P=0.0001). Whereas the slope of infected and non-
infected males was similar (post hoc test, P=0.739),
infected males generally attained a longer sixth abdominal
segment than non-infected (by 5.8 % based on grand
means) (Fig. 3b; equality of grand means: SMA, Wald
test, W
1
=75.34, P\0.0005). Hence, the abdomen of
sacculinized males is generally morphologically feminized,
but, in a male way: infected males do not adopt the inherent
allometric relationship characteristic of females.
Accompanying these abdominal changes of sacculinized
males, the usually fused third to fifth segments (Fig. 1d)
develop an articulation, although they never become
capable of independent movements as in females. In
addition, many of the sacculinized males acquired shorter
or longer setae along the margin of the abdomen as females
also possess (Fig. 1b, c). However, in not a single case did
they develop any trace of the setose pleopods 2–5 char-
acteristic for females. The morphology of these pleopods
also appeared unaffected by infection in females.
Fig. 3 Dimensions of abdomen, copulatory appendages and peropods
(legs) as a function of carapace width (CW) in infected and uninfected
Carcinus maenas from Rønbjerg. aStandardized Major Axis (SMA)
analysis of the width of the sixth abdominal segment for infected
males (open circle;n=84; y=0.173x?0.498; r
2
=0.860,
P\0.0005), non-infected males (closed circle;n=118;
y=0.169x-0.432; r
2
=0.922, P\0.0005) and non-infected
females (open square;n=109; y=0.530x-8.687; r
2
=0.673,
P\0.0005). bStandardized Major Axis (SMA) analysis of the
length of the sixth abdominal segment for infected males (open circle;
n=84; y=0.173x?0.498; r
2
=0.860, P\0.0005), non-infected
males (closed circle;n=118; y=0.169x-0.432; r
2
=0.922,
P\0.0005) and non-infected females (open square;n=109;
y=0.530x-8.687; r
2
=0.673, P\0.0005). c. Standardized Major
Axis (SMA) analysis of the length of the copulatory appendage
(pleopod 1) for infected males (open circle;n=84;
y=0.264x?0.187; r
2
=0.949, P\0.0005) and non-infected
males (closed circle;n=117; y=0.255x?1.017; r
2
=0.953,
P\0.0005). d. Standardized Major Axis (SMA) analysis of the
length of the longest third pereopod (second walking leg) for infected
males (open circle,n=77; y=1.47x-6.551; r
2
=0.969,
P\0.0005), non-infected males (closed circle;n=107;
y=1.49x-5.796; r
2
=0.949, P\0.0005) and non-infected
females (open square;n=100; y=1.22x?1.16; r
2
=0.961,
P\0.0005)
Mar Biol (2012) 159:2015–2023 2019
123

Male copulatory appendages
In the combined Venø Bay, Bjørndrup and Lovns Broad
samples, the length of the first pair of the copulatory
appendages was positively related to carapace width in all
three male categories: non-infected, scarred and externae
present (Linear regression, r
2
[0.809, P\0.0005). Cor-
rected for carapace width (i.e., the appendage:carapace
ratio), there was no statistically significant effect of
infection status on the relative size of these appendages
(One-way ANOVA, F
2,159
=2.099, P=0.126). An SMA
analysis of the Rønbjerg data demonstrated a small but
significant length reduction in the copulatory appendages
of infected individuals (2.2 % based on grand means)
(Fig. 3c; equality of slope: T=1.115, P=0.296; equality
of grand means: Wald test, W
1
=21.402, P\0.0005).
However, the morphology of the appendages seemed
unaffected.
Claw dimensions
In the combined Venø Bay, Bjørndrup and Lovns Broad
samples, claw height (Fig. 1g) was positively related to
carapace width in all investigated crab categories (Linear
regression, r
2
[0.638, P\0.0005). Corrected for cara-
pace width (i.e., claw:carapace ratio), relative claw height
was significantly affected by gender and male infection
status (Fig. 2b). Infected and uninfected females had sim-
ilar sized claws that were significantly smaller than those of
males irrespective of their infection status. Sacculinized
male crabs possessing an externa expressed a significantly
smaller claw:carapace ratio than non-infected males,
thereby approaching the claw dimensions of females. In an
analysis including also the female categories, scarred males
could not be statistically separated from either non-infected
males or males with externae (see Fig. 2b). But a separate
analysis including solely the three male categories suggests
that the relative claw size of scarred males is similar to
those with externae, but significantly smaller than those of
non-infected males (Kruskal–Wallis test, v
2
2
=22.392,
P\0.0005. Post hoc test: scarred versus externae,
P=1.000; scarred versus non-infected, P\0.0005).
