Ectosymbionts of the non‐indigenous Asian shore crab, Hemigrapsus sanguineus (Decapoda: Varunidae), in the western north Atlantic, and a search for its parasites

Over 560 Asian shore crabs, Hemigrapsus sanguineus, collected mainly in the spring of 2005 and 2006 from rocky intertidal locations in southern New Jersey, were examined for epibionts. Small numbers of the sympatric green crab, Carcinus maenas, and the Atlantic mud crab, Panopeus herbstii, were examined for epifauna to compare with H. sanguineus. Blue mussel spat, Mytilus edulis, and the encrusting, cheilostome bryozoan, Conopeum tenuissimum, were the dominant ectosymbionts of the shore crab, with prevalences of 22.2 and 32.1%, respectively; ranges of intensity were 1–146 spat/crab and 1–31 colonies/crab. Both species are incidental symbionts. Larger crabs had higher prevalences and intensities of C. tenuissimum colonies, and these colonies displayed a distinct pattern of attachment to the carapace which seemed to be related to crab morphology and habitat. Much less common was the encrusting, ctenostome bryozoan Alcyonidium albescens, a facultative symbiont species with a prevalence of 3.4%. Other epibionts were the encrusting, cheilostome Membranipora tenuis, the tubicolous polychaetes Hydroides dianthus, Sabellaria vulgaris, and Spirorbis sp., the barnacles Balanus improvisus and Semibalanus balanoides, and unidentified thecate hydroids, all incidental symbionts with prevalences from 0.2 to 3.2%. The total number of known ectosymbionts of H. sanguineus, including additional species found previously in the USA and the western Pacific, is 13. Carcinus maenas and P. herbstii share some of the same symbionts. Rhizocephalan externae were not observed in any of the crab species used in this study, nor were gill parasites or internal parasites found among 248 specimens of H. sanguineus.


Introduction
Hemigrapsus sanguineus (de Haan, 1853) was found for the first time in waters of eastern USA along the coast of New Jersey in 1988 (Williams and McDermott 1990;McDermott 1991). Since then it has been reported from Maine to North Carolina (McDermott 1995(McDermott , 1998a(McDermott , 1998b(McDermott , 1999(McDermott , 2000Lohrer and Whitlach 1997;Lohrer 2001;Seeley and McDonald 2003). The biology and life history of H. sanguineus in its native range in the sites of previous biological studies of H. sanguineus (McDermott 1998aH. sanguineus (McDermott , 1998bH. sanguineus (McDermott , 1999. At these locations monthly water temperature and salinity ranged from approximately 1 to 28uC and 29 to 32%, respectively; lowest temperatures were from December to February and the highest were from July to September (McDermott 1998a). Most of the collections were made in the spring when low water temperatures minimized moulting and consequent shedding of organisms attached to the exoskeleton; one other collection was made in the fall of each year. Crabs were preserved in 10% seawater formalin and later transferred to 70% ethyl alcohol. All crabs were sexed and their carapace widths (CW) were measured with vernier calipers to 0.1 mm; all crab measurements, unless otherwise noted, refer to this parameter. Crabs were examined dorsally and ventrally for the presence of epibionts, including the cephalothorax, sternum, abdomen (dorsal and ventral), and the five pairs of pereopods (chelipeds and walking legs). Small symbionts (e.g. zooids of bryozoan colony, minute barnacles) were measured to 0.01 mm with a calibrated ocular micrometer. Symbionts were identified to the lowest taxon. The terms prevalence and intensity, as defined by Bush et al. (1997), were used to describe the populations of symbionts on H. sanguineus. Two other sympatric species of brachyurans, Carcinus maenas (Portunidae) and Panopeus herbstii (Xanthidae), were collected and their epifauna compared with that of the Asian crabs. Crabs used for photography were dried at room temperature in order to produce better contrast between attached symbionts, such as calcareous bryozoans, and the crab exoskeleton.
