The Biology of the Mussel Crab, Fabia subquadrata, from the Waters of the San Juan Archipelago, Washington
JACK B. PEARCE!
THE MUSSEL CRAB, Fabia subquadrata, described sults of the present study indicate that F. sub by Dana in 1851 from material collected in quadrata passes through stages comparable to Puger Sound, is placed in the subfamily Pinno those reported for P. pisum by Atkins and P. therinae Miln e-Edwards, one of the two sub ostreum by Stauber and by Christensen and Me families of the Pinnotheridae indigenous to the Dermott (1958: 150) . There are, however, im Americas. portant differences in the developmental cycle Most species of pinnotherids live in associa of P. subquadrata. tion with a host organism. Th e relationships Th e present known range of the mussel crab have been variously described as parasitism, is from the coast of Alaska to southern Cali commensalism, or mutualism. Although the fornia. Rathbun ( 1918:102 ) noted it in waters mussel crab is usually found in association with 250 m deep and W ells (1940 :47) found it in the horse mussel, Modiolus modiolus, several mussels dredged at a depth of 220 m.Hart other pelecypod hosts as well as a tunicate have (personal communication) has found it in M. been reported (W ells, 1928:289). Th e present modiolus taken int ertid ally near Victoria, Van research has revealed additional bivalve host couver Island, British Columbia. species. It is frequently reported from both spe While it is true that pinnotherid crabs have cies of MytiltIS in waters sout h of Puget Sound, been known from ancient times, only recently although in the latter waters it does not appear have there been any comprehensive studies of to frequent these hosts. any membe r of this family. Atkin's early ob The life cycle of F. subquadrata is complex servations (1 926 ) on the moul ting stages of and, as with many pinnorherids, includes several P. pisum laid the groundwork for future studies. developmental forms subsequent to the typical Thi s was followed by Hart's investigations decapod larval stages (i .e., the zoea and mega ( 1935) in which she reported success in hatch ing the eggs of Pinn oth eres taylori and rearin g lops) and before the definitive adult stage is them through the first true crab stage. Sandoz reached. As in most marine decapods, the early and H opkins (1947:250) were able to rear pinnotherid zoea and megalops are planktonic. P. ostreum to this same stage. Th ese investiga Upon moulting from the megalops into the first tions extended the earlier work of Atkins, in true crab stage the animal, it is thought, leaves which the hard and posthard stages subsequent the plankton and becomes associated with its to the first crab stage had been described. At host. kins ( 1955) later raised two species of Brit Posrplankronic developmental stages of a pin ish pinnotherids, P. pisum and P. pinnotheres, norherid were first described by Atkins ( 1926: through the megalops stage. 475) for Pinnotheres pisum, which is common Most of the workers cited above were con to the coast of the British Isles. Later Stauber cerned largely with the early developm ent of the (1 945:269) found that the developmental cycle crabs rath er than with their ecology or associa of the North American east coast pinnotherid, tion with the hosts. Wells ' studies ( 1928, 1940 ) Pinnotheres ostreum, was very much the same as were among the first published papers concerned that previ ously described for P. pisum . The re- with the biology of Ameri can species of pinno therids. Later Stauber ( 1945) investigated the 1 Department of Zoology, University of W ashing postlarval development and habits of the oyster ton, Seattle 5. Present add ress : Marine Laboratory, Humboldt State College, Arcata, California 9552 1. crab, P. ostreum. This work was followed by that Manuscript received June 25, 1964. of Christensen and McDermott (1 958) which
3 4 PACIFIC SCIENCE, Vol. XX, January 1966 represents the most comprehensive study of this versity of Washington, Friday Harbor, San Juan or any other species of the Pinnotheridae. Island, Washington. Observations were made on Except for purely taxonomic studies, these the contents of host mussels collected at least papers are the main reports concerning the pin once but frequently twice per month throughout notherids despite Rathbun's (1918: 10) early the period of study. In addition, materials were admonishment concerning the lack of knowl collected once a week during the summer edge of this family and the inherent rewards to of June, July, and August of 1958 and 1959. A be found in its study. total of 3,480 host mussels were examined dur Other than Wells' data (1928, 1940) there ing this period. have been no extensive reports concerning rhe The mussels were collected by dredging in biology of F. subquadrata. It is, therefore, one localities where they are known to occur. The of the least studied species of the Pinnotheridae, dredging gear included either a standard rock Until Wells' work of 1928 the male of the spe dredge or beam trawl, depending upon the type cies had not been recognized and was, in fact, of bottom from which the mussels were to be described as a separate species in a different removed. genus, Pinnotheres concharum. Several areas within the San Juan Archi The present paper is concerned with the bi pelago, selected as dredging sites, were chosen ology of this neglected species. The principal as being representative of a variety of depths study was conducted over a period of one and and bottom types. The deepest stations were one-half years, from June, 1958 to January, located in President Channel northwest of Orcas 1960, but many observations made subsequent Island (48 °39'45"N, 123°1'W), where the wa to the main investigation have been incorpo ter is 195 m in depth. The shallowest station is rated in this paper. Information on the develop off Point Lawrence, Orcas Island (48 °39'30"N, mental cycle, reproductive biology, relationship 122°44'45"W), where the water is 22-30 m to the host organism, distribution and size in in depth. Other stations were located near Point relation to water depth, and ecdysis is reported Caution, San Juan Island (48°34'N, 123°0'48" here. W) in water 130 m in depth; off Mineral Point, The author is especially indebted to Dr. Dixy San Juan Island (48°35'10"N, 123°3'35" W) 1. Ray, as well as to Dr. Robert Fernald, Dr. in waters 55 and 130 m deep; and in East Pea Ernst Florey, and Dr. Paul Illg, whose valuable Vine Pass (48°35'30"N, 122°47'30"W) in 48 assistance and helpful criticisms were most use m of water. ful in the preparation of this paper. The many Upon being brought to the surface the mus valuable suggestions of Dr. J. F. 1. Hart of Vic sels were immediately placed in live boxes with toria, British Columbia, and the critical reading circulating sea water. The drains of these boxes of the manuscript by A. M. Christensen are also are covered with screening of a gauge sufficient acknowledged. to insure that any swimming stage crabs would Finally, I should like to acknowledge the sum be retained should they leave their hosts. The mer cooperative fellowship provided me by the mussels were then brought into the laboratory National Science Foundation in 1959, the Na where they were opened and examined alive for tional Institutes of Health predoctoral fellow the presence of crabs. In those mussels that were ship (GF 10,872) awarded me during the years infested, any damage which may have occurred 1960 through 1962, and the NIH postdoctoral as a result of a crab's presence was noted. A fellowship (GPD-10, 872-C3 ) given for study dissecting microscope was always used in these at the University of Copenhagen's Marine Lab examinations. Each mussel was measured the larger ones (greater than 10 mm in le;gth) oratory, Holsinger, Denmark. with a vernier caliper, the smaller with a dial caliper. The larger mussels were measured to MATERIALS AND METHODS the nearest 0.5 mm, the smaller (less than 10 All field work involved in this study was car mm) to 0.1 mm. All crabs collected after July ried out at the Marine Laboratory of the Uni- 15, 1958 were measured with an ocular microm- Biology of Fabia subquadrata-PEARCE 5 eter to the nearest 0.01 mm. The greatest width stages a very anomalous instar appears. First de of both carapace and abdomen were noted. scribed by Atkins (1926:478) for P. pisum as The crabs were then placed in standard house the Stage I crab, this instar is, in its morphology hold polyethylene ice cube trays. Each tray con and behavior, entirely different from any of the sists of 14 cubicles and 1 crab was held in each stages preceding or following it. The exoskele of these. A "vaporize" pen was used to number ton is well calcified and very hard. It is, in many each cubicle with the crab's respective catalog species, highly pigmented with definite patterns number. In this manner several hundred crabs on the carapace. Above all it is highly modified could be retained, facilitating observations on for a temporary, freeswimming planktonic ex their behavior, ecdysis, and subsequent changes. istence. The setal ornamentation found on the The crabs were kept at temperatures approxi pereiopods is extensive and, in addition, these mating those of their natural environment, and appendages are broad and flattened in contrast either a flow of water from the sea water system to the cylindrical condition noted in the pre or several daily changes were used to maintain hard instars. They thus serve as very effective adequate environmental conditions. swimming appendages. It has been reported for P. pisum (Atkins, 1926 :475) and P. ostreum DEVELOPMENTAL STAGES IN F. subquadrata (Stauber, 1945 :272; Christensen and McDer morr, 1958: 152 ) that at this stage of develop As previously noted the life cycle of F. sub ment the males leave their host to seek out quadrata is similar to that described for P. pisum females, copulating with them in their host. (Atkins, 1926:475) and P. ostreum (Stauber, To this point of development the male and 1945 :272 ; Christensen and McDermott, 1958: female crabs have paralleled each other. The ex 150 ). The typical planktonic zoeal and rnegalo ternal morphology of both sexes is very similar pal stages are followed by a series of true crab throughout the prehard series and the Stage I instars, The first of these is the invasive crab instar. Only by the examination of the external (Christensen and McDermott, 1958 :150). Fol genitalia can the two sexes be distinguished. lowing the invasion of the host organism several Following this stage a dichotomy occurs in the instars occur which are collectively designated developmental cycle of the two sexes. The male as prehard stages. These prehard crabs have a is thought to remain in the hard stage, dying soft, membranous exoskeleron. With the excep after breeding. The female, however, moults tion of the initial invasive stage there is little soon after copulation and the new posrh ard in setal ornamentation on the pereiopods, which star is soft, with a membranous exoskeleton are cylindrical in shape. comparable to that of the earlier prehard stages. While the prehard instars were thought to oc The first posthard stage is referred ro as the cur chey were not described for any pinnorherid Stage II female. It is followed, both in P. pisum until the investigation of P. ostreum by Chris and P. ostreum, by Stages III, IV, and V. These tensen and McDermott (1958: 147). The total stages are characterized by an overall increase in number of prehard instars is still not known for size, greater complexity of the pleopods, and an any species, although Hart (personal commu increase in the width of the abdomen relative nication ) has found up to five prehard instars to the carapace width. The Stage V crab is the in F. subquadrata. Since the terminal one of terminal adult female. Although subsequent these is smaller than the smallest Stage I crabs moults may occur, they result largely in an in of this species it is certain that at least seven crease in size and there is little morphological prehard instars normally occur between the change. megalops and the Stage I crab. Because there is The Stage II crab is very similar in appearance a considerable variation in the size of the termi to the terminal prehard instar, In F. subquadrata nal prehard instar it may be assumed char the there is little increase in carapace or abdomen total number of prehard insrars also varies width during the terminal prehard-Stage I and somewhat. Stage I-Stage II moults. This stage (II) is very Subsequent to the series of prehard crab difficult to identify unless the actual Stage 1- 6 PACIFIC SCIENCE, Vol. XX, January 1966
Stage II moult is observed. Stages following it phology, become highly modified . The signifi can be readily distinguished, however , on the cance of this sudden cransformation is discussed basis of the differential growth of the abdomen in a later portion of this paper. and increased complexity of the pleopods. The pleopods of the first crab stage and the Since Atkins (1926:475) recognized only the first few prehard instars subsequent to it are Stage I-Stage V crabs and did not describe the merely small knobs protruding from the ven prehard series, the nomenclature originally ap tral surface of the abdomen. At this time th ere plied by her to the pinnotherid developmental is no differentiation into endo- or exopodites. instars is no longer adequate. However, as all the In the later prehard insrars immediately preced preh ard stages have not been described for any ing the Stage I or hard instar, the pleopods pinnorherid crab it would be difficult to rename become very conspicuous and show clear differ or renumber these forms at this time. For this entiation into endo- and exopodire portions. reason her original terminology, with some The smallest F. subquadrata found within a modifications made by Chri stensen and Mc mussel measured 0.85 mm in carapace width. Dermott ( 1958), has been retained in this This crab is somewhat larger than the first crab investigation . stage of this species reared by Hart (personal communication ), which had a carapace width of Invasive and Prebard Stages 0.76 mm. This difference in size may be ac counted for by assuming that the formerly The carapace of the invasive first crab stage planktonic first crab stage undergoes a moult of F. subquadrata is more square in outline than very soon after entering the host mussel. Con are the later prehard stages, which tend to be sequently it would be difficult to find a true first ovoid in shape. The eyestalks and p ereiopods of stage crab in a host mussel. the first crab stage are also proportionately However, comparison of the supposed first larger in relation to the rest of the body than crab stages removed from mussels with the are those of succeeding pr ehard instars. The known first crab stage raised by Hart indicates pereiopods of this instar are covered with swim that morphologically they are very similar or ming' hairs or setae. Th e distributional pattern identical. of these hairs is different, however, from that of As earlier noted , Hart has reared F. sub the Stage I or hard crab, which also has similar quadrata through five pr ehard insrars, the largest setae. The pereiopods of the first crab stage oc of these still being somewhat smaller than the cur with the hairs distributed over much of the smallest Stage I insrar yet observed ( 1.3 mm ). surface, giving the appendage a bottle-brush ap For this reason it may be suspected that several pearance. Th e hairs of this stage are also much instars intervene between the aforementioned more sparse and the entire structure does not forms. Christensen and Mcffermotr ( 1958: 150 ) appear to be as efficient an arrangement for . found that the smallest P. ostreum collected swimming as that of the pereiopods of the Stage measured 0.59 mm. Th ey suggested that at least I crab. Since the first stage crab apparently seeks four moults occur before a crab would rnouk out or in some manner becomes associated with into the Stage I instar. Th e smallest Stage I a host immediately after moulting from the insrar in their collection also measures 1.3 mm . megalops, appendages well adapted to extended Since, however, th e Stage IF. subq uadrata is swimming activities are not necessary. This in normally somewhat larger , it is suspected that star is able to swim, however , as is demonstrated at least seven moults occur between the invasive by its activities in the labor atory. first crab stage and the average Stage I instar, As an individual crab progresses through the When the method used by Hiatr (1948: 165) to series of prehard moults the swimming hairs extrapolate the number of intermoult periods in found on the per eiopods , as well as the swim Pachygrapsus crassipes was applied to F. sub ming abilities and activity, are lost until the quadrata it was confirmed that approximately Stage I or hard instar is attained. At this point seven ,to eight moults occurred between the first the swimming hairs, as well as the general rnor- crab stage and the average size Stage I instar. Biology of Fabia subquadrata-PEARCE 7