Hence, infection by Sacculina in males results in a claw
size intermediate between non-infected males and females.
Pereopod length
In the Rønbjerg sample, the length of the longer third
pereopod (Fig. 1h) increased linearly with carapace width
for non-infected females, non-infected males and infected
males, albeit at a significantly lower rate in females than in
males irrespective of infection status of the latter (Fig. 3d;
equality of slope among groups: SMA, T=51.844,
P=0.0001; post hoc test: females versus male groups,
P=0.0001). Whereas the slope of infected and non-infec-
ted males was similar (post hoc test, P=0.604), infected
males generally attained shorter pereopods (8.1 % shorter
based on grand means) than non-infected males (Fig. 3d;
equality of grand means: SMA, Wald test, W
1
=28.76,
P\0.0005). Thus, as for the abdomen, the walking legs of
sacculinized males are morphologically feminized, but
without adopting the inherent female allometry.
Discussion
Modification of male crabs
Most male C. maenas parasitized by S. carcini showed a
modification of secondary sexual characters which can be
summarized as follows: a redivision of the abdomen, which
broadens to various degrees albeit rarely to an extent
comparable to that of healthy females (Figs. 2a, 3a); sac-
culinized males acquire marginal setae on the abdomen
(Fig. 1c); the length of the abdomen, as expressed by the
sixth segment (Fig. 3b), shows a slight increase which, in
companion with the increase in width, results in an overall
larger abdomen; a slight diminution of the copulatory
appendages (expressed by pleopod 1, see Fig. 3c); and no
additional pleopods appear. The claws show the highest
degree of modification, since they often tend to approxi-
mate those of normal females of same size (Fig. 2b).
Finally, the walking legs are reduced in length as expressed
by the third pereopods (Fig. 3d), thereby approaching the
relatively shorter female pereopods. These results largely
agree with those of Potts (1909), Foxon (1940) and Werner
(2001), who studied the same association but only con-
sidered the modification in abdominal shape, and also those
of Day (1935), who examined males of L. holsatus para-
sitized by S. carcini. As described by Veillet (1945),
S. carcini causes much more modification of males when
parasitizing C. aestuarii, a sibling species of C. maenas
found in the Mediterranean. These modifications comprise
a resegmentation of the abdomen, which may become
nearly as broad as in sexually mature females, acquisition
of pronounced marginal setae, and appearance of female-
like biramous and setae pleopods (swimmerets) that often
become asymmetrically developed. The copulatory
appendages become shorter and more fragile, and the
second pleopod disengages from the first and may acquire
segmentation and setae. The claws also diminish in size.
Nielsen (1970) similarly observed that the modification of
males caused by hermit crab infesting rhizocephalans may
vary with host species.
Even more extensive modification of sacculinized males
occur in other species of crabs parasitized by sacculinid
2020 Mar Biol (2012) 159:2015–2023
123

rhizocephalans. Infested males not only may acquire an
abdomen as broad and vaulted as in the female counterpart,
but may also show a varying number of pleopods, degen-
eration or disappearance of the copulatory appendages, and
reduction in the size of the chelae (Giard 1886; Smith
1910; Okada and Miyashita 1935; Matsumoto 1952;
George 1959; Galil and Lu
¨tzen 1995). The almost total
feminization of Portunus sanguinolentus (Herbst, 1783) by
Heterosaccus ruginosus Boschma, 1931, led Nair and
Gurumani (1956) and Srinivasagam (1982) to believe that
each of the 179 sacculinized crabs investigated were of the
female sex. This highlights the importance of taking par-
asitic feminization into account in population studies of
sacculinized crab populations.
Several studies have shown that the degree to which
male hosts become modified decreases with size (Pe
´rez
1933; Day 1935; Veillet 1945). The explanation given for
this is that the modifications occur stepwise and at each
moult and that smaller crabs on an average go through
more moults than larger ones before the cessation of
moulting caused by the parasite sets in. Our data support
this explanation because our analyses of covariance dis-
close that smaller sacculinized male crabs are relatively
more modified than larger males (Fig. 3). This agrees with
the observations made by Foxon (1940), but is contrary to
those of Potts (1909), two authors who studied the same
parasite-host association as we did. These covariance
analyses also suggest that the amount of modification in
exact metric terms to be similar in infected males regard-
less of their size.