All specimens of H. sanguineus and the sympatric brachyurans collected during this twoyear study were examined for the externae of rhizocephalans, and about half of the Asian crabs were examined for macroscopic parasites in the gill chambers and the organs of the haemocoel. Table I lists the numbers of male and female Asian shore crabs collected on each date and their combined mean CW and range. The male/female sex ratio was 1.06. In the spring of 2005 (late March to early May) only four species of symbionts were found among 120 crabs examined (Table II). The 75% prevalence of Mytilus edulis spat in March coincided with the usual, large, recruitment of young mussels on rocks and shells in the immediate area. Mussel intensity on the crabs was 4.7¡5.1 mussels/crab (n575, range 1-16). Mean length of 65 mussels removed from crabs was 1.00¡0.23 mm (0.39-1.60 mm), which corresponded to the 1.10¡0.34 mm (0.74-2.62 mm; n552) length of mussels in the surrounding habitat. Prevalence of mussel spat decreased to 24.3% in April (intensity 0.7¡1.9, n550) and 11.8% in May (intensity 0.2¡0.6, n58). Mean length of 41 mussels attached to crabs in April was 1.34¡0.55 mm (0.7-3.5 mm); the two largest mussels, 2.8 and 3.3 mm, were from under the abdomen of a female crab. Half of the mussels in the May collection were still recent recruits, i.e. ,1.0 mm long. A slightly gaping abdomen of an 18.0 mm male was invaded by three mussels. All parts of the exoskeleton of crabs, but mainly the bases of the walking legs (incurrent areas to the branchial chambers) and sockets of the antennae and eyes, were sites for mussel attachment. No young Mytilus spat occurred among the 54 crabs examined in October 2005.

Prevalence and diversity of symbionts
The encrusting, calcareous bryozoan Conopeum tenuissimum was not found in March but was prominent in April and May (prevalence 17.3%, 18 of 104 crabs) (Table II). This bryozoan was the most common and conspicuous symbiont in the entire study, its colonies being attached to most regions of the crab's exoskeleton ( Figure 1A). No colonies were present in October. The distribution and intensity of C. tenuissimum on H. sanguineus will be presented below.
Whereas there was heavy recruitment of Semibalanus balanoides intertidally in March 2005 (J. J. McDermott, unpublished data), only six of 120 crabs (5.0%) collected that spring harboured individuals of this species. Fifteen barnacles were recorded; mean carinorostral length of 13 individuals was 2.45¡1.08 mm (1.21-4.42 mm); the other two barnacles had just metamorphosed, and were less than 1.0 mm. Barnacles attached locally to shells and rocks had a mean length of 1.21¡0.34 mm (0.82-2.21 mm, n560), and cyprid larvae were seen undergoing metamorphosis or searching for favourable sites on the substrata. In October of 2005, 18.5% of the crabs harboured recently settled colonies of the encrusting, calcareous bryozoan Membranipora tenuis Desor, 1848. In addition, an unidentified thecate hydroid and three tubicolous polychaetes, Hydroides dianthus (Verrill, 1873), Sabellaria vulgaris (Verrill, 1873), and Spirorbis sp., were noted. Whereas spirorbids were found on the carapace, subhepatic area, and pereopods of 11 crabs (20.4%), a total of only 20 young worms was recorded (most were ,1 mm diameter; maximum 1.5 mm); one crab had six worms. Recently metamorphosed specimens of Sabellaria were present on five crabs, one of which (35.8 mm male) had 35 worms on the carapace, subhepatic surface, chelipeds, and walking legs.
The diversity of species on Hemigrapsus sanguineus in the spring of 2006 (early March to early June) was about the same as in 2005 (Table II). Attached mussels were common during late March and April 2006 (Table II), again coinciding with recent local recruitment. Prevalence in March was 2.5 times that in April and only one crab had mussels in June. There was no significant difference in the prevalence of M. edulis between males and females in the 29 March collection (x 2 50.012, P.0.05, df51; 40 of 102 males showing Conopeum tenuissimum colony spanning the left hepatic-branchial region of the carapace, a small medial colony and a colony in the right hepatic region, plus numerous colonies on the pereopods. (C) Male (18.5 mm CW) with a single colony of Conopeum tenuissimum covering more than half of the carapace, with some smaller colonies on the pereopods. (D) Female (24.9 mm CW; left cheliped missing) with four colonies of Conopeum tenuissimum on the carapace, two in hepatic and two in branchial areas. Barnacle is Semibalanus balanoides. and 41 of 112 females). Mean length of mussels taken from crabs in March was 0.90¡0.39 mm (0.40-2.68 mm, n594). Of the 731 mussels removed from all crabs, 85% were from the March collection. The mean intensities of mussels attached to crabs in March and April were 8.0¡18.1 (644 mussels on 81 crabs, range 1-146) and 8.5¡9.6 (85 mussels on 10 crabs, range 1-26), respectively. The crab with 146 mussels was a 30.3 mm male collected in March; these minute mussels had no apparent effect on the condition of the crab. Mussels attached not only to the bases of the walking legs and depressions such as the sockets occupied by antennae and eyes, but also to the edges or under the female abdomen and often attached to the pleopods. One female crab (26.1 mm) had an abnormally gaping abdomen due to the presence of several relatively large mussels whose byssus threads entwined and immobilized many of the pleopods (mean mussel length52.2¡1.8 mm, range 0.6-9.5 mm, n528) ( Figure 2). Stumps of autotomized pereopods were often sites for spat attachment.