The abdomen width of the smallest F. sub and well ornam ented with funct ional swimming quadrata removed from a mussel was 0.26 mm hairs; in F. subquadrata and P. ostreum only the or approximately one-third the carapace widt h. second and third pereiopods bear the long plu This is a ratio t hat is approximated in all de mose swimming hairs, whereas they are present velopmental instars through the Stage II post on all the walking legs of P. pisum (Christensen hard. W ith the exception of a few abnormal and McD ermott, 1958:152) . While Darbishire females it is true for the hard Stage I form. ( 1900 ) is quoted (Christensen and McDermott, App arently it is rare for a male Stage I F. sub 1958:152) as stating that the Stage I P. pisttm quadrata to moult into a posthard, soft carapace uses the third and fourth pereiop ods for swim crab. As will be discussed later, however, such ming, in contrast with P. ostreum which uses males do occasionally occur and, in fact, may be the second and third, recent observations by more common chan suspected. In order to obtain Christensen (personal communi cation) confirm an approximation of the size of the terminal that P. pisum uses primarily the second and prehard instars, both males and females, each third pereiopods, as does F. subquadrata. indivi dual collected was measured. If, within a The carapace of the Stage I F. subquadrata, week, the crab moulted into a Stage I insrar the like that of P. pisum, is quite convex. Th e sur previous dimensions were recorded as chose of a face of this structure has a distinct pattern of termi nal prehard. bright orange marki ngs (see Maerz and Paul, The carapace width of 19 male prehard crabs 1930, plate 10, E-12). Th is pattern is very varied from 3.0- 5.3 mm, with a mean width constant and is found in almost all Stage I crabs. of 4.3 mm. The abdomen width ranged from Th e background is a brilliant whi te. The orange 1.1-2.1 mm and averaged 1.7 mm. It should be pattern tends co fade to a dull brown (Maerz kept in mind , however, that males may occa and Paul, plate 13, G-ll ) some weeks after be sionally moul t into a soft instar from the hard ing removed from the host mussel. Other stages Stage I form. Furth ermore, as will be discussed of this species, both pre- and posthard, do not later, this soft instar may subsequently revert to present any indication of this pigm enrarion. the hard form . Such a moulting sequence may Macroscopically the exoskeleton in these latter thus invalidate these measurements. forms appears colorless, although microscopic Since the female 'r egularly moults from a examination reveals isolated black and red term inal prehard form into t he Stage I instar chromatophores. Finally, as reported for the any dimensions of these forms can be accepted comparable stage of P. ostreum by Christensen as valid. Thirteen such moult s were observed and McD ermott (1958:152 ) , the Stage I F. and the individuals involved ranged from 2.7 subquadrata was noted to have two cylindrical 5.1 mm in carapace width prior to the moult. rods conn eoring ,the dorsal and ventral sides of The average carapace width of these terminal the body. Th ese structures , along with the al prehard females was 4.1 mm . ready discussed exoskeletal rigidity, may be modifications for a freeswimm ing existence. T he Stage I (Hard) Crabs Finally, in addition to these differences, the This is one of the stages originally described Stage IF. subquadrata varies from the other for P. pisum by Atkins ( 1926:478) and sub stages in having a heavy pub escence along the sequently applied to the comparable instar of antero-lateral margins of the carapace. This P. ostreum by Stauber (1945 :272-276) . The pubescence appears somewhat heavier in the latter suggested that it was during this stage male, but such differences are hard to quantitate. that P. ostreum invaded its oyster host. The average carapace width of 54 male Stage The Stage I instar of F. subquadrata is in I crabs, selected at random from collections many ways morphologically similar to the Stage made on July 29, 1959, is 4.1 mm, with a range I form in both P. pisum and P. ostreum. In all of from 1.3-6.8 mm. The mean of 29 female three species the exoskeleton is well calcified Stage I crabs collected on the same date is 3.5 and very hard. Th e pereiopods are flattened mm, with a range of 1.5-6.2 mm. This does not 8 PACIFIC SCIENCE, Vol. XX, January 1966
appear .w be as large a size difference between where tapering occurs, ,they are, in P. ostreum sexes as was found for P. ostreum by Christen and F. subquadrata, slender and lanceolate. sen and McDermott (1958 ). In addition to a sexual dimorphism in size the Stage I crabs have The Stage II Females other sexual differences. The abdomen of the female is different in shape from that of the There is no apparent increase in body size of male; .the lateral margins of che male's abdomen this insrar over the Stage I female. The average are concave, whereas those of the female are carapace width of seven Stage II female crabs straight. It has also been noted that an occasional which were observed to moult from the Stage I Stage I female will have an abdomen which is instar is 3.4 mm, with a range of 2.9-3.9 mm. The average abdomen width is 1.1 mm, with a relatively wider than the 1:3 abdomen-carapace range of 0,9-1.3 mm. ratio which is characteristic of most of the Stage The exoskeleton of the Stage II crabs is soft I crabs both male and female. Finally, the ab and membranous, as is that of the prehard in domen'of the female bears four well developed stars. There are few swimming setae or hairs pairs of pleopods which contrast markedly w!th to be found on che pereiopods, nor is there the two pairs of highly modified reproductive any pubescence along .the anterolateral carapace appendages borne by the male. margins. Both the male and female Stage I crabs have Th e appendages are subcylindrical, not flat much stouter chelipeds than eith er the pre- or tened as in the Stage I instar. The carapace is posrhard growth forms. The merus and carpus ovoid; the angles of the subpencagonal carapace are heavier and both fingers of the chela are of the Stage I form have become rounded. It is swollen. during this stage that the lateral carapace sulci As observed by Stauber ( 1945: 274) in P. (one of the definitive characters of the genus ; ostreum, the Stage I F. subquadrata possesses a Rathbun, 1918:101) become pronounced. They locking mechanism whereby the abdomen may appear faintly in che prehard stages and are be 'secured in the sternal groove. On the fifth hardly present at all in the Stage I instar. Wells thoracic segments of the sternal groove chere ( 1928:289 ) notes that these sulci are present are pairs of antero-venrrally directed knobs. in the newly moulted Stage I crab but are lost These knobs hook under shelves found on the with subsequent hardening. As previously noted, opposing ventral surfaces of th e abdomen in the typical pigmentation of the Stage I instars such a manner as 'to become securely locked is lost in rhe Stage II forms. when any attempt is made to lift forcibly Stauber (1945 :275) indicates large differ the abdomen of the living crab. Consequently, ences between the pleopods of Stages I, II, and whereas it is easy to displace the abdomen of the III in P. ostreum. Christensen and McDermott pre- and posthard instars it is very difficult to (1958:152 ) suggest that Stauber's series of free ehe abdomen from the sternal groove in the Stage II crabs may have included some prehard Stage I crabs. individuals. At any rate, no such marked dif The reproductive app endages of the male ferences could be found between the pleop ods Stage I mussel crab are very similar to those de of terminal prehards and Stages I, II, and III of scribed for P. ostreum by Stauber (1945: 276 ), F. subquadrata. and quite dissimilar from the reproductive ap There is little or no widening of the abdo pendages of P. pisum as described by Atkins men relative to the carapace in the Stage II crab. ( 1926:476 ). Atkins described the first copula The ratio between the two is approximately the tory appendage of P. pisum as blade-like and same as that of the Stage I forms. The sternal hairy. Recentexamination of preserved P. pisum groove remains deep and is only as wide as the material by the present author verified a con abdomen. The locking mechanism which was siderable difference. While the appendages of present and functional in the Stage I crabs no P. pisum are broad with almost parallel margins, longer operates. As was surmised by Stauber except for the distal one-fourth of its length ( 1945:278 ) for P. ostreum, this may be due to Biology of Fabia subquadrata-PEARcE 9
.rhe diminished rigidity of the exoskeleton in Th e Stage V Females the posrhard ins-tars. This is the definitive adult female crab and the stage most commonly found t hroughout the The Stage III Females year. As in previous posthard forms the ex oskeleton is membranous and, while the body This is the first stage subsequent to the Stage shape is similar to the Stage IV crabs, the rela I instar in which there is an increase in cara tively large growth of the abdomen causes this pace width over .that of the preceding instar, the ins-tar to become very awkward in its move Stage II form . The average width of 41 Stage III ments, especially when compared with the earlier females was 5.4 mm, with a range of 4.0- 5.9 stages. The abdomen is as wide or wider than mm. This instar is also the first in which the the carapace and normally it protrudes laterally abdomen is more than one-third as wide as the beyond the coxopodires and anteriorly to the carapace. The average abdomen width of rhe mouth parts. above Stage III crabs was 3.6 mm . The range There is a great deal of variability in this was 3.2-4.2 mm. stage, especially in the size and width of the Except for the relatively wider abdomen and abdomen relative to the carapace. From ob larger overall body size the Stage III crab is, servations made on moulting Stage V crabs it externally, morphologically similar to the Stage has been found that this stage consists of not II form. The carapace is soft and membranous, just one instar, as is usually true in the previous th e pereiopods are slender and subcylindrical stages, but rather of a series of growth instars, and devoid of swimming hairs. The sternal in which the general morphology remains the groove, however, is shallower and che abdomen same but with each succeeding insrar become no longer lies within the confines of this de somewhat larger than the one preceding. This pression. Rather, it extends both laterally and results in a wide range of size within this one, anteriorly beyond the borders of che groove. The arbitrarily designated stage. The smallest Stage pleopods are almost identical in both structure V crab observed measured only 4 mm in cara and serarion with those of the Stage II instar. pace width, whereas several Stage V crabs were found to measure 14 mm . The average carapace width of all observed Stage V crabs (831) was The Stage IV Females 9.5 mm, and the average abdomen width was The average carapace width of 33 Stage IV 10.3 mm. crabs is 5.8 mm . They range in width from 5.3 Christensen and McDermott (1958:162) dis 6.1 mm . The average abdomen width of these cuss the effect of the presence of P. ostreum in crabs is 5.4 mm with a range of from 4.8-5.9 slow-growing spat. They suggest that, while the mm. This stage is not only larger than the Stage growth of the crab is retarded in such host III instars but in addition obvious external oysters, the development is not affected to a changes indicate that it is sexually more mature similar extent. The data gathered on che F. sub than those stages which precede it. Ovaries con quadrata-M. modiolus relationship would sug taining large numbers of developing eggs were gest that a similar situation prevails. The very observed in 29% of the Stage IV crabs. Also, small, below average in size, Stage V F. sub while no ovigerous Stage IV crabs have been quadrata are usually found in relatively smaller rioted, it is significant that at this time the pleo host mussels. In a more recent study Houghton pods undergo the greatest change since cheir ( 1963: 254) reports a similar situation for P. initial appearance. These modifications in the pisum. pleopods involve changes in size, proportion, In addition to a positive correlation between and setal decoration. This is in preparation for crab and host size it has been determined that the deposition and attachment of eggs. Finally, there is a negative correlation between the size the abdomen is now nearly as wide as the cara of rhe Stage V crabs and .the depth of the water pace and is more concave than in previous instars. from which they were removed. Crabs reaching 10 PACIFIC SCIENC E, Vol. XX, January 1966
90 I I 80 ~ ~ o------I -, ILl " 70 u lO cl " Q. " , cl " 60 a« a: '0 ...... cl z o 9 "'l 5 0 fTI u, 1Il o -..... -l :to :c 40 -l l- ...... e 8 -.....- '"0- 0 o z ~ 3 0 Z et ILl ~ 7 20
10