Modification of female crabs
Parasitized females of C. maenas did not exhibit any
morphological modification. The mean relative abdominal
width and the claw size remain unchanged (Fig. 2). This is
concordant with most other studies showing that sacculi-
nized females undergo little, if any, modifications (Day
1935; Hartnoll 1967; Phillips and Cannon 1978; Yamag-
uchi and Aratake 1997; Werner 2001). Nevertheless, some
crabs, including C. aestuarii parasitized by S. carcini, may
be subject to major changes (Veillet 1945). Although such
cases are deceptively known as hyperfeminization, the
abdomen is usually of normal width and only rarely
excessively broadened as seen, for instance, in Callinectes
sapidus Rathbun, 1896, parasitized by Loxothylacus tex-
anus Boschma, 1933 (Reinhard 1950). Moreover, rather
than increasing in size, the pleopods may degenerate more
or less completely (Okada and Miyashita 1935; Matsumoto
1952; Galil and Lu
¨tzen 1995). In female crabs of some
species, the adult secondary sexual characters may appear
at a considerable smaller body size than usual as a result of
sacculinization (Smith 1906,1910;Pe
´rez 1933; Veillet
1945; Vernet-Cornubert 1958). In the very few cases where
the relative claw size in sacculinized females has been
analysed, it never differs from that of non-parasitized
females (Veillet 1945; Yanagimachi and Aratake 1997).
Survival of externae on modified crabs
Veillet (1945) and Lu
¨tzen (1984) showed that the ratio of
externally infected female crabs to that of the males
gradually increases with the age of the externae. In our
presently unpublished material, comprising more than
17,000 parasitized C. maenas from Limfjorden, the pro-
portion of infected male crabs with scars from lost externas
is 23.8 %, while the same proportion for female crabs is
only 14.6 %. Together, this suggests that females are better
hosts. The much broader abdomen of the female, particu-
larly in the larger individuals having relatively broader
abdomen, offers a better protection of the externa than the
narrower abdomen of males. Regeneration of a lost externa
occurs only rarely in S. carcini (Lu
¨tzen 1981). Externa loss
usually leads to the death of the parasite. This means that
the survival rate of the parasite is favoured by the devel-
opment of a broad female-like abdomen that protects
against accidental and loss of the externa. In the sacculinid
Heterosaccus dollfusi Boschma, 1960, the parasite induces
an abdomen as broad as that of healthy females of its host
Charybdis longicollis Leene, 1938. This results in an
equally high survival of the externae in both host sexes
(Galil and Lu
¨tzen 1995). This is born out by the fact that
the number of scarred crabs, each indicating a lost externa,
makes out only 2.3 % of the number of crabs with exter-
nae, a percentage that is much lower than in most other
species of sacculinized crabs (Veillet 1945;Lu
¨tzen 1984).
This probably means that most externae of H. dollfusi stay
with the host for life and are rarely lost. In Portunus
pelagicus L., 1758, sacculinized males develop an abdo-
men almost as broad as that of the females, and the per-
centage of scarred males is accordingly also low, \1%
(Weng 1987). Our data on S. carcini infesting C. maenas
show a more variable pattern. In contrast to the examples
mentioned above, the degree to which S. carcini modifies
the abdomen of a male host varies greatly from one host
specimen to another; from almost no change to close to
female proportions (Figs. 1,2,3). The data furthermore
demonstrate that crabs with scars from lost externae show
significantly less pronounced feminization of the abdomen
than found in males, still carrying viable externae (Fig. 2a).
This may reflect the presence of a co-evolutionary arms
race scenario between host and parasite: a parasite unable
to cause a male host to produce a protectively broadened
female-like abdomen is selected against. The host, on the
other hand, is exposed to strong selection pressure to pre-
vent the parasite-induced broadening of the abdomen. The
Mar Biol (2012) 159:2015–2023 2021
123

less morphological affected an infected male host, the
greater is the chance that the emerging externa is physi-
cally damaged and dies due to the poor protection provided
by a narrow male abdomen. The death of the parasite might
enable the crab to resume moulting and eventually repro-
duce. In contrast to the sacculinids on C. longicollis and
P. pelagicus, where this arms race seems already won by
the parasite, the S. carcini/C. maenas system may be at a
transitional evolutionary state, where full feminization of
the abdomen in male hosts has been only partly estab-
lished. As a result, the capability of each individual parasite
to modify the host still varies greatly within the population.