Prevalence of Conopeum tenuissimum colonies in March and April was nearly equal (55.6 and 55.2%, respectively), but was greatly diminished in June. The gelatinous, encrusting bryozoan Alcyonidium albescens was present on 6.8% of the 281 crabs collected in March and April (Table II), occurring on the carapace, subbranchial surfaces, pereopod segments, and stumps of autotomized pereopods ( Figure 1A). The intensity of the A. albescens was 1.6¡1.3, range 1-6. Adjacent colonies of Conopeum were partially overgrown by four of 29 Alcyonidium colonies.
Only 2.0% of the 344 crabs examined from the spring collections of 2006 had spirorbids attached. Eleven worm tubes were 0.86¡0.26 mm in diameter (0.60-1.15 mm). These worms were found on pereopods and in subhepatic areas. Unidentified thecate hydroids (short, dead stalks) were present on 2.3% of these crabs.

Host distribution and intensity of Conopeum tenuissimum
This bryozoan was found on Hemigrapsus sanguineus in all spring collections except that of 30 March 2005 (Table II). Colonies were more frequent on larger (i.e. older) crabs ( Figure 3A). There were no significant differences in the prevalence of Conopeum tenuissimum between male and female crabs in the total spring collection of 2006  Figure 1B) with a gaping abdomen caused by the accumulation of young blue mussels Mytilus edulis attached to the pleopods (white arrow shows largest mussel, 9.5 mm long). Colonies of the calcareous bryozoan Conopeum tenuissimum present on many of the pereopods. Black arrow points to the polychaete Spirorbis sp. On 29 March 2006 the mean intensity of Conopeum tenuissimum was 5.6¡6.0 colonies/ crab (660 colonies on 119 crabs, range 1-29 colonies). The 29 colonies on a 27.4 mm female were distributed on the pereopods (21), carapace (7), and abdomen (1), and the host also harboured five colonies of Alcyonidium albescens on the pereopods and one on the carapace. Intensity of C. tenuissimum in crabs on 29 March showed a significant positive correlation with crab CW ( Figure 3B). Mean intensity in April was 7.8¡9.4 colonies/crab (290 colonies on 37 crabs, range 1-31 colonies). The relationship between intensity and CW was also significant for crabs in the April collection (regression equation: y5213.328+0.989 x; r 2 50.377, P(0.0001, n537). There was a significantly greater intensity of Conopeum colonies (in March 2006) on the carapace of female crabs compared to males (x 2 525.2, P,0.001, df51; 87 on 42 males and 167 on 64 females).

Location of Conopeum tenuissimum colonies
The dorsal carapace of Hemigrapsus sanguineus was the major site for attachment of colonies to the cephalothorax of crabs ( Figure 1B). Many colonies that began their growth on the lateral parts of the dorsal carapace often continued development ventrally on to the subhepatic and subbranchial regions of the cephalothorax. Some subhepatic and subbranchial colonies, however, originated in these locations. Conopeum tenuissimum was seldom found on the sternum, abdomen, or the third maxillipeds, but chelipeds and walking legs were common sites for attachment; some colonies completely covered the individual segments of these appendages. Colonies ranged from those recently settled (1-10 zooids) to large colonies with hundreds of zooids, some covering one-half of the carapace ( Figure 1C).