20 40 60 80 10 0 11 0 120 14 0 160 18 0 200- DEPTH IN METERS FIG. 1. Curves showing: dash line, correla tion between mean width of Stage V crabs and depth of water from which their host mussels were collected; solid line, correlation between per cent of inf estation of host mussels and depth of water from which they were removed. matur ity in relatively shallow waters were, larger Stage V instar is reached ,that egg deposition oc on an average, th an crabs which have developed curs. The smallest oviger found measured 5 mm and are collected from deeper waters (Fig. 1). in carapace widt h; the largest was 13.4 mm . The Thi s relationship was noted to exist thr oughout largest crabs collected (i.e., those 14 mm in cara che entire one and one-half years that the crabs pace width), were not ovigerous. H owever, rheir were studied. A further discussion of these cor gonads did contain large numbers of well de relations is deferr ed to anoth er par.e of this veloped eggs and it appeared ,that these were paper. about to be spawned. The highly colored eggs contained in che The average carapace width of 187 ovigers gonads show clearly rhrough the thin mem collected from two depths off Mineral Point ( 55 branous exoskeleton. The color varies during de and 130 m ) was 8.5 mm . These crabs were re velopment; initially appearing chrome yellow, moved from mussels colleoted during a period they app ear coffee brown immediately prior to (November, 1959 ) when th e ovigerous females their deposition (see Maerz and Paul, plate 9, constituted almost 60% of the total population. K-2 and plate 15, A-ll). Unless a crab has just These ovigers were, on the average, 1 mm less become ovigerous, eggs are almost always pres in carapace width than the average of all the em in some stage of developm ent. It has been Stage V crabs collected during the per iod of observed ,that within a week after egg deposition study. This is undoubtedly due Abnormal Instars subcylindrical and with few swimming hairs. Such moults are not accompanied by significant Although th e sequence of developmental in grow th. In no case has a posthard male been stars, as already described, represents the normal observed to und ergo further moulting, as was situation, investigations preceding this one have observed by Atkins in P. pisum . revealed occasional deviations from this general These posrhard male F. subquadrata were ob pa ~t ern by other pinnotherid species. Orton served only during the summer months of July (l921 :533 ) described a single male P. pisum and August. Th is does not necessarily mean that which was morphologically similar to a soft, rhey do not occur at other times, since they could posthard female. Stauber ( 1945:280 ) discussed easily have been mistaken for prehard forms had a second stage, posthard male P. ostreum which they not been observed moulting from the Stage appeared externally to resemble the Stage II or I instar. III females. He notes chac they were found in Apparent abnormalities are found not only in "appreciable" numbers and that their size distri the males but also in Ithe morphology of the bution was somewhat greater than that of rhe Stage I females. In these cases Stage I females typical Stage I males.He suggested th at these are noted whose abdomens are precociously atypi cal males might be ". .. the result of some widened. This increase in width over that of the SOrt of parasitism as Mercier and Poisson (1929) normal individuals is quite large, the abdom en have reported for P. pisum ." Stauber further sug carapace ratio approaching that found in the gested that rhese posthard male forms were Stage III females. Other morphological aspects copulatory partners for zhe larger posrhard fe of chese ind ividuals tend to be normal, although males. Christensen and McDermott ( 1958 :152 ) thes e forms are invariably larger than the aver suggest that che abnormal P. ostreum referred 'to age female Stage I instar, Of 183 Stage I females by Stauber were actually prehards and that the examined, 5 were of this anomalous type. greater size range of Stauber's second stage, pOSt Christensen and McDermott (1958:152) re hard male over his Stage I series of male crabs pON similar anomalies in the Stage I females of was probably due 'to a sampling error. They do P. ostreum . In two cases they found individuals make the reservation that a hard Stage I male considerably larger than t he normal Stage I may, "now and rhe n," moult to a soft -shelled females. Both these crabs had abnormally formed form. Atkins (1958) presents evidence that, at pleopods. However, they do not mention any least in P. pisum , the hard or Stage I males do extraordinary increase in the relative abdomen quite frequently undergo a metamorphic moult width of these crabs. It was their opinion that into a sofr posthard form . She has repeatedly ob they had been retarded in their development, served the same crab change from one form 'to another with usually two or three soft forms in ECDYSIS IN F. subquadrata tervening between hard instars. These soft post hard males are usually found during the summ er Ecdysis is one of the most significant events months, June- August inclusive, in southwest in the life history of any crustacean. In a few England. It is during this period that the males crustaceans it has evolved to be primarily a moult and young crabs are found in mussels. mechanism allowing an increase in size to occur. Because of this she suggested that the soft post This is true both in rhe freshwater decapods , hard males occur during the periods of rapid the Potamonidae, crabs which hatch from the growth. egg as a replica of the adult (Rathbun, 1918 : A similar situation has been found with re 11 ), and in a species of the Oxyrhynca or spider gard to F. subquadrata. D uring the summer of crabs, N acioides serpulifera ( Rat hbun, 1914: 1959 eight Stage I males were observed to moult 653 ). In most marine crustaceans, however , into sofc posthard forms. The latcer are similar moulting is accompanied not only by increased in body shape to the Stage I insrars but are soft size but also by considerable morphological and membranous. The pereiopods of these soft change. In no group is this more true than posthard males are, as those of posrhard females, in the family Pinnotheridae. Certainly other 12 PACIFIC SCIENCE, Vol. XX , January 1966 crustacean groups have representatives which held. Th ese authors, th erefore, had co resort to undergo extensive changes rhrough ecdysis, but other techniques in order to obtain moulting few others, particularly among the brachyuran specimens. families, have fiued into the postlarval (post N o external change in color or opacity her plankt onic) portion of their life cycle such alds approaching exuviation in the hard Stage I complex morp hological changes as accompany crabs. Only the somewhat more flexible nature ecdysis in the pinnotherids. of t he exoskeleton and the appearance of a crack While previous investigators have described along che postero-lateral margins of the carapace ecdysis and accomp anying phenomena in other indicates chat ecdysis is under way. The cara brachyurans ( Drach, 1939; H i a ~t , 1948; Guysel pace of the Stage I form does not become as soft man, 1953; and Knudsen, 1957), Iietle informa or decalcified as is indicated for some other tion is available concerni ng these processes in brachyuran species (Hiatt, 1948:156); however, the Pinn otheridae. For this reason careful notes a recent paper by Kn udsen (1957:134) states were made of any moulting activities of F. sub that in the Californi a Xanthidae che exoskele quadrata during the period of this study. Sub ton does not become fragile p rior to ecdysis. It sequent studies of ecdysis in F. subquadrata as may be that, a-t least in the case of F. sub well as other W est Coast pi nnorherids have been quadrata, since che following posthard instars made ( Pearce, 1962b). Th ese studies involved are not heavily calcified che hardening salts re the use of both ligh t and electron microscopes in main in the exuviae of the Stage I instar rather determining cissue changes which occur during than being retained in the crab's tissues to be ecdysis. Th ese data will be included in a separa te subsequently redeposited in th e new exoskeleton. paper, the present work noting only the macro Abom one day after the onset of the opaque scopic aspects of ecdysis in F. subquadrata. appearance in the pre- and posrhard crabs a Two disti nct phases of ecdysis can be recog crack appears along rhe epimeral line, and at nized in all brachyurans. T he first is preparaorory this time the active phase begins. Th e body now and, to all outward appearances, is passive in expands due to the uptake of water ( Drach, na-ture al-though there can be no doubt that 1939; Guyselman, 1953:129 ). This in effect physiologically the animal is very aotive. The lifts and frees the posterior portions of the cara second, or active phase, involves the actual ex pace. In t he pre- and posrhard stages the old uviation of rhe cast. This phase is characterized integument being shed has the consistency of by a great deal of movement by the crab. heavy, wet cellophane. Further, because of its Most observations concerning the moulting of supple nature, it is never lifted oro the extent of F. subquadrata were made on animals recently a 30° angle as was noted in Paehygrapsus eras rem oved from a host mussel. A total of 134 sipes by Hiaor (1948: 157) or as is found in the moults were recorded. In 61 of these the dimen Stage I mussel crabs. Rather, t he old integument sions were noted both before and after ecdysis. lies free upon the dorsal surface of th e new Prehard and posrhard crabs rhat are about to integument of the carapace. mouLt can be easily detected. One to two days prior ro exuviation animals in ,this state become As is noted by Knudsen (1957: 136 ) for the "creamy" and very opaque in appearance. Unlike xanthid crabs, it is evident that muscular move other species which have been studied (H iatt, ments occur during th is period since th e new 1948:155) , they do remain quite active. All integument can be observed to be pulled inward , stages of Fabia have moulted under laboratory form ing surface depressions. conditions-some after being held as long as six Following the freeing of che posterior por weeks. Christensen and McDermott ( 1958: 150 ) tions of the carapace the last pair of thoracic found it difficult to obtain moulting P. ostreum appendages, t he fourth pereiopods, and the ab under the laboratory conditions in which they domen are simultaneously freed from rhe old worked. Unless crabs were "obviously ready to integument. Th is is a procedure intermediate moult on arrival Ito the laboratory" no moulting between that observed by Knudsen (1957: 136 ), occurred in the Petri dishes in which they were who insists that the abdomen is freed first in Biology of Fabia subquadrata-PEARCE 13 the xanthid crabs which he studied, and Hiatt once following ecdysis no measurable expansion 0 948 :157) who reports that in the grapsoid, was noted after th e first posrrnouk measurement P. crassipes, th e legs are first removed, then the was made. The first posrrnoult measurement was abdomen. routinely taken 30 minutes following the com As soon as ,the posterior pair of appendages pletion of exuviation. This implies th at the crab are free ,the animal then apparently uses them expands to its postexuvial dimensions during to exert pressure against th e old integument in and immediately following ecdysis, with little or such a manner as to push che rest of the body no increase occurring over an extended period free of the exuvia or cast, By thi s time the crab following ecdysis. This agrees favorably with the has moved far enough posteriorly within the cast minim um cime requ ired for the final expansion to allow rhe more anterior pereiopods and of the freeliving xant hid crabs (Knudsen, 1957 : mout h parts to be freed. Th e former can then 141). The lacter required from 30 minutes to be pull ed into the area vacated by the cephalo 2 hours. thorax proper. Th is description is true of all the Th e degree of postexuvial size increment in pre- and posthard crabs which were observed. F. subquadrata varies not only with the stage at In the case of t he Stage I crab it is more difficult which rhe moult occurs but also oro some extent to determine the manner in which the anterior between individuals of the same stage. Prehard appendages are freed since the exoskeleton of crabs moulting into new prehard instars had an this stage is completely opaque. average increase of 16.5 % , with the smaller The aotive phase of exuviation varied between crabs, i.e., the second or rhrid posrplankronic 15 and 45 minutes, with the average time being instars, increasing as much as 20% .Similar 20 minutes. The larger crabs (g reater than 10 increases have been found in the early instars mm in carapace width) took, on an average, of other brachyuran species (Olmstead and somewhat longer. There were exceptions, how Baumberger, 1923; Broekhuysen, 1941; McKay, ever. Th e longest period observed was taken by 1942; H iatt, 1948 ) . However, no F. subquadrata a Stage III female 5.7 mm in carapace width. of any stage ever showed the 400% variation Th ere was Iirtle difference in the average time indicated by Hiatt ( 1948:163 ) for P. crassipes. required by hard Stage I or pre- and posthard Generally, during the moulting from the termi crabs. nal prehard instar into rhe hard Stage I crab, In only 2 our of 61 closely observed moult and from th e Stage I into th e Stage II instar, ings was F. subquadrata seen oro moult during there is little or no increase in size. the daylight hours. This might seem surprising Only two Stage II females were observed in view of the fact chat mussel crabs are rarely undergoing ecdysis. The first of these crabs in a photic situation and thus darkn ess would showed no increase in carapace width although not be of protective advantage during the cru the abdomen became 20% wider. The other cial period of moulting. However, since most of observed Stage II crab increased 10% in cara the freeliving brachyurans do moult at night pace width and 40 % in abdomen width during ( Broekhuysen, 1941; McKay, 1942; Hiatt, 1948; the moult into the Stage III form. It is during and Kn udsen, 1957) , it can be hypothesized this moult that a significant change occurs for that F. subquadrata retains an inherited mechan the first cirne in the carapace-abdomen width ism involving the inhibition of moulting by ratio ,N eedham ( 1950) has discussed the quan light. As noted by these authors such a mechan titativ e aspects of rhis allometric growth in ism would have obvious adaptive advantages co P. pisum. freeliving forms, but i,t would be of little sig The moult from the Stage III to the Stage IV nificance to a symbiotic crab living in a non instar, in six observed cases, was accompanied by photic situation. an average 12.2% increase in carapace width Th e length of time required for t he maximum and a 65% increase in th e widt h of the abdo postexuvial expansion oro occur was not deter men. During chis moult occurs the greatest in mined in every observed rnouk, However, in the crease in abdomen width relative oro the carapace cases in which a crab was measured more than width. 