This calls for experimental studies on the extent to which
C. maenas that loose a S. carcini externa resume growth
and reproduction. Although scarred C. maenas have a
higher mortality rate than uninfected crabs, they may sur-
vive for an extended period of time (Lu
¨tzen 1981). The
additional parasitically induced morphological effects on
the male (reduced copulatory appendages, chelae and
pereopods) cannot easily be interpreted as advantageous to
parasite survival. They may simply be pleiotropic side
effects from the main adaptive value of broadening the
abdomen through hormonal manipulation. However, pos-
sibly also less understood physiological or behavioural
advantages of feminizing male hosts could be involved.
Phylogenetic remarks
Species of the family Sacculinidae Lilljeborg, 1860, exhibit
a number of surprisingly uniform morphological traits.
Attached to the underside of the brachyuran host’s abdo-
men, the left and right sides of the externa are flattened to
fit into the narrow space between the sternum and the
abdomen, with the dorsal sides and ventral sides always
directed towards the same sides with respect to the host’s
abdomen. Molecular data have nevertheless shown the
family and the type genus Sacculina Thompson, 1836, to
be diphyletic (Glenner and Hebsgaard 2006). To date, few
species of Sacculina have been relegated to the Sacculin-
idae s. str., which besides the type species, S. carcini,
probably contains all species of Heterosaccus Smith 1906
and Loxothylacus Boschma, 1928. No family name has yet
been proposed for the other group, which is known with
certainty to contain only a few species that were earlier
referred to as Sacculina plus all three species of Polyascus
Glenner, Lu
¨tzen and Takahashi, 2003.
When comparing morphological modifications caused
by sacculinid and non-sacculinid species of the ‘‘Saccu-
linidae’’, the general pattern is the same: males invariable
acquire a broadened and resegmentated abdomen, which
may or may not develop pleopods, and the copulatory
appendages may degenerate while the claws almost always
become reduced in size. The females, in contrast, are rarely
affected. The underlying physiological cause of the femi-
nization of parasitized males is still unknown. It has been
explained as a parasitic destruction of the androgenic
gland, which in healthy crabs induces masculinisation in
the males (Veillet and Graf 1958; Høeg 1995), but alter-
native physiological pathways might be possible. In any
case, it is most likely that the mechanism operating in both
groups of ‘sacculinids’ is identical.
Acknowledgments We are very grateful to the Carlsberg Founda-
tion for covering all expenses connected with the study (grant no.
2008-01-0491). We also wish to thank the staff of the Danish
Shellfish Center, Nykøbing Mors, Denmark, for collecting some of
the material and providing laboratory facilities, and Dr. David R.
Nash, Section for Social Evolution, Department of Biology, Univer-
sity of Copenhagen, Denmark, for advice and assistance.
References
Bulgurkov K (1938) Study of Rhizocephala and Bopyridae from the
Bulgarian Black Sea coast. Trudy na Chernorskata Biol Sta.
v’Varna 7:69–81
Day JH (1935) The life-history of Sacculina. Q J Microsc Sci
77:549–583
Foxon GEH (1940) Notes on the life history of Sacculina carcini
Thompson. J Mar Biol Assoc UK 24:253–264
Galil BS, Lu
¨tzen J (1995) Biological observations on Heterosaccus
dollfusi Boschma (Cirripedia: Rhizocephala), a parasite of
Charybdis longicollis Leene (Decapoda: brachyura), a lessepsian
migrant to the Mediterranean. J Crust Biol 15:659–670
George AI (1959) Heterosaccus ruginosus (Boschma), a rhizoceph-
alan parasite of the crab Neptunus sanguinolentus (Herbst).
J Zool Soc India 11:171–204
Giard A (1886) De l’influence de certains parasite
`s rhizoce
´phales sur
les caracte
`res sexuels exte
´rieures de leur ho
ˆte. C R Hebd Acad
Sci 103:84–86
Glenner H, Hebsgaard MB (2006) Phylogeny and evolution of life
history strategies of the parasitic barnacles (Crustacea, Cirripe-
dia, Rhizocephala). Mol Phylogenet Evol 41:528–538
Hartnoll RG (1967) The effects of sacculinid parasites on two
Jamaican crabs. J Linn Soc (Zool) 46:275–295
Høeg JT (1995) The biology and life cycle of the Rhizocephala
(Cirripedia). J Mar Biol Assoc UK 75:517–550
Høeg JT, Lu
¨tzen J (1985) Crustacea Rhizocephala.Marine invertebrates
of Scandinavia 6. Norwegian University Press, Oslo, pp 1–92
Lu
¨tzen J (1981) Field studies on regeneration in Sacculina carcini
Thompson (Crustacea: rhizocephala) in the Isefjord, Denmark.