An analysis of the number of Conopeum tenuissimum colonies at different locations on the dorsal carapace of Hemigrapsus sanguineus showed that the colonies occurred primarily on the hepatic and branchial areas-basically at the four corners of the carapace (Table III; Figures 1D,4). In many cases the colonies at each of these locations coalesced to produce a lateral band from the anterior to the posterior end of the carapace ( Figure 1B). Colonies were relatively fewer and less developed on mid-carapace locations (i.e. frontal, protogastric, cardiac, etc.). Among the 119 crabs infested with Conopeum tenuissimum from the 29 March 2006 collection, there were 254 colonies on the dorsal carapace and 406 on the chelipeds and walking legs for a total of 660 (intensity 55.6¡6.0). Colonies on the ventral carapace were not included in this tabulation.

Symbionts on sympatric brachyurans
Green crabs (Carcinus maenas) and mud crabs (Panopeus herbstii) were examined for ectosymbionts. Eighty-six green crabs were collected in spring and fall of 2005 and 2006  , 1870), and the bryozoans Anguinella sp. and a ''Triticella-like'' sp. were found only with green crabs, all of which were of low prevalence (1.2-3.5%) and intensity. All surfaces of green crabs, particularly the legs, were prone to the development of Conopeum and Alcyonidium colonies, and colonies of the latter tended to overgrow those of Conopeum. Barnacle and polychaete larvae tended to metamorphose in the grooves of the green crab's carapace ( Figure 5A).

Parasites
Externae of rhizocephalans were not found on any of the crabs in this study. Two hundred and forty-eight specimens of H. sanguineus were examined for gill and internal parasites  (Table III)  (136 males, 112 females (11 ovigerous); mean CW514.4¡4.3 mm, range55.3-28.2 mm). No bopyrid or entoniscid isopods, nemertean egg-predators (Carcinonemertes spp.), or metacercarial cysts of digenetic trematodes were present. A 23.0 mm male with carapace damage and gill damage on the left side harboured two Mytilus spat in its gill chamber.

Diversity, prevalence, and intensity
In the present study, 10 species of ectosymbionts were recorded from Hemigrapsus sanguineus (Table II)-a bivalve mollusc, three species of encrusting bryozoans, three species of polychaetes, two barnacles, and an unidentified hydroid(s). The occurrence of attached hydroids refers to possibly more than one species, but all were so damaged (consisting of short dead stalks) that specific identification was not possible. Alcyonidium albescens was listed as A. polyoum (Hassall, 1841) in my previous biological studies of H. sanguineus (McDermott 1998a), but Winston and Key (1999) determined that A. polyoum was of European origin, and thus created the new species A. albescens. Five of the 10 ectosymbionts found in the present study (Mytilus, Balanus, Semibalanus, Alcyonidium, Conopeum) were noted in earlier studies (McDermott 1998a; J. J. McDermott, unpublished data), but hydroids and the polychaete worms, Hydroides, Sabellaria, Spirorbis, were not noted previously. The barnacle Chelonibia patula was found only in the earlier studies; it occurred on the carapace of the largest H. sanguineus ever recorded, a 43.9 mm male that had likely reached terminal ecdysis (McDermott 1998a(McDermott , 1999. Thus, the Asian shore crab is host to 11 species of ectosymbionts in the waters of the USA. Worldwide, with addition of the bivalves Hiatella arctica and Mytilus trossulus from the western Pacific, attached only to crabs with externae of the rhizocephalan Polyascus polygena (Isaeva et al. 2001), the total is 13 species. Relationships of all known ectosymbionts from H. sanguineus are considered incidental (see Williams and McDermott 2004), except for the crab's facultative relationship with C. patula, a species known to have a predilection for attaching to the exoskeletons of a variety of brachyuran crabs and horseshoe crabs (Limulidae), but also found sometimes to be free-living or on gastropods, sea snakes, and sea turtles (Key et al. 1997).
There are 13 recognized species within the genus Hemigrapsus (McLay and Schubart 2004; Asakura and Watanabe 2005), but ectosymbionts have been reported only from H. sanguineus. Epibionts have not been recorded in numerous studies of intertidal members of the Grapsoidea sympatric with H. sanguineus from the western Pacific (e.g. Kikuchi et al. 1981;Fukui 1988;Lohrer et al. 2000a).