14 PACIFIC SCIENCE, Vol. XX, January 1966 Onl y a single Stage IV crab was observed apparently normal exoskeleton consistency is undergoing ecdysis. A 10% increase in carapace reached following ecdysis. However, more recent width occurred during this moult. The abdomen work ( Pearce, 1962b) on related species indi of the new instar (Stage V ) increased 34 % over cates that, whileexternal app earances suggest a that of the original Stage IV crab. Assuming "normal" intermoulr conditio n, the actual de this single observed moult 12 !a en >- II g 15 I&J ~ z » 10 -l en 111 lD ::u 6 A M J J A s o N 0 J F M 1959 1960 FIG. 2. Cur ves showing: dash line, surface water temperatures taken in the San Ju an Archipelago duri ng 1959 and early 1960; solid line, per cent of Stage V crabs in ecdysis during the period from February, 1959 to March, 1960 . Biology of Fabia subquadrata-PEARcE 15 middle August. During this period as much as ton found a Stage I female with its spermathecae 13% of the observed population was undergo filled with viable sperm. This latter discovery ing ecdysis. JUSt as a correlation was noted by was repeated by Atkins (1926:478 ) and was Hiatt ( 1948:161) for P. crassipes, so was a indicative of a precocious copulation in P. correlation noted between monthly tempera.cure pisum. means and the percentage of ecdysis for F. sub Examination of mid-water plankton trawls quadt'ata ( Fig. 2 ). The cemperacures are surface made by members of the Department of Ocean water recordings made by the U. S. Coast and ography of the University of Washington in Geodetic Survey in 1958-59 for the waters of the southern waters of Puget Sound during May, the San Juan Archipelago, and closely approxi 1957 resulted in the finding of 56 Stage I F. mate the means recorded over a five-year period subquadrata of both sexes. Examination of ma by Phifer and Thompson (1937) . While chese terials collected in a similar manner prior to temperatures are taken at che surface, it is as and subsequent to this period revealed a paucity sumed that they would be valid at the depths of these crabs, only 2 being found. Microscopic from which the mussels were removed since examination of the spermathecae of the females Phifer and Thompson (1937 :34) note that the taken in the trawls, as well as those removed waters of San Juan Channel are very homogene from host mussels, revealed that F. subquadrata, ous co a depth of 100 m. like P. pisum and P. ostreum (Christensen and It was noted that che percentage of Stage V McDermott, 1958), copulates precociously. females in ecdysis was greatest during August, Since the sample taken from these plankton regardless of the depth from which they were tows represents the only time that female and removed. Figure 2 represents only Stage V male F. subquadrata have been found together, crabs and does not include those immature crabs it is probable that copulation in this species oc observed undergoing developmental moults curs during a period when both the male and throughout the year. female Stage I crabs leave their host bivalves While increased temperature may not in itself and assume a temporary freeliving existence.P be the direct cause for the onset of ecdysis, it Of the total of 56 Stage I F. subquadrata taken may be an indirect .factor. Certainly the moult in these tows, 29 were males and 27 females. ing sequence is correlated in a number of ways Conversely, in only 3 out of 2,088 total observed with reproduction, the success of which in turn infestations were double infestations of the host depends upon factors favoring the survival of mussel ever found. In 2 of these cases 2 male the zoea. Since it is known that phytoplankton Stage I crabs were found together, and in the standing crops vary with water temperatures and remaining example a Stage I female was found in turn are important for the development and together with an unsexable prehard crab. survival of many zooplankton larval types, in If copulation occurs while the female is in the cluding zoea and megalops, so temperature may Stage I form, as seems likely from present evi- be correlated with the periodicity of ecdysis. 2 It has only recently come to my attention that Sakai (1 939 :60 4 ) has report ed a similar swarming THE DEVELOPMENTAL CYCLE in the pinnotherid , Tritodynamia horvathi, which is found in Japanese and Korean waters. He notes that Early investigators (Thompson, 1835) as both males and females swarm together in large num sumed that the male pinnotherid sought out his bers. The swarming or migrations occur in Japanese copulatory partner by moving from host to host waters ". .. from the middle of autumn to the begin during the reproductive period. Later Orton nin g of winter." In Tinkai Bay, Korea, however , Ka mira (1935 :36 ) reports it as occurring in June and ( 1921:533) tended to substantiate these views July. The crabs are often found in such dense num by his discovery of individual male P. pisum bers that they are harvested and used for fertilizer. caught between the valves of the host mussel, In addition, large schools of fish follow the moving Mytilus edulis. The implication was that the crabs, obviously feeding upon them. No ecological significance was attached to the swarming, and Sakai males were trapped while seeking to gain en does not distinguish at what stage of the life cycle trance to the host of a female. In addition, Or- the swarming occurs. 16 PACIFIC SCIENCE, Vol. XX , January 1966 dence, and if copulation were consummated in through the middle of February, with a peak at the host mussel, it would be expected that oc the last of January. As many as 75 (60% ) of casional pairs of males and females would be the total adult population of 126 Stage V fe found in the large numbers of mussels examined males collected off Point Caution on January 28, throughout the year. Such double infestations 1959 were ovigerous. Collections made during have not been found , however. Furthermore, the the winter of 1959-60 indicate that the ovigers finding that Stage I males and females swarm were more numerous earlier in the season during together in open water gives evidence that the this particular year. A sample of 22 Stage V copulatory act takes place outside the host. crabs collected off Mineral Point and examined This is contrary to what other investigators on November 23, 1959 contained 19 ovigerous have found to be true in related pinnotherids crabs ( 87% ). Large percentages of ovigers were (Thompson, 1835; Orton, 1921 :533; and Chris found at the other collecting sites during this tensen and McDermott, 1958: 166 ). period. More recent evidence, involving the attrac Since the swarming females copulate in late tion of swimming Stage I crabs to a "night light" May, an inter val of some 21 to 26 weeks would used at the Friday Harbor Laboratories, indi ensue before the start of egg deposition in cates that the swarming behavior is probably November and December. During this period restricted to late May and early July. Of eight the precociously inseminated female must un crabs taken in this manner during the spring dergo the series of growth and metamorphic and summer of 1961 only one was obtained later moults which have already been described. Be than mid-June and none earlier than the 4th of cause there is no overwintering of immatures as May. reported for P. astreum by Christensen and It is of interest that the average sizes of the McDermott (1958:158 ), the number of imma swarming Stage I males and females, taken in tures, both male and female, present in the the mid-water trawl and at the "night light," are population is low during the winter months of approximately the same. The average carapace November, December, and January. width of 32 males is 3.51 mm, while the same The immature crabs are found in greatest numbe ~ of females averaged 3.58 mm . numbers during the early- and mid-summer Collection data show that swarming was not months. On July 21, 1959 they constituted 56% restricted to only one station or limited area. of a total sample of 119 crabs collected off Point Rather, it was found to occur at several widely Caution. They formed a comparable percentage separated stations in the San Juan Archipelago of the population at the other stations during and Puger Sound. Since the host mussel is widely the period of June 15-August 1. The Stage I distr ibuted in these waters, it would be ex crabs were particularly prevalent in the samples pected that the crab symbiont is found equally taken during May. The collection taken at Min dispersed. eral Point on May 4, 1959 included 37 Stage I A current investigation of Pinnath eres macu instars. These 37 crabs constituted 31% of the latus as part of the benthic community in the total population. The remainder were mostly Woods Hole area indicates that this species also prehard stages which would moult into the Stage engages in a copulatory swarming. In 1963 this I form before the month was over. The early swarming reached its peak during the last two posthard forms ( the Stage II and III instars) weeks of October. A more detailed account of were more in evidence during June and early this and other aspects of the biology of P. ma ca July, with the later posthard forms (the Stage latus will appear in a later paper. IV and V instars ) becoming prevalent in August When all the material collected from the San and September. Juan Archipelago area in 1958-59 is considered, From the discussion above and Figure 3 it is it is clear that the greatest number of ovigerous obvious that the ovigerous females occur pre females appears during the winter months. Large dominantly during the months of November, numbers are initi ally found in early November December, and January. During Februar y the and form a significant portion of the population eggs begin to hatch and the new larvae spend Biology of Fabia subquadrata--PEARcE 17 60 +--+ 0_ t "0_ I I 50 /0_0\ -0 Z I 0 I ... -+", ~ + ...J 40 0 ~ \ / , .- Q. \ / \ 0 \ Q. \ + / 30 '0 --+ p'- / IL <, 0 -, a-t <, <:» \ / 20 \ 0/ \ ...- ...- '0/ / 10 A s o N 0 J F M A M J J A s o N o 19 5 8 1959 FIG . 3. Curves showing per cent of total crab population constiruted by: solid line broken by circles, ovig erous Stage V females; solid line broken by crosses, nonovigerous Stage V females ; dash line broken by circles, immature crabs of both sexes. Samples collected from San Juan Channel off Point Caution, San Ju an Island . Washington between .August 8,1958 and November 23, 1959. upwards of eight weeks in attaining the first true moults which terminate in the definitive adult, crab stage. Hart (personal correspondence) has the Stage V crab. recently reared F. subquadrata. She found that The immature posthard and Stage V crabs re eggs laid on October 27, 1958 hatched on Febru tain the full spermathecae which result from the ary 6, 1959 and moulted into the first true crab precocious copulation of the Stage I insrar, Non stage on Apri l 6, 1959. As already discussed, it ovigerous Stage V females are present in the is assumed that the first crab stage enters an greatest numbers during August, September, and initial bivalve host, usually Mod iolus modiolus, October, and this is the period when the num and completes a series of prehard moults which bers of immature crabs are sharply decreasing. culminate in the hard, Stage I instar. This is true During late October and early November the both of the male and of the female. It is sus ovigerous females become more numerous and pected that in general the prehard series occu once again, during mid-winter, they constitute pies the interval between early April and late the majority of the population. May. At this time the Stage I crabs leave their There is a period of three months between hosts and engage in the swarming activities the time the Stage V crabs are first found in which culminate with copulation. Following increasing numbers and when the ovigerous these activities the Stage I female, and at least females are noted to represent a significant por some males (since larger Stage I males occur in tion of the population. Furth ermore, since many the collections throughout the remainder of the Stage V crabs are found moulting duri ng this year) , return to a bivalve host, where the female period one may assume that at least one or two undergoes a series of at least four posthard moults will occur in this period. This assump- 18 PACIFIC SCIENCE, Vol. XX, January 1966 tion is supported by the fact that the earlier It is possible that they are the offspring of second Stage V crabs (found in late summer) are al year females. ways smaller on an average than are the ovigers The immature crabs which are found later in taken later in the year at the same station and the season and throughout the winter ( Fig. 3 ), depth. when the Stage V crabs are predominant, are The only previous study concerning the com almost certainly the result of the egg deposition plete postplanktonic life cycle of a pinnotherid subsequent to the period when the majority of is that made by Christensen and McDermott the crabs spawn. Similarly, Stage I males are (1958). They found that: (1) the number of found throughout the winter. Since they too are Stage I P. ostreum was at a minimum during larger than the average Stage I males collected the winter and spring months and that in June during May, when the swarming occurs, it is large numbers of Stage I crabs appear in the believed that they represent a remnant of the oyster hosts; (2) the Stage I crabs represented, male population of the previous summer. dur ing this period, up to 60 % of the total popu Atkins ( 1955:689 ) has shown that it is pos lation; ( 3 ) 45% of the Stage I crabs were sible for one implantation of sperm to success males; (4) by the end of July only 5% of the fully fertilize a second egg deposition which crabs collected were males and by early Septem might be produced by a female P. pisum . Chris ber not a single male could be found . tensen and McDermott ( 1958 :167 ) also present In addition to the ultimate disappearance of evidence tending to confirm the same thing in the males from the P. ostreum population they P. ostreum, as do Wells and Wells (1961 :275) noted that during a period in late June a sig in Pinnaxodes floridensis, and Pearce (1962b) nificant number of double infestations began to in Pinnixa faba and P. littoralis. app~~r . They suggested that these could only be Whether it is possible for a second copula due to new invasions by male Stage I crabs seek tion to occur in F. subquadrata is extremely ing copulatory partners within the host oyster. doubtful. Christensen and McDermott (1958: In view of this fact, and of the present observa 167) report what -mighr be a copulation be tions that F. subquadrata engages in swarming tween a male, Stage I crab and a female, Stage and the male survives through the summer, it V instar of P. ostreum . They doubt, however, can only be assumed that P. ostreum and F. sub that such a copulation would normally occur in quadrata have diverged widely in their repro this species. If copulation in F. subquadrata is ductive habits. restricted to the period of swarming in open It is believed that some female F. subquadrata waters it is obvious that, because of its morphol live for more than one year and also reproduce ogy, the Stage V instar can never leave the host more than once. Figure 3 shows that there are and is thus unable to copulate a second time in always residuals of the nondominant stages. For open water. Since no Stage V female has ever instance, in the summer when the immature in been noted to be accompanied by a hard stage stars (of all stages ) predominate, there is always male or, for that matter, a male of any other a percentage of Stage V crabs and occasionally stage, it can probably be assumed that a second an ovigerous female. Since these crabs were fre infestation is not tolerated and it is thus doubt quently much larger than average (often greater ful that a second copulation would or could oc than 12 mm in carapace width ) it is thought cur. Furthermore, as was