J Exp Mar Biol Ecol 53:241–249
Lu
¨tzen J (1984) Growth, reproduction and life span in Sacculina
carcini Thompson (Cirripedia: rhizocephala) in the Isefjord,
Denmark. Sarsia 69:91–106
Matsumoto K (1952) On the sacculinization of Charybdis japonioca
(A. Milne-Edwards). Biol J Okayama Univ 1:84–89
Nair NB, Gurumani ON (1956) On the occurrence of three Sacculina
parasitising the edible crab Neptunus sanguinolentus. J Bombay
Nat Hist Soc 53:730–732
Nielsen S-O (1970) The effects of the rhizocephalan parasites
Peltogaster paguri Rathke and Gemmosaccus sulcatus (Lillje-
borg) on five species of paguridan hosts (Crustacea: decapoda).
Sarsia 42:17–32
Okada YK, Miyashita Y (1935) Sacculinization in Eriocheir japoni-
cus de Haan, with remarks on the occurrence of complete sex-
2022 Mar Biol (2012) 159:2015–2023
123

reversal in parasitized male crabs. Mem Col Sci Kyoto Imp Univ
B 10:169–208
Øksnebjerg B (2000) The Rhizocephala (Crustacea: cirripedia) of the
Mediterranean and Black Seas: taxonomy, biogeography, and
ecology. Israel J Zool 46:1–102
Pe
´rez C (1933) Action de la sacculine sur les caracte
`res sexuels
exte
´rieures du Pachygrapsus marmoratus. C R Soc Biol Paris
113:1027–1029
Phillips WJ, Cannon LRG (1978) Ecological observations on the
commercial sand crab, Portunus pelagicus (L.), and its parasite,
Sacculina granifera Boschma, 1973 (Cirripedia: rhizocephala).
J Fish Dis 1978:137–149
Potts FA (1909) Observations on the changes in the common shore-
crab caused by Sacculina. P Camb Philos Soc 15:96–100
Reinhard EG (1950) An analysis of the effects of a sacculinid
pasrasite on the external morphology of Callinectes sapidus
Rathbun. Biol Bull Wood’s Hole 98:255–288
Smith G (1906) Rhizocephala. Fauna Flora Golf Neapel 29:1–123
Smith G (1910) Studies in the experimental analysis of sex. Part 3.
Further observations on parasitic castration. Q J Microsc Sci
55:225–240
Srinivasagam S (1982) An instance of large scale parasitism of
Heterosaccus dollfusi Boschma on the edible crab, Portunus
sanguinoletus (Herbst) in Pulicat Lake. J Inland Fish Soc India
14:88–90
Veillet A (1945) Recherches sur le parasitisme des crabes et des
Galathe
´es par les Rhizoce
´phales et les e
´picarides. Ann I
Oce
´anogr Paris 22:193–341
Veillet A, Graf F (1958) De
´ge
´ne
´rescence de la glande androge
`ne des
Crustace
´sde
´capodes parasite
´s par les Rhizoce
´phales. Bull Mens
Soc Sci N 18:123–127
Vernet-Cornubert G (1958) Recherches sur la sexualite
´du crabe
Pachygrapsus marmoratus (Fabricius). Arch Zool Exp Ge
´n
96:101–276
Weng HT (1987) The parasite barnacle, Sacculina granifera
Boschma, affecting the commercial sand crab, Portunus pelag-
icus (L.), in populations from two different environments in
Queensland. J Fish Dis 10:221–227
Werner M (2001) Prevalence of the parasite Sacculina carcini
Thompson 1836 (Crustacea, Rhizocephala) on its host crab
Carcinus maenas (L.) on the west coast of Sweden. Ophelia
55:101–110
Yamaguchi T, Aratake H (1997) Morphological modifications caused
by Sacculina polygenea in Hemigrapsus sanguineus (De Haan)
(Brachyura: grapsidae). Crust Res 23:89–101
Mar Biol (2012) 159:2015–2023 2023
123