Information in Table II shows clearly that Mytilus edulis and Conopeum tenuissimum were the most prevalent species among the 564 crabs examined in this two-year study (22.2 and 32.1%, respectively). Alcyonidium albescens had a prevalence of 3.4%, although not found in any of the 174 crabs examined in 2005. It was not surprising that the two basically freeliving species of bryozoans were found on the non-indigenous H. sanguineus, because they have been recorded frequently from other benthic brachyurans and horseshoe crabs [Limulus polyphemus (Linnaeus, 1758)] (Watts 1957;Key et al. 1996Key et al. , 1999J. J. McDermott, personal observations). The Asian horseshoe crab, Tachypleus gigas (Mü ller, 1819), has a high prevalence of similar, encrusting, cheilostome bryozoans (Key et al. 1996). The presence of bryozoans on the sympatric green crabs and mud crabs mentioned here emphasizes that H. sanguineus is not a unique arthropod host. Colonies of Membranipora found only in the fall collections of 2005 and 2006 were all recently settled and of low intensity. The other ectosymbiotic species (barnacles, polychaetes, and hydroids) covered only minute areas on the exoskeleton of H. sanguineus.
It was also not surprising that Mytilus spat were so prevalent on H. sanguineus and C. maenas in the spring because most other hard or soft (algae) surfaces in the intertidal zone were heavily covered with these newly attached mussels. There they remain attached and grow rapidly, unlike their relatively ephemeral existence on the exoskeleton of crabs, especially young crabs with their short intermoult periods. Bryozoans and all the other attached organisms suffer the same fate. This was the reason for collecting crabs in the spring for evaluating the ectosymbiont load, when moulting was expected to be minimal, thus preserving organisms that may have accumulated on crabs during their lower intertidal or subtidal existence during the colder months of the year. More year-round data are needed to substantiate this point, but the present information (Table II) and previous observations with large numbers of crabs suggest that the spring affords a more fruitful period for studying the crab's epibionts. The present study suggests that annual differences in the prevalence of ectosymbionts is probably due, at least in part, to the size of the existing free-living populations and the many environmental factors related to their recruitment.
Rhizocephalan barnacles are known usually to prevent moulting or prolong normal intermoult periods of infected crabs. Hemigrapsus sanguineus is host for the rhizocephalan Polyascus polygenea, originally described as Sacculina polygenea by Lü tzen and  with a later generic change to Polyascus by Glenner et al. (2003). Isaeva et al. (2001) found ectosymbionts only on sacculinized H. sanguineus collected at Vistok Bay, Sea of Japan. To date, however, externae of P. polygenea have not been found in crabs collected in the Atlantic Ocean in previous studies (McDermott 1998a;Torchin et al. 2001) or in the 564 crabs examined in the present study. Abello and Corbera (1996) showed that infestations with the bryozoan Triticella flava Dalyell, 1848 were significantly greater in Goneplax rhomboides (Linnaeus, 1758) infected with an unidentified parasitic barnacle. That rhizocephalan infections can be an unequivocal factor in increasing the prevalence and intensity of epibionts in Carcinus maenas from Danish waters was shown by Mouritsen and Jensen (2006). In .1100 crabs examined, 75% of sacculinized crabs (Sacculina carcini Thompson, 1836) had a variety of epibionts compared to only 29% of the uninfected animals. The normal burying response of C. maenas was reduced by more than half in infected crabs in the laboratory, which would make the exposed crabs more vulnerable to epibionts. Such an effect on behaviour was shown also in another portunid crab, Charybdis longicollis Leene, 1938, infected with the rhizocephalan Heterosaccus dollfusi Boschma, 1960 (Innocenti et al. 1998).