pointed out by Stauber that they are remnants of the previous year's ( 1945 :270) for P. ostreum, the size difference adult population and not merely precociously between the average adult female and the hard developed individuals of the present year. Chris stage male probably makes copulation mechan tensen and McDermott (1958:159) present evi ically impossible. dence that P. ostreum may live as long as three years. There can be no doubt, however , that GROWTH AND DEVELOPMENT CORRELATED some of the smaller, "residual," Stage V F. sub WITH HOST SIZE quadrata are the result of spawnings occurring Wells (1940:45) found that a definite posi somewhat earlier in the year than the majority tive correlation could be established between the Biology of Fabia subquadrata- PEARCE 19 carapace width of the mussel crabs and the They further note that while smaller crabs may length of their host mussels: by grouping shell occasionally be found in proportionately larger lengths and plotting them against the mean cara bivalves, the converse seldom occurs, i.e., large pace width of all the crabs found in each shell crabs are rarely found in proportionately smaller length group, he could obtain a curve indicating oysters. that the carapace width was, in general, propor The present work extends that of Wells, who tional to the shell length of the host. Atkins used a relatively small sample of III unstaged (1926:482 ) was able to find a size correlation crabs in his study. During the present investiga between 34 P. pisum and their bivalve hosts. tion examination of 305 crab-mussel associa More recently Christe nsen and McDermott tions has resulted in the curve presented as (1958:160 ) established a similar relationship Figure 4. This curve, derived from material col between all stages of P. ostreum, as well as for lected at one station during the period July 8 only the Stage V instars of that species, and their to November 11, 1959, is indicative of this host, the American oyster, Crassostrea virginica. correlation. The latter proposed that external factors, with After examination of the host mussels for the the amount of food probably being the most im presence of the pinnotherid crabs two rather portant, act upon both the bivalve and its sym interesting facts emerge. One was that immature biont crab so as to regulate the growth of the crabs are almost always found in the propor crab in such a manner that it "fits" the host. tionately smaller host mussels. This is true even NO. OF CRABS 29 30 32 19 22 43 62 41 23 4 14 12 ::& ::& I (J) 10 a:l ~ 0: U u, 8 0 :I: l- -- e -: 6 ~ z ~ w ::& 4 -: 2 vr 0-1011-2021-3031-4041-5051-6061-7071-8081-9091-100 LENGTH GROUPS OF MUSSELS - MM FIG. 4. Curve showing positive correlation between carapace width of all stages of F. Jubquadrata and length of host mussels from which they were removed . Both mean width and size range are given for the crabs. 20 PACIFIC SCIENCE, Vol. XX , January 1966 though the larger mussels could obviously accom "wastage" noted by Th orson ( 1950:3 ) 10 so modate them. Second, the very large mussels, i.e., many mari ne forms. over 85 mm in length, seldom contain a crab. It The second suggestion involves the hypo is easy to see, as did Christensen and McDermott thesis that the crabs are attracted, selectively, to ( 1958:161) , why small mussels only rarely con the smaller mussels by a "host factor." Recent tain relatively larger crabs. The reverse, how studies by Davenport ( 1950, 1953a, b), Johnson ever, is not so easy to explain, i.e., why are ( 1952 ), Hickok and Davenport (1957) , and smaller crabs seldom found in relat ively larger Sastry and Menzel ( 1962) indicate that in cer mussels? Also, why is it rare that few crabs, of tain cases of symbiosis the commensal or para any size, are found in the very large mussels site is indeed attracted to the host by a diffusible above 85 mm in length, the latter in spite of the factor from the host. The same factor may be fact that such mussels are relatively abundant? used to maintai n the relationsh ip once it has The occurrence of the early posrplanktonic been established. Preliminary work recently car stages has been noted only in the mantle cavity, ried our by Davenport (personal communica and not throughout the entire water-conducting tion ) at the Friday Harbor Laboratories did not system as noted for P. ostreum by Christensen reveal evidence to suppOrt the existence of any and McDermott ( 1958:173 ). Therefore, it is such mechanism between F. subquadrata and its not felt that such crabs have been overlooked in host, M. modiolus. It is important to note, how the microscopic examination of the larger host ever, that these experiments were conducted only bivalves. with the Stage I and older instars; and it is quite Because the very small prehards and invasive possible, in fact probable, that such an inter stages (l.0 mm or less in carapace width) are action might be found only between the invasive most frequen tly found in the smaller mussels stage and its host, and perhaps even at a par (15 mm or less in shell length), it is thought ticular time during the insrar's existence. that the newly moulted true first crab srage, Johnson (1952) reported that a chemotaxis when settling our of the plankton, usually selects existed between the pinnotherid, Dissodactylus a mussel of this size for its host. Figure 5 in mellitae, and the echinoid, Mellita. H is work dicates that 80 % of the true first stage crabs are with two other pinnotherid species, however, found in mussels rangi ng from spat to 20 mm did not reveal the existence of any attractive in -valve length. The mechanism affecting this mechanism between -them and their hosts. He selection is not definitely known, bur at least two suggested that the chemotaxis between Disso possibilities might be suggested. dactylus and Mellita acted to enable the pinno It is possible that the smaller mussels use therid to maint ain a continuous relationship qualitatively or quantitatively different food ma with the host in an environment (heavy surf) terials than do the larger mussels. Further, it is in which they might readily become separated. quite possible that the invasive and other pre In the Pinnixa chaetopterana-Chaetopterus re hard instars require a similar size or type food lationship, as well as the Pinnotb eres- Ostrea and are thus obliged to infest initially the relationship, both of which he studied, it was smaller, immature mussels. Small crabs which suggested that the negative evidence for the ex fortuitously find their way into compatible small istence of a host factor might be the result of mussels would survive, while those crabs that using the experimental devices with a stage of infest the larger mussels would not. The fact the crab's life cycle which is not attracted to that occasionally small, invasive stage crabs are the host. It might well be that other stages do found in larger mussels ( Fig. 5) could be ex respond. plained by regarding these crabs as in a transient The recent study by Sastry and Menzel situation, in which the mussels have only re ( 1962) , while indicating that Pinnotheres macu cently become infested by the invasive crab, latus is attracted both to the bay scallop, Aequi which would soon be eliminated. The waste pecten irradians concentricus, and the penshell, which would accompany this elimination is an Atrina rigida, makes no mention of the stage of other example of the normal larval or juvenile the females used in the experiments. They do Biology of Fabia subq uadrata-PEARcE 21 distinguish between adult and "early" male demissus, on the basis of oxygen consumption stages, although this distinction is not evident (used as an index of metabolic activity).One in the summary of their experimental results. of four factors affecting the metabolic rate was The hypothesis that the invasive stage F. sub the age of the prey (the others were species, quadrata is selectively attracted to the immature growth rate, and feeding). Both Haskin (1950) spat of the host mussel is made more plausible and Carriker (1955:49) have shown that there in the light of recent evidence presented by is a predilection by the drills in their choice of Blake (1960). He has found that the predator younger prey. It is hoped that further investiga oyster drill, Urosalpinx, is attracted selectively tion of this aspect of the relationship between to its prey, Crassostrea virginica and Mod iolus F. subquadrata and M. modiolus can be made in a, ~ 0 Q:: (!) :I: t- o Z w 60 -.J -.J W en 50 en ~ ~ Q:: w 40 a.. a z ~ 30 0 u, en ow 20 II) 10 a, :J 0 I- (I) 0:: 25 -J (!) LLJ (J) J: (J) I- J (!) - ~z 20 - u, l1J o .-J. 0:: :I: LLJ 0 CD FIG. 6. Hi stogram showing per cent of total populati on of M. modiolus constituted by each length group. Biology of Fabia subquadrat~PEARCE 23 Observations on the percentage of infestation are commens al in habit. A current investigation of the various size groups in the mussel popu of this species in the temperate waters off Cape lation tend to confirm this speculation. From Cod indicates that, as with P. ostreum and F. Figure 4 it may be seen that the average adult subquadrata, all the stages, with the exception of size (8 mm) of the crabs on this date is reached the invasive first true crab stage and the swarm in the mussel length group of 51-60 mm. This ing Stage I instar, are normally symbiotic. same average size is also typical of the crabs It has been observed that Stage V crabs and found in the larger mussel length groups, i.e., posthard forms can and do vacate a host mussel larger than 60 mm , although the total number of which is moribund. This is, however, the only adult crabs found in each mussel length group time that these stages have been found outside a decreases above the 61-70 mm range ( Fig. 5). host organism under laboratory conditions. This decrease in the number of crabs per grou p During August 1959 several immature crabs is probably due to the posrreproductive mortality were found in bivalves not previously recorded following the adult and ovigerous stages. It can as hosts for F. subquadrata. The new host species also be noted that the greatest number of mus include A starte compacta, Cardita oentricosa, sels is found in the 71- 80 mm length group Crenella columbiana, and K ellia sp. All these (Fig. 6); and furthermore, while there are almost species are quite small. N one, according to Old as many mussels to be found in the 81-90 mm royd ( 1928 ), reaches a length grea ter than 25 group as in the 61-70 mm range, the number of mm. One of them, C. colum biana, rarely exceeds infested mussels in the former group is only 16 mm. All the crabs found in these hosts were one-third that in the latter. Since this is gen either prehard or Stage I crabs. Because none of erally true of the mussel populations at all the these bivalves normally attains a size sufficient collecting sites, it seems to indicate that the crabs to contain an adult F. subquadrata it is thought are outlived by their hosts which, usually, are that the Stage I crabs, after leaving these small not subsequently reinfested. initial host species to take part in the copul atory Closely related to the problem of the size re swarming, do not return to the smaller bivalve lationship existing between F; subquadrata and species but rather secondarily infest a larger host, its host is the problem of how permanent is the usually M . m odiolus. If the smaller initial host relationship between the crab and its individual species are reinfested by postswarrning. crabs it host mussel once this has been established. Does is qu ite likely that the definitive adult crab stage the crab remain with in the host after having is not attained. Examin ation of 262 individuals initially infested it, or is the relati onship tran si of these small bivalve species has not revealed tory, with the crab at some period leaving the the presence of a single adult crab. host? A comparable host change has been described Rathbun ( 1918:62 ), Orton ( 1921:533 ) , and for P. pisum by Christensen ( 1958: 3). He Berner (1952 :345) all suggested that the vari found that on the west coast of Sweden the first ous hard stage male pinnotherids with which crab stage initially infests the lamellibranch they worked were freeliving. Christensen and Spisula salida and that later, upon reaching the McDermott ( 1958 :175 ) doubt this for P. os Stage I insrar, it leaves S. salida and secondarily treum and P. pisum , and they report that the infests M . m odiolus. The major difference be male leaves the host only temp orarily duri ng tween F. subquadrata and P. pisum , in this re the copu latory period to seek a mate. They s.ate gard, is that in the latter the host change be also that this migration is only a phase in the tween lamellibranch species seems to be regular ordinarily commensal or parasitic life of the or obligate, whereas F. subquadrata appears only crab. More recently, Sastry and Menzel (1 962 : occasionally to undergo interspecific change, the 390), in their discussion of the chemotactic re usual situation being intraspecific. In a more re sponses of the pinnotherid, Pinn otberes maw cent report, Christensen ( 1962 :6) notes the latus, to its bivalve hosts, quote Rathbun ( 1918: occurrence of an ovigerous P. pisum in a pr i 76 ) to the effect that the early stages of the mary host, S. salida, collected at Frederiksh avn, male are freeliving whereas the later adult males Denmark, in the summer of 1960. Th is indicates 24 PACIFIC SCIENCE, Vol. XX, January 1966 that P. pisum is occasionally capable of reaching the eroded and undamaged ctenidia of a mussel. adulthood in the primary host. The damage seems to be caused by the constant contact of the chela against the edge of the THE RELATIONSHIP BETWEEN F. subquadrata ctenidium. It is also noted that the crab "nips" AND ITS HOST, M. modiolus at the food string with the chela, and very likely the gill margins will also be pinched and cut as By placing various stages of F. subquadrata a result of this action. Damage to the palps often within mussels which have had a "window" accompanies ctenidial erosion, and they may be opened in one of the valves, in the manner re much reduced in length as well as malformed ported by Orton (1921) and MacGinitie and due to the presence of a pinnotherid crab. MacGinitie (1949:313 ), it has been possible to Besides the effects of the chelae during feed observe the behavior of the crabs within the ing, the dactyls of the pereiopods, used to sup host. port the crab, also contribute to the gill erosions. Probably the most noticeable feature of the By repeatedly inserting the dactyls into the gill relationship is the comparative inactivity of the lamellae the crab causes progressive erosion dor Stage V crab within the bivalve host. Most of sally from the point of the initial damage. Once the movements noted were associated with feed the entire underlying gill has been eroded away ing activities. the crab maintains its position by inserting the The adult female, without exception, is found dactyls into the mantle. This results in a patho occupying the anterior half of the mussel's logical condition in which the constant irritation mantle cavity. This is generally the widest part by the dactyls causes a blister or'cyst-like forma of the mussel. The crab's abdomen is always tion. This anomaly was found to be present in placed against a pair of demibranchs, with the 55% of the mussels infested with Stage V crabs. carapace facing the center of the mantle cavity A similar condition has also been reported in and the frontal region oriented ventrally with Anomia simplex infested by Pinnotheres (Me respect to the mussel. Such a position insures Dermott, 1962a :163 ). As the erosion progresses that .the chela and mouth parts of the crab are dorsally toward the suspension of the crenidium, in a position which facilitates feeding . The crab the food groove is continually reconstituted. maintains this position by inserting the dactyls Without such a continuous regeneration it is of the pereiopods into the gill filaments and/or doubtful whether the crab-mussel relationship mantle tissues. could long endure, since the food groove is In general the feeding is, as described by necessary to the feeding process of both organ Orton (1921) for P. pisum, a matter of picking isms. Atkins (1931) has reported a similar the mucous food strings from the food grooves regeneration of the food groove following de of the crenidiurn upon which the crab is sitting. liberate mechanical damage to the gills of My The chelae are used initially to catch the strings tilus edulis. which are passed to the mouth parts. The an As reported by Stauber (1945:284) for P. terior pair of pereiopods are observed sometimes ostreum, the ctenidial damage inflicted by the to play a part in the manipulation of the food immature F. subquadrata, especially the Stage I strings. forms, differs markedly from that described for Prolonged feeding in this manner ultimately the adult crabs. Because the Stage I crab is results in extensive ctenidial erosion, as was re considerably smaller, and much flatter dorso ported as occurring in Crassostrea uirginic« ventrally, than the Stage V crab, it is able to (Stauber, 1945 :284) and Mytilus edulis (Me move about more extensively within the con Dermott, 1962a: 163) due to the presence of fines of the host. As a result of this movement P. ostreum. Ctenidial or gill erosion caused by the gill erosion is not restricted to the area be adult crabs involves the entire portion of the neath and in immediate contact with the crab, crenidium underlying the crab. This portion of but is found along the entire margin of the the gill is eventually destroyed. Figure 7 reveals ctenidium. Moreover, erosion of both ctenidia is this damage and shows the difference between common . These erosions quite often cause the Biology of Fabia subquadrata-PEARcE 25 FIG. 7. Photograph showing damage done to ctenidia of M. modiolus by a Stage V crab 9.3 mm in cara pace width. Note that the ctenidia lying in the left (lower) valve are eroded dorsally to the line of suspen sion. The ctenidia in the right (upper) valve are almost entire and serve for comparison. Cystlike anomalies can also be noted on the mantle tissues of the left valve. Y3 X . gill edges to appear serrated. Indentations caused infestation by a large adult crab. by immature F. subquadrata are only one-fourth It is noteworthy that only 58% of the small to one-third the dorso-ventral dimension of the crab infestations seen during the months of June gill, whereas the gill damage caused by the adult and early July are accompanied by extensive crab almost always appears as a single, crescent erosion, while Stage I infestations prior to and shaped erosion, often extending to the base of following this period are almost invariably ac the crenidiurn (Fig. 7). companied by extensive, small-crab type ero It appears likely that the erosion of a mus sions. This effect could quite possibly be due to sel's ctenidia always starts with the small-crab the infestation of previously uninfested mussels type and, as the crab matures, the damage be following copulatory swarming by the crabs, comes that typical of infestations with the adult which thus did not have time to cause extensive crabs. Christensen and McDermott (1958:171) erosion. note that the erosion caused by P. ostreum in Both kinds of damage appear to be about the American oyster also progresses from an equally detrimental to the individual host mus initial small-crab type damage to the more ex sel. Because the smaller Stage I or prehard crabs tensive destruction of tissues connected with the are usually associated with smaller mussels, the 26 PACIFIC SCIENCE, Vol. XX , January 1966 small-crab type erosion is prevalent in hosts of Within the waters surrounding the San Juan this size range. Similarly, while the small-crab Archipelago this relationship of the degree of type of erosion may not appear to be as exten infestation to the water depth appears to hold sive as the adult-crab type, it may well result in regardless of the geographic area from which as much, or more, relative surface area being the mussels are removed. Mussels from the removed from the ctenidia. In addition, the food shallow waters off Point Caut ion, San Juan groove in the former type of erosion may be Island, have more than an 80 % infestation mutilated over much of its entire length, and it just as do mussels removed from the shallow is this structure which is essential to the feeding waters off Point Lawrence 0:1 Orcas Island, and of the host mussel as well as the symbiont crab. Pea Vine Pass, all relat ively separated areas. Finally, small-crab damage, unlike adult-type Mussels removed from any of the deeper waters erosions, usually affects both ctenid ia, of President Channel always have a low per It is apparent from the foregoing that F. sub centage of infestation. quadrata is not the harmless commensal that it It was also found that the mussels themselves has been considered to be (W ells, 1940 :26; removed from the deep waters of President MacGinitie and MacGinitie, 1949:312) . As Channel were in very poor condition. The does P. ostreum this pinnotherid induces an gonads usually appeared atrophied, while the actual physical damage to the mussel host which visceral masses were, in general, very much re by its very nature must be harmful to some de duced. The valves of these mussels, while on the gree. Under normal conditions, however, the average the same approximate length as those crab is a very effective parasite. Its presence does from shallower waters, were thin, brittle, and not seem to affect the growth of or cause the much more subject to breakage than were those death of the host, nor does it appear to so of the shallow water mussels. weaken the host as to render its own position Wright (1 917) reports that Pinnotheres precarious. However, under other than normal never, or at the most very rarely, occurs in poorly conditions the effect of the relationship may be nourished mussels, although it was frequent in more serious. As Hopkins (1957:414 ) points those forms from areas where the host mussels out, ''.. . an organism which robs its host of were obviously well nourished and making rapid nourishment must be harmful in some degree, growth. even if the host shows no apparent effect, but Thu s it can be concluded that when host mus under favorable conditions it is probably not sels occur in an environm ent that is deficient difficult for the host to compensate or even over in some factor the infestation by the pinnotherid compensate for the loss by ingesting more food. crabs is reduced, either primarily or secondarily, Under conditions of food scarcity the same para by the same limiting factor. On the basis of site might become harmful. " known information it cannot be determined That such a situation may possibly occur can what the limiting factor might be in the present be shown by studies of the distribution of the case. Considering the poor physical condition of crab with regard to water depth. Mussels re the mussels found there, it is possible that there moved from relati vely shallow waters, i.e., 20 may be a deficiency in the amount or kind of 60 m, have a much higher percentage of infesta available nutrients in the deeper waters. tion than do mussels taken from waters at a A number of recent papers (Blake, 1960; depth of 200 m. Mussels collected from waters Haskin, 1940, 1950; Jan owitz, 1956 ) have in of intermediate depths (120-140 m ) have a dicated that certain gastropod predators are able percentage of infestation intermediate between to locate their prey on the basis of the latter's those typical of shallower and of deeper waters. relative metabolism. Apparently the diffusible As seen in Figure 1, mussels removed from 30 end products of the metabolic processes of the 60 m of water are consistently more than 80 % host species form gradients up which the preda infested. Mussels removed from waters 200 m tor is able to move. Similarly, Wilson ( 1948) and greater in depth are rarely more than 2% has recently demonstrated the ability of certain infested and frequently less than 1%. larval forms to "select" appropriate substrates Biology of Fabia subquadrata-PEARcE 27 upon which to settle. Therefore, it does no: unable to accommodate the acnvmes of more seem improbable that the settling first stage than one pinnotherid. It appears obvious that, crabs and the postswarming Stage I crabs might in the case of the F. subquadrata-M. modiolus use similar metabolic "targets" in selecting their relationship, an infestation by two adult crabs hos s. This not only would account for the lack would reduce the food gathering surface of the of infestation in the poorly nourished, and ctenidia to a level below the minimum required hence slower metabolizing, mussels found in the to sustain three organisms. Since double infesta deeper waters, but also could explain, as pre tions are not ever observed between the host viously discussed, the propensity to infestation and two immature crabs, it would appear that of smaller, and probably more rapidly meta the supposed mechanism operates below the bolizing, mussels by the early postplankronic adult level, i.e., at the first stage and/or Stage I first crab stages. levels. Finally, the author does not discount the fact As noted by Wells (1940: 34) , Stage V fe that the hazards encountered by these crabs, in males display a marked hostility toward each settling in deeper water, are much greater and other when placed together in finger bowls. that hence the low levels of infestation found in However, the present investigation has re such conditions may only reflect the loss incurred vealed that immature forms (even Stage IV in during the extended settling period. stars ), similarly situated, do not demonstrate the Houghton (1963:257 ) reports that there is marked aggressive behavior which characterizes a correlation between the incidence of infesta the adults' relationships. It is notable, however, tion of Mytilus edulis by P. pisum and the tidal that when two Stage I crabs are placed in a level at which the host mussels were collected. mussel one, and sometimes both, will imme He suggests that this is because the first true diately vacate the host. This was true in six crab stage of this species is photonegative, and replicate trials. On the other hand, when a single hence it is not likely that mussels found on the crab is inserted it will remain within the host. shore or at the surface on floats will be invaded. The relationships between the mussel crab One further testimony to the delicate balance and the alternative, smaller species of bivalve of this relationship is the fact that double in hosts already mentioned are not known as yet. festations occur only very rarely. As reported A cursory examination of these hosts did not re elsewhere in this paper only three such cases veal extensive damage to gills or other parts. were noted in this study. This is dramatically The infesting crabs were mostly immature pre different from the condition reported by Stauber hard crabs (94 out of 120 such infestations, or (1945:281) and Christensen and McDermott 78% ), and extensive damage probably would (1958: 155), who found that multiple infesta not have had time to occur. tions by immature P. ostreum , even of spat, were F. subquadrata has also been recorded as oc very common during certain periods of the life curring in Mytiltts edulis (Wells, 1928:289) , cycle. However, they did not ever observe two M. californianus (Wells, 1928:289; Ricketts posthard crabs occurring in the same host. Using and Calvin, 1952:164), and Saxidomus sp. the same survey techniques as those employed (Hart, personal correspondence). Ricketts and during my investigation of F. subquadrata, they Calvin report that the mussel crab is found in recently observed that frequently a single, adult 3% of the full grown California mussels. Giles Mytilus edulis is infested with up to six pre (personal correspondence) has found F. sub hard P. maculatus as well as with an adult fe quadrata in only 6 out of 805 M. californianus male. In this respect, then, the behavior of F. collected from Bodega Bay and Tomales Bay, subquadrata differs markedly from that of both California. This is less than 1% infestation. The P. ostreum and P. maculatus. present author has examined some 300 M. edulis The strong tendency toward single infestation and 104 M. californianus taken from the inter observed in the case of F. subquadrata certainly tidal zones of San Juan Island without finding a suggests some mechanism which selects against single mussel crab. multiple infestation of a host organism that is Hart (personal correspondence) has collected 28 PACIFIC SCIENCE, Vol. XX , January 1966 F. subquadrata from an unusual bed of inter subquadrata does not, at any stage in its life his tidal M. modiolus. About 18% of the mussels tory, engage in the multiple infestations which removed from this area (l ocated at Ten Mile occur during the early stages of P. ostreum. Point, Victoria, Vancouver Island, British Col While prev ious investigations (Christensen umbia ) are infested. and McDermott, 1958 ) suggest that swarming is not typical of members of the genus Pinno DISCUSSION tberes, the recent observations of swarming P. maculatus would indicate that at least one spe F. subquadrata, as do the other two pinno cies of this group takes part in a copulatory therid species which have been sufficiently swarming. It is suggested, therefore, that other studied, P. pisum and P. ostreum, has a complex pinnotherid species should be studied with re postplanktonic life cycle. The anamolous Stage gard to their reproductive behavior. This is I instar is present in the life cycle; and the pre particularly true in view of Sakai's paper (see hard and posthard instars, while not identical footnote 2 ) in which he menti ons a swarming with those of the two species of Pinnotheres, or migration as being characteristic of the Asi are quite similar. Also, as was demonstrated by atic pinnotherid, Tritodynamia horvathi. Miyadi Christensen and McD ermott ( 1958: 150 ) for (1 941 ) has described a benthic community on P. ostreum, the first true crab stage following the basis of a large number of pinnixid crabs the megalops is the invasive stage. found covering the bottom of certain areas of There are extensive differences, however, be the Ise-wan, Kii Peninsula, Japan. At one station tween the subsequent life cycle of F. subquadrata he reports that these crabs, Pinnixa rathbuni, and that reported for P. ostreum by the latter occur in densities of up to 3441/m2• They were authors. Present evidence indicates that the Stage found associated with several bottom types. I