It was expected that ectosymbionts would have a greater prevalence and intensity in larger Asian crabs because of their longer intermoult intervals and possibly due to their greater size. Figure 3A, B shows this relationship to be true for the prevalence and intensity of Conopeum tenuissimum colonies. Such relationships have been demonstrated for the epifauna of other species of brachyurans, e.g. the deep-water portunids Bathynectes superbus (Costa, 1853) (Lewis 1976) and B. piperitus Manning andHolthuis, 1981 (Abello et al. 1990), the tanner crab Chionocetes bairdi Rathbun, 1924 in Alaska (Dick et al. 1998), the cancroid crabs Cancer gracilis Dana, 1852, C. magister Dana, 1852, and C. productus Randall, 1839(McGaw 2006, and the New Zealand portunid Ovalipes catharus White, 1843 (Miller et al. 2006). Abello and Corbera (1996), however, found that juvenile males of Goneplax rhomboides had lower prevalences of Triticella flava than mature crabs, but that there was no difference in the female population. On the other hand, in the blue crab, Callinectes sapidus Rathbun, 1896, Key et al. (1997Key et al. ( , 1999 showed that Chelonibia patula and bryozoan epifauna were not more prevalent on older crabs. In both of these studies, however, unlike the Asian crabs examined in the present study, no juvenile crabs were used in their analyses. How do the prevalences and intensities of ectosymbionts associated with Hemigrapsus sanguineus compare with those observed in other brachyurans? A variety of factors need to be considered in making interspecific comparisons. Most other brachyurans whose epibionts have been studied are strictly aquatic, whereas the Asian crab is mainly intertidal and exposed to the air for substantial periods during the tidal cycle and for considerable periods during the whole year. This difference should favour larger ectosymbiont loads in the aquatic species. Carcinus maenas is basically an aquatic species, but some crabs spend periods in the lower intertidal usually under the cover of rocks (often used as a refuge for moulting; personal observations). The present study showed that the prevalence and intensity of epibionts in the larger green crab were noticeably greater than in the Asian crab. Panopeus herbstii is aquatic, but may be found periodically in burrows in the lower rocky intertidal often covered with mud. The information presented here for Panopeus is inadequate for any quantitative comparison with the Asian shore crab.
Recently Miller et al. (2006) compared the epifauna on two portunid crabs from New Zealand, the non-indigenous species Charybdis japonica (A. Milne-Edwards, 1861) and the native Ovalipes catharus. Nearly 100% of the native species were infested with the obligate bryozoan epibiont Triticella capsularis Gordon and Wear, 1999, a species not found on the alien crab. Charybdis japonica, however, was host only for serpulid polychaetes and balanomorph barnacles. Substrate preferences and differences in burying behaviour of the crabs seemed to be responsible for this disparity. Whether innate differences between the two portunids are in some way responsible for susceptibility to attachment of the obligate T. capsularis is not known.
Although there were no significant differences in the prevalence of Conopeum tenuissimum between male and female Asian crabs, the intensity of this bryozoan was significantly greater in females. This difference may be related to the overall longer intermoult periods in females, associated with reproduction. Key et al. (1999) found, for an unexplained reason, that only female blue crabs harboured bryozoans. Perhaps the higher salinity water occupied by females compared to males may be involved in this difference.
Another factor to consider in determining prevalence of epibionts is the effect of the particular method of collection (e.g. trawls, baited traps, gill nets, hand collecting) on the age distribution of the host population. In some cases only mature crabs are recovered (Paul and Paul 1986;Key et al. 1997Key et al. , 1999Dick et al. 1998;McGaw 2006;Savoie et al. 2007) while others may include both immature and mature specimens (Lewis 1976;Abello et al. 1990). Lower prevalence values are expected with greater numbers of immature crabs. As in the present study, prevalence of ectosymbionts may vary seasonally, especially in relation to peak ecdysial periods when all attached epibionts are shed with exoskeletons. Crabs collected during periods of prolonged anecdysis or those species known to have a terminal ecdysis are likely to yield large prevalence and intensity values (Lewis 1976;Abello et al. 1990;Savoie et al. 2007).
It is to be understood that the prevalence and intensity values reported here for the epibionts of Hemigrapsus sanguineus are minimal because of missing pereopods due to autotomy. For interspecific comparisons of the diversity and prevalence of brachyuran symbionts, concentration only on the carapace and ventral aspects of crabs may yield more comparable data ).