male oyster crab must leave its host and seek Since he thought that such a large number of out the female within her host in order to crabs cannot occupy a bottom area for an ex copulate. Following copulation the male leaves tended period of time, he suggested that among the female and perishes, either in the open other reasons, the crabs could be ". . . in the re water or within a secondary host. The results of productive period." As both Tritodynamia and the present study would indicate that copulation Pinnixa are relared -tin the subfamily Pinno in F. subquadrata occurs during a freeswimming thereliinae it is not improbable that the phe period, the swarming, in which both the male nomenon observed by Miyadi was actually a and the female participate. There is no indica swarming comparable to that observed by Sakai; tion that the male subsequently perishes. In fact, and, in fact, both might be associated with re following the swarming period in late May and production. Thorson (195 7:518 ) describes a June, males are frequently taken throughout crab community found in the Persian Gulf as the ent ire summer. Wells and W ells ( 1961: being a parallel of Miyadi's community. The for 275) have also noted the continued persistence mer community has as a dominant a pinnotherid, of males of Pinnaxodes floridensis following Xeno phthalamus pinnotheroides, which occurs copulation." in densit ies of up to 1,500 mature individuals/ In addition to the copulatory swarming and rn". Again, while Thorson suggests that this persistence of the male following swarming, F. community is stable, it is not impossible that 3 W hile at present the genus Pinnaxodes is often pl aced in a different subfamily (the Pinn othereliina e ) the Pinn oth ereliin ae. Members of the latter subfamily, from both Pinnotheres and Fabia (which are in the investigated in a recent study (Pearce, 1962b ), as subfamily Pinnothe rinae), there is some question as well as in the recent descriptions (W ells and W ells, to the validity of this arrangement. Rathbun (1918: 1961) of the life history and morphology of Pinna 179 ) states that Pinnaxodes tomentosus ". .. is very xodes [loridensis, differ very markedly in their life likely a Pinnotheres." A thorough study of the life history from both Fabia and Pinnotheres. Sakai history of the members of this genus will possibly (1939:582 ) , in his review of the Japanese Brachy indicate closer affinities with the subfamily Pinn o rhyncha, placed the genus Pinnaxodes, along with th erinae, including Pinnotheres and Fabia, than with Pinnotberes, in the subfamily Pinnotherinae. Biology of Fabia subquadrata-PEARcE 29 the dense population observed is actually a re no actual feeding or efforts to maintain a nor productive swarming. To the present author's mal environmental temperature, in the confine knowledge neither Miyadi or Thorson was able ment of a finger bowl until the following spring to study subsequently the respective areas, and (May 1960) . Similarly, many male Stage I crabs hence it is unknown whether or not these popu were held throughout the summer months at the lations were maintained. Friday Harbor Laboratories. Of particular sig The present investigations substantiate the nificance was the fact that eight such crabs did finding of Christensen and McDermott (1958: moult into soft, posthard forms. It is thought 174) that the soft-shelled, posthard females do that the slightly greater size of the male Stage I not normally leave their host. As indicated crabs may be a reflection of a growth moulting earlier, however, at least the immediate post which has previously been regarded as anoma hard insrars of Fabia are able to leave their dying lous or as not actually occurring. host, and a small Stage V Pinnotheres puget With regard to the crab-host relationship a tensis has been observed, and photographed, number of interesting conclusions can be drawn. leaving its ascidian host, Halocynthia igaboja. There is no doubt that the relationship between Another aspect which should be investigated F. subquadrata and M. modiolus is parasitic in further is the size differential between male and nature, especially if the broad definition of Hop female Stage I crabs of at least two species, and kins ( 1957:4 13) is used. The extensive, almost the possibly related phenomenon in which males ubiquitous, damage to the ctenidia as well as to of this stage were observed to moult into soft, the underlying mantle and palps cannot be con posthard forms as reported by the late Dr. At strued as anything but a result of a parasitic re kins (1958). She regarded these moults as hav lationship. As with some parasitic relationships ing possible significance in the growth of male (Allee et al., 1949) it seems to have developed crabs. The results reported by both Christensen with a degree of specificity. By this it is meant and McDermott (1958: 153) and the present that many species of pinnotherid crabs, includ investigation indicate a somewhat larger size for ing Fabia, are primarily found , at least in the the Stage I male than for the comparable female adult instar, in a single host species. There are, instar. The former found that the female P. however, exceptions to this generalizations, both ostreum ranges from 1.3-2.7 mm in carapace for Fabia and the other pinnotherid species.Al width, while the male ranges from 1.4-4.6 mm. though in the waters of Puget Sound adult Female F. subquadrata range from 1.5-6.2 and female F. subquadrata almost invariably occur average 3.5 mm (29 individuals), while the in the definitive or primary host, M. modiolus, males range from 1.3-6.8 and average 4.1 mm Wells (1928:289) reports it with both Mytilus ( 54 individuals). Atkins ( 1958) stated that edulis and M. californianus as well as in the hard Stage I males would moult into a soft stage branchial sac of the ascidian, Styela gibbsii. The which was frequently followed by two more crab found in the latter host was noted, how moults. Thus in P. pisum, which Atkins studied, ever, as being immature. During the present in one to three thin-shelled instars intervened be vestigation no crabs, of any stage, were found in tween successive thick-shelled or hard forms. either M. edulis or M. californianus. In the more Since Christensen and McDermott (1958:164) southern extensions of its range adult F. sub were unable to keep alive the male Stage I P. quadrata has recently been found in M. cali ostreum under their laboratory conditions, they fornianus. The status of the definitive host, M. could not observe such moultings. modiolus, in these waters is not known. Male Stage I F. subquadrata survive in the The closely related pinnotherid, Pinnotheres laboratory as long or longer than the same pugettensis, which from present information female instar. Such a crab was taken to the may have a life history very similar to that of Zoology Department of the University of Wash Fabia, was found only in the large ascidians, ington following the end of the normal summer Halocynthia igaboja and Ascidia paratropa. session at the Friday Harbor Laboratories (Au While Wells (1928:286) reports it only from gust 30, 1959). This crab survived, in spite of Halocynth ia (T ethyum) aurantium collected by 30 PACIFIC SCIENCE, Vol. XX , January 1966 Prof. Kincaid in the Friday Harbor region, the bunda. With three of the former worms he found present author was unable to find it in the P. eburna. Thus there is an indication that the limited number of specimens which were avail substratum, not the worm, determines the pres able for examination. All three of these tuni ence of the crab. During the present investiga cates are large and quite similar in their basic tion, however, several hundred worms of both morphology. species were subsequently checked and no anom A report by McDermott ( 1962a:163 ) also alies were found in the crab-worm association , contains information which indicates that, as in i.e., P. eburna was always found in association the Fab ia~Modiolus relationship, the oyster crab, with A . vagabunda and P. schmitti with A. pa P. ostreum, may develop through the hard Stage cifica. One final example of the specificity of the I instar in a secondary host, M. edulis. After at crab-host relationship is the relationship of Pin taining the hard stage ". . . both sexes leave the nixa fava and P. littoralis to the large larnelli hosts and seek some other molluscs (oysters or branch host, Schizothaerus capax. As with F. jingle shells ) in which the females may grow to subquadrata, the juveniles of these two pinni xids matur ity." McDermott suggests, therefore, that are found with a wide range of small bivalve it is possible for P. ostreum occasionally to species, although interestingly enough these are utilize two hosts in completing its life cycle. never the same species in which the immature In the same paper he reported that both P. Fabia occur. The adult P. faba and P. littoralis, ostreum and P. maculatus are able to develop to however, are found only in association with S. maturity in the jingle shell, Anomia simplex. capax. Even though closely related, and similar This is, as he notes, another new record for in size, Schizothaerus nuttali never contains the P. maculatus, and emphasizes further ". .. its adult pinnixids and rarely the immatures. The profound lack of host specificity." This pinno reason for this specificity is detailed in a recent therid has previously been reported from a wide paper (Pearce, 1962b :48). range of hosts ( Rathbun, 1918:76 ) , and since Thus, while many pinnotherid species may Grey ( 1961:357) has reported it from the tubes occasionally be found in other than the primary of Chaetopterus it must be assumed to be quite host during their early postplanktonic stages, widely distributed. the adults of most species so far studied appear While McDermott finds mature P. ostreum in typically to be found in a definitive or primary Anomia he emphasizes that "the incidence and host species or type. The reasons for this spe survival of P. ostreum in Anomia is not com cificity undoubtedly center in the fact that the parable to what we have seen in the oyster. The crabs have evolved in many ways to fit the incidence is much lower as is its survival to environment provided by a specific host or maturity." ganism . Apparent exceptions to the general rule, Representatives of the subfamily Pinnothere such as P. maculatus, must be more thoroughly liinae, which includes the several species of pin investigated. nixids found in Puget Sound, appear to be Although at the present time there is not a equally specific in the selection of, or survival in, great body of evidence from which to extra their hosts. On the tidal flats of False Bay, San polate, there are indications that the pinno Juan Island, two species of lugworm occur and therids, as a group, are actively involved in a each is associated with a particular species of process of adaptation. For instance, within the pinnotherid crab. Abarenicola pacifica, a worm subfamily Pinnotherinae there is considerable living in muddy sand, is almost invariably found variation in the use of secondary hosts. F. sub with Pinnixa schmitti; while A. vagabunda, a quadrata utilizes a wide range of bivalve species, species dwelling in the clean sandy bars protect of several families, for secondary hosts although, ing the entrance to False Bay, is usually found as earlier pointed out, in the waters of Puget with Pinnix» eburna ( Healy and W ells, 1959: Sound the adults have almost invariably been 325) . However, Healy reports in the same paper found in the definitive host, M. modiolus. an instance in which A . pacificawas found in the P. ostreum, upon occasion, will infest M. clean sand substrate favored by the A. vaga- edulis and A. sim plex (McDermott, 1962a ), Biology of Fabia subquadrata-PEARcE 31 although it apparently matures only in the latter range of latitude, is found in associatio n with secondary host, and here there is a noticeable many different types of hosts, including poly reduction in survival and percentage of infesta chaete worms, mussels, oysters, and scallops. tion when comp ared to the primary host rela Even in temp erate waters there is an advan tionship with Crassostrea virginica. In addition tage to the infestation of multiple host species, McDermott ( 1962b: 2) has observed that there if only by the immature crabs. In the event are annual fluctuations in the incidence of in of an epizootic involving the definitive host, festation of the secondary host bivalves by P. those individual crabs which have infested the ostreum, and these fluctuations should be studied secondary hosts during the postplanktonic to determine if there are any correlations with hard stage of their development would still be fluctuations or relative abundance of the primary available to reinfest the surviving, pr eviously host organism, in this case C. virginica. uninfested individuals of the primary host Finally we have the case of P. pisum which, population. according to evidence presented by Christensen Intergradations of morphology have been (1958), almost always develops, in the Kattegar, found between comparable instars of the species first in Spisula solida and then undergoes an so far studied. Both the first crab stage and Stage obligatory host change to M . modiolus. In a I instar of P. ostreum have rigid, well calcified more recent paper, however, Christensen ( 1962: exoskeletons and possess the rod-like structures 6) notes occasional exceptions to this general which connect the carapace to the sternum pattern : he has found an ovigerous female in a ( Christensen and McDermott, 19'58:150 ). In Spis ula and, in addition, has found several new Fabia the first crab stage is not hard, while the genera which can serve as the initi al host. Th ese Stage I form is; the latter stage of this species are Glycim eris, Cardium, Laevicardium, and does have the well developed columnar struc Mactra. tures linking the dorsal and ventra l surfaces of Thus, it would appear that there is a tendency the body. Similarly, in P. pisum the first crab for certain species to fill several niches, at least stage is not hard, while the Stage I insrar is. during the juvenile stages. Whether, in an evolu In addition to these variations in the hosts tionary sense, these species are progressing from utilized, and in the morphology of the invasive an original intraspecific or single-host condition instars, there are the differences already discussed to one in which a number of host species are in the reproductive biology of the various spe infested is not known. It might be suspected, cies. P. ostreem mates within the bivalve host of however, that while the adults of many of the the female while, from all available evidence, species thus far stud ied appear to be quite spe both F. subquadrata and P. maculatus engage in cific in their use of hosts, a definite advantage a freeswimming swarming during which copula would accrue to a species which was able to de tion occurs. While nothing definite is known velop in more than one host or niche. Th is about this aspect of the biology of P. pisum , col would be especially true of forms living in trop lections made by Christensen (personal corre ical waters, where a greater speciation has oc spondence ) ind icate simultaneous occupancy of curred but the total number of anyone species, the host bivalves by pairs (male and female) and hence of porential hosts, might not be so of Stage I crabs. On the other hand, the author great as in temperate or boreal waters. Sakai is in possession of a female, Stage I Pinn otheres ( 1939: 589) reports one species of Pinnotberes taylori taken freeswimming in a rnidwater as occurring in at least five different bivalves plankton trawl. found in Japanese waters (a lthough there is no Because of these intergradations between the statement as to the stages involved ) ; and the various species it is extremely difficult, as sug tropic al species listed by Rathbun (1 918 ) are gested by Christensen and McDermott ( 1958: frequentl y taken from several hosts. Again, how 177 ) , to generalize in any way with regard to ever, there are no statements as to the stage of the biology of the pinnotherid crabs. The fact the crabs. It is of interest that Pinnotb eres macu that in many species the Stage I female is modi latus, which is distributed throughout a wide fied for a freeswimming existence, and yet does 32 PACIFIC SCIENCE, Vol. XX, January 1966 not leave the host during this phase of the de sult of similar influences. A comparative quanti velopment, seems incongruous. It may be that tative study of the digestive tract contents of the modified morphology of the Stage I female host mussels and their symbiont crabs might be of P. ostreum is an adaptation for the possible of value in determ ining whether the limiting host changes which do occasionally occur in this factors include total available food. species. Christensen and McDermott ( 1958: From the foregoing discussion it is obvious 175 ) point out that when several female P. that the investigation of the family Pinnotheri ostreum invade the same host the excessive crabs dae is still as desirable today as it was several must either "perish or migrate to other oysters." decades ago, when Rathbun (1 918: 10) made According to Christensen's ( 1958) inform ation her admonishment concern ing the family. It is on P. pisum, it may be assumed that this is thought that, particularly from a systematic definitely the case in this species. It seems more point of view, a more intensive comparative reasonable, however, to regard the anomalous study of the biology of various species will pro Stage I instar, which seems common to many duce an unsuspected amount of information. pinnotherids, as a remnant of an earlier an N ot only will this information be of interest in cestral life cycle in which both the male and itself, but it might be profitably applied to the female were hard and occasionally freeliving, study of benthic and pelagic communities. The and both simultaneously infested the host or effects of the symbiotic crabs on their host or ganism. Later, during the evolution of the group, ganisms would undoubtedly influence the role when the symbiotic existence was definitely as of the latter within the community. This is espe sumed, the female started to moult into the cially true when the host is one of the dominants large soft, posthard stages capable of producing within a community. In fact, M. modiolus has the greater number of eggs necessary to the suc been considered a dominant organism in the cess and survival of a cryptic, parasitic form. At Modiolus faciation of the Strongylocentrotus the same time the male became less necessary, Argobuccinum biome (Shelford, 1935:287) and today, in at least some species, seems to be which is typical of the San Juan Channel. Pinno eliminated following copulation. It is suggested theres ostreum has been investigated with regard that this is a more reasonable hypothesis regard to its effects on the economically important ing the evolution of these forms than that pro oyster beds of the east coast of the United States posed by Christensen and McDermott. They (Stauber, 1945; Haven, 1958). Hancock (1962 ) state (1958:176) that posthard males, com has reported that infestation by P. pisum reduces parable to the existing female stages, probably the average volume of the meats of the edible existed somewhere in the line of evolution. Since mussel, Mytilus edulis. to this author's knowledge no known males of More recently a study has been initiated to any freelivin g brachyuran species assumes the ascertain the role which P. maculatus plays in soft habitus of the posthard pinnotherid females, the benthic mussel communities found in the it seems unlikely that such males were present waters surrounding Cape Cod and the Elizabeth in the early evolutionary history of the Pinno Islands. Some of the preliminary results of this theridae. study are contained in this paper. Finally, as was mentioned earlier, a definite correlation has been noted between the depth SUMMARY from which the host mussels are collected and the percentage of infestation, as well as between 1. The posrplanktonic stages which succeed depth and the size of the symbiont crabs. Pos the megalopal instar are described and their sible reasons for the correlation between the dimensions given. depth of water in which the mussels are found 2. Like Pinnotheres ostreum, Fabia subqua and the amount of infestation have been dis drata is shown to invade the host organism dur cussed previously. It is suggested that the smaller ing the first true crab stage. size of the crabs taken from mussels removed 3. "Abnormal" instars of both sexes are de from relativel y deeper waters (Fig. 1) is the re- scribed. Biology of Fabia subquadrata-PEARcE 33 4. The external manifestations of ecdysis in --- 1955. The post-embryonic development F. subquadrata are described. of British Pinnotheres (Crustacea). Proc. 5. Copulation occurs in open water, with both Zool. Soc. Lond. 124:687-715. the male and female crabs leaving their sym --- 1958. British pea-crabs (Pinnotheres). biont host as the hard , Stage I insrar, Following Nature 181: 1087. copulation the female crabs return to a host or BERNER, 1. 1952. Biologie de Pinnotberes pisum ganism to continue their development. Some, if Penn. (Decapoda, Brachyoure). Bull. Soc. not all, surviving males return to a host. Zool. France 77: 344-349. 6. Only during the copulatory swarming have BLAKE, J. 1960. Oxygen consumption of bivalve males and females been found together. Only prey and their attractiveness to the gastropod, three multiple infestations of the host have been Urosalpinx cinerea. Limnol. Oceanog. 5:273 noted ; two of these were between two males 280. and the third was between probably incom BROEKHUYSEN, G. 1941. The life-history of patible stages of a male and female. Cyclograpsus pun ctatus M. Edw.: breeding 7. Copulatory swarming occurs in Puget and growth. [Trans.] Roy. Soc. So. Africa, Sound during late May. This is followed by a Cape Town 28:331-366. period of 21- 26 weeks, during which the pre CARRIKER, M. 1955. Critical review of biology cociously inseminated females pass through the and control of oyster drills, Urosalpinx and five posthard developmental instars. Ovigerous Eupleura. U. S. Fish and Wildlife Ser., Spec. females are first noted in November; the eggs Sci. Rept., Fish. No. 148. hatch in February. CHRISTENSEN, A. 1958. On the life-history and 8. The growth rate of F. subquadrata ap biology of Pinno theres pisum . XV Inter. pears to be positively correlated with the growth Congo Zool., Sect. III, paper 15. rate of the definitive host, the horse mussel, --- 1962. Nogle parasitiske krabbers biologi. Modiolus modiolus. Suggestions are given to ex Naturens Verden, Kebenhavn. plain the fact that immature crabs are less com --- and J. McDERMOTT. 1958. Life-history monly associated with the relatively larger host and biology of the oyster crab, Pinnotberes mussels. ostreum Say. BioI. Bull. 114: 146-179. COE, W. 1948. Nutrition, environmental condi 9. Four new bivalve hosts are given for F. tions, and growth of marine bivalve mollusks. subquadrata. J. Mar. Res. 7:586-601. 10. Evidence is presented which suggests that DANA, J. 1851. Conspectus Crusraceeorum quae the mussel crab is a-true' parasite causing exten in Orbis Terrarum circumnavigatione, Carpio sive physical damage to the host organism. Wilkes e Classe Reipublicae FaederatoeDuce, 11. Crabs found in mussels removed from lexit et descripsit. Proc. Acad. Nat. Sci. Phila. deeper waters tend to be smaller than those re 5:247-254. moved from hosts taken in relatively shallow DARBISHIRE, A. 1900. Investigations made at waters. the Marine Biological Laboratory, Plymouth. Repr. Brit. Assoc. Adv. Sci., Bradford, p. 399. REFERENCES DAVENPORT, D. 1950. Studies in the physiology of commensalism, I. The polynoid genus Arc ALLEE, W., A. EMERSON, O. PARK, T. PARK, tonoe. BioI. Bull. 98:81-93. and K. SCHMIDT. 1949. Principles of animal --- 1953a. Studies in the physiology of com ecology. Saunders, Philadelphia. xii + 837 pp. mensalism, III. The polynoid genera Acboloe, ATKINS, D. 1926. The moulting stages of the Gattyana, and Lepidashenia. J. Mar. BioI. pea crab (Pinnotheres pisum). J. Mar. BioI. Assoc. 32: 161-173. Assoc. 14:475-493. --- 1953b. Studies in the physiology of --- 1931. Note on the regeneration of the commensalism, IV. The polynoid genera gill of Mytilus edulis. J. Mar. BioI. Assoc. Polynoe, Lepidashenia, and Harmothoe. J. 17:551-566. Mar. BioI. Assoc. 32: 273-288. 34 PACIFIC SCIENCE, Vol. XX, January 1966 DRACH, P. 1939. Mue et cycle d'intermue chez KAMITA, T. 1936. Crabs from the coast of Corea les crustaces decapodes. Ann. l'Inst. Oceanog. Strait. J. Chosen Nat. Hist. Soc. 21. [In Japa (n.s.) 19: 106-377. nese. Quoted from Sakai, 1939.} GREY, 1. 1961. Changes in abundance of the KNUDSEN, J. 1957. The act of molting in the commensal crabs of Cbaetopteras. Biol, Bull. California Xanthidae, the pebble crabs. Bull. 120:353-359. So. Calif. Acad. Sci. 58:133-142. GUYSELMAN, J. 1953. An analysis of the molt MACGINITIE, G., and N. MACGINITIE. 1949. ing process in the fiddler crab, Uca pugilator. Natural history of marine animals. McGraw BioI. Bull. 104: 115-137. Hill, New York. ix + 473 pp. HANCOCK, D. 1962. Starfish and mussels. Inter. MAERZ, A., and M. PAUL. 1930. A dictionary Coun. Expl. Sea, Shellfish Committee, paper of color. McGraw-Hill, New York. 207 pp. 83,7 pp. McDERMOTT, J. 1962a. The incidence and host HART, J. 1935. The larval development of Brit parasite relat ions of pinnotherid crabs (Deca ish Columbia Brachyura, 1. Xanthidae, Pinno poda, Pinnotheridae). Proc. First Natl. Coastal theridae (in part), and Grapsidae. Canadian Shallow Water Res. Conf., pp. 162-164. J. Res. 12:411-432. --- 1962b. Life-histories, biology, and host HASKIN, H. 1940. The role of chemotropism in parasite relations of the pinnotherid crabs. food selection by the oyster drill, Urosalpinx Second Annual Progress Rep t., Natl. Sci. cinera Say. Anar. Rec. 78:95. Found. Grant No. G-11448, pp. 1-5. --- 1950. The selection of food by the com McKAY, D. 1942. The Pacific edible crab, Can mon oyster drill, Urosalpinx cinerea Say. cer magister. Fish. Res. Bd. Canada, Bull. Proc. Natl. Shellfish Assoc., 1950, pp. 62-68. 62:1- 32. HAVEN, D. 1958. Effects of pea crabs, Pinno MERCIER, 1., and R. POISSON. 1929. Alteration , tberes ostreum, on oysters, Crassostrea vir de certaines caracteres sexueles secondaires du ginica. Proc. Natl. Shellfish Assoc., 1958, pp. male de Pinnotheres pisum 1. parasite par un 77-86. entoniscien (Pinnotberion vermiforme Giard HEALY, E., and G. WELLS. 1959. Three new et Bonnier). Bull. Soc. Zool. France 54:301 lugworms (Arenicolidae, Polychaeta ) from 304. the North Pacific area. Proc. Zool. Soc. Lond. MIYADI, D. 1941. Marine benthic communities 133:315-335. of the Ise-wan and the Mikawa-wan. Mem. HIATT, R. 1948. The biology of the lined shore Imp. Marine Obs., Kobe 7:502-524. crab, Pachygrapsus crassipes Randall. Pacific NEEDHAM, A. 1950. The form transformation Sci. 2:135-213. of the abdomen of the female pea-crab Pinno HICKOK, J., and D. Davenport. 1957. Further th eres pisum Leach. Proc. Roy. Soc. Lond. studies in the behavior of commensal poly B, 137:115-136. chaetes. BioI. Bull. 113:397-406. OLDROYD, 1. 1928. Marine shells of Puger Sound HOPKINS, S. 1957. Parasitism. In : Treatise on and vicinity. Publ. Puger Sound Biol, Sta. marine ecology and paleoecology, Geol. Soc. 4:1-271. Amer., Mem. 67, 1:413-428. HOUGHTON, D. 1963. The relation between OLMSTEAD, J., and J. BAUMBERGER. 1923. Form tidal level and the occurrence of Pinnotberes and growth of grapsoid crabs. A comparison pisum (Pennant) in Mytilus edu!is 1. J. Anim. of the form of three species of grapsoid crabs Ecol. 32:253- 257. and their growth at molting. J. Morph. 38: JANOWITZ, E. 1956. Food selection by Urosal 279-294. pinx cinerea (Say). Paper presented at Conv. ORTON, J. 1921. The mode of feeding and sex Natl. Shellfish Assoc. [Quoted from Blake, phenomena in the pea-crab (Pinnotberes 1960.} pisum ). Nature 106:533- 534. JOHNSON, 1. 1952. The demonstration of a PEARCE, J. 1962a. Adaptation in symbiotic "host-factor" in commensal crabs. Trans . Kan crabs of 'the family Pinnotheridae. Biologist sas Acad. Sci. 55:458-464. 45: 11-15. Biology of Fabia subquadrata-PEARcE 35 --- 1962b. The biology of some pi nnotherid STAUBER, 1. 1945. Pinnotberes ostreum, para crabs from the waters of Puget Sound and th e sitic on the Am erican Oyster, Ostrea (Gry San Juan Archipelago. Unpub!. Ph .D. thes is, phaea) virginica. Bio!. Bul!. 88: 269- 29 1. Univ. of W ash., Seattle, ii + 279 pp. THOMPSON, J . 1835. Memo irs on the metamor PHIFER, 1., and T. T HOMPSON. 1937. Season al phosis and natural his'ory of the Pinnotheres varia tions in the surface waters of San Juan or pea-crabs. Entorn, Mag. Lon d. 3:85-90. Channel during the five year period, January, THORSON, G. 1950. Reproductive and larval 1931 to Dec ember 30, 1935. J. Mar. Res. ecology of marine bottom invertebrates. Biol, 1:34-59. Rev. 25 : 1-45. RATHBUN, M. 1914. Stalk-eyed crustaceans col --- 1957. Bottom communities . In : Treat ise lected at the Monte Bello Island s. Proc. Zoo!' on marine ecology and paleoecology, Geol. Soc. Amer., Mem. 67, 1:461-534. Soc. Lond . 2:653-665. WELLS, W . 1928. Pinnotheridae of Puget Sound. --- 1918. Grapsoid crab s of Ameri ca. U. S. Pu b!. Puget Sound Biol, Sta. 6:283- 314. N atl. Mus., Bull. 97:1-461. --- 1940. Ecological studies on the p inno RICKETTS, E., and J. CALVIN. 1952. Between therid crabs of Puget Sound. Univ. Wash. Pacific tides. Stanford University Press, Stan PubI. Oceanog. 2: 19-50. ford, Calif. xiii + 502 pp. WELLS, H ., and M. W ELLS. 1961. Observat ions SAKAI, T. 1939. Studies on th e crabs of Japan, on Pinnaxodes floridensis, a new species of IV. Brachygnatha, Brachyrhynca. T okyo (publ. pinnotherid crustacean commensal in holo by author). pp. 365- 741. thurians. Bull, Ma r. Sci. Gulf Carib. 11:267 SANDOZ, M., and S. HOPKINS. 1947. Early life 279. histor y of the oyster crab, Pinnotheres os WIBORG, K. 1946. Undersokelser over oskjeller treum (Say ). Bio!. Bul!. 93 :2 50-258. (M odiolus modiolus (L». Fisk. D irekr, Skri f. SASTRY, A., and R. MENZEL. 1962. Influence of 8: 1- 85. hosts on the behavior of th e commensal crab WILSON, D . 1948. The relatio n of the sub Pinnotberes maculatus Say. Bio!. Bull. 123: stratum to the metamorphosis of Ophelia lar 388-395. vae. J. Mar. Bio!. Assoc. 27:723-760. SHELFORD, V. 1935. Some marine biotic com WRIGHT, F. 1917. Mussel beds ; their produc munities of the Pacific Coast of N orth Amer tivity and maintenance. Ann. Applied BioI. ica. Ecol, Mon og. 5:249- 354. 4: 123-125.