Distribution of bryozoan colonies
Distinct distributional patterns of epibionts living on a number of brachyurans and horseshoe crabs have been reported (Abello et al. 1990;Abello and Corbera 1996;Key et al. 1997Key et al. , 2000Dick et al. 1998;Dietl et al. 2000;McGaw 2006). Conopeum colonies on the carapace of Hemigrapsus sanguineus are more frequently attached laterally, often extending to subhepatic and subbranchial locations ( Figures 1B, D, 4). This crab lives mainly in narrow crevices in rocky intertidal areas where its carapace comes in contact with the overlying rocks (Fukui 1988;McDermott 1998a). Since the carapace slopes laterally, the medial regions of the carapace have greater contact with the rocks, which likely minimizes the attachment and growth of bryozoan colonies. Scratch marks on the carapace are more conspicuous medially. Bryozoan larvae that metamorphosed medially in the grooves separating the distinctive regions of the carapace usually produced only small colonies in these locations. The chelipeds and walking legs were prime sites for the attachment of the bryozoan colonies, but they were less common on the two distal segments (propodus and dactylus), which are in greater contact with the substratum than the more proximal segments.
Colonies of Alcyonidium albescens were less common on the dorsal carapace of H. sanguineus, occurring more frequently on the ventral aspects of the carapace and locomotory pereopods and chelipeds. Key et al. (1999) reported that A. albescens preferred settling on lateral portions of the dorsal carapace of Callinectes sapidus, but that colonies also attached ventrally. The tendency of A. albescens colonies to overgrow those of C. tenuissimum requires further study, but such interspecific dominance is common among encrusting species of bryozoans (Jackson 1979;McKinney and Jackson 1989).
The nudibranch Corambe obscura is a known predator of Alcyonidium albescens colonies that encrust the gastropod shells harbouring hermit crabs (McDermott 2001), colonies of Conopeum tenuissimum (Franz 1967;Dudley 1973), and other bryozoan species (Franz 1967). This nudibranch was found with some of the sympatric green crabs infested with A. albescens and C. tenuissimum but not with H. sanguineus. It seems likely, however, that the larvae of C. obscura may at times attach and metamorphose on shore crabs bearing these bryozoans.
What are the costs and benefits of the Hemigrapsus sanguineus-ectosymbiont relationships described? It is unlikely that any of the epibionts described here in any way benefit the crab host. Perhaps the bryozoan colonies might be considered to have some camouflage effect, but this would appear to be unnecessary for the survival of animals that lead a secretive existence in their rocky habitat. The ecdysis-prone exoskeleton substrate of crabs does not usually allow for the accumulation of a heavy epibiont growth that would provide a more useful disguise for avoiding predation. Older brachyurans and those in terminal anecdysis are more likely to have heavier growths of epibionts, but in most cases do not accumulate the heavy growths of symbionts seen on the appropriated shells of many hermit crabs (Stachowitsch 1980;Williams and McDermott 2004). Are there any advantages gained by the ectosymbionts of Hemigrapsus sanguineus? Almost all of them are basically free-living species that usually have more than ample substrates to which they may attach, compared to the relatively minute surface areas provided by the ephemeral body surface of shore crabs. However, the exoskeleton of crabs in terminal anecdysis may be a protective refuge from predators and sedimentation and provide time for the completion of their life cycles.
What are the possible adverse effects of these symbionts on Hemigrapsus sanguineus? It appears that only Mytilus edulis has the potential for producing harmful effects on this crab. Mussels attached to the underside of the abdomen of older females in anecdysis may grow sufficiently to cause gaping, entanglement of the pleopods with byssus threads, and interference with normal attachment and development of embryos ( Figure 2). Although H. sanguineus, in the laboratory, is known to produce viable, successive broods from previously stored sperm (McDermott 1998b), normal copulation in females would be inhibited in cases of heavy mussel infestation.

Parasitism in Hemigrapsus sanguineus
The Asian shore crab has been a member of rocky intertidal communities in the northwestern Atlantic for at least 20 years (McDermott 1998a), but there is still no evidence of parasitism. As with several other introduced species of brachyurans, H. sanguineus remains released from its native parasite load, i.e. the rhizocephalan Polyascus polygenea and the trematode metacercaria Maritrema setoensis (Bridgman 1971;McDermott 1998a, present study;Torchin et al. 2001). Charybdis japonica, introduced into New Zealand from the northwestern Pacific, is without its native rhizocephalan Heterosaccus papillosus Boschma, 1933 (Miller et al. 2006), whereas C. longicollis, a Lessepsian migrant to the eastern Mediterranean from the Red Sea, brought with it H. dollfusi (Galil and Lü tzen 1995;Innocenti et al. 1998). It remains to be seen if H. sanguineus in the Atlantic will ever become reassociated with its native parasites or modify its resistance to similar parasites in its new environment.