Sunteți pe pagina 1din 9

International Association for Ecology

Filter Feeding in the Hermit Crab, Pagurus bernhardus Author(s): S. A. Gerlach, D. K. Ekstrm and P. B. Eckardt Reviewed work(s): Source: Oecologia, Vol. 24, No. 3 (1976), pp. 257-264 Published by: Springer in cooperation with International Association for Ecology Stable URL: . Accessed: 09/10/2012 07:19
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact

Springer and International Association for Ecology are collaborating with JSTOR to digitize, preserve and extend access to Oecologia.

Oecologia (Beri.) 24, 257-264 (1976)

tJCCOlOgUl ? by Springer-Verlag 1976

Filter Pagurus






bernhardus D.K. Ekstrom, and P.B. Eckardt

S.A. Gerlach,

Marine Biological Laboratory, Helsing0r, Denmark

Summary. The hermit crab, Pagurus bernhardus is able to remove both ArteCrabs with wet weights mia nauplii and unicellular algae from suspension. all of the 300 Artemia nauplii contained of 1.1-9.2 g consumed in 200 ml 0.7-1.1 g wet weight filtered susof sea water within 1 h. Crabs weighing of 10-350 million cells pended Dunaliella algae (8 ???) from concentrations a of 2 liter at rate and within and 26% 47% 5 A similar h, per respectively. result was obtained with an 11 g crab. During filter feeding activity a water current is generated by the flagella of the exopods of the second and third Artemia nauplii are caught by grasping movements of the endomaxillipeds. whereas filtering of unicellular algae is probably pods of the third maxillipeds, achieved by the two maxillae. It is assumed that filter feeding activity depends on the same structures and behavior as in deposit feeding. P. bernhardus is one more example of a benthic marine animal which may use any food source which becomes available in the course of the seasons.

Introduction Even It is well known that hermit crabs of the genus Pagurus are omnivorous. can small seize if they are not in general particularly voracious, many species or morsels of animals or crustaceans and polychaetes, bivalves, echinoderms, the to with the lift them part of the third proximal chelipeds plant material, A second source and mandibles. them between and tear maxillipeds maxillipeds, of food, and probably the most important one, is the organic content of sediments. The means of gathering the sediment varies from species to species. Most commonly this is accomplished by the minor left chela which tosses this action Sometimes of the third maxillipeds. the sediment to the endopods is preceded by a pushing of sand forward with the first and second pereiopods, of the third maxillipeds and in other cases the endopods directly sweep the mouth the smaller to of The final transfer substrate. parts is acparticles This the of the second pattern of behavior by maxillipeds. endopods complished


S.A. Gerlach et al.

has been described in P. bernhardus (L.) by Jackson (1913) and by Orton (1925). Thorson up 0.8 ml of sand (1966) reports that a 2.8 g hermit crab shoveled study of this mode of feeding, and the role of the per minute. A thorough in P. longicarpus and P. pollicaris was given different mouth parts involved by Roberts (1968). There are some general remarks in the literature on filter feeding by hermit crabs. Yonge (1949, p. 158): "in their feeding habits they show an interesting blend of the omnivorous of the true crabs and the filtering mechascavenging nisms of the allied porcelain crabs; they are thus capable of taking food from the bottom and from the surrounding waters". Brightwell (1951, p. 280), referring to the feeding of Pagurus bernhardus: feeding by "setting up a current, apparently " " thus fine particles by the breathing apparatus and the third maxillipeds, wafting of food toward itself". G.E. and N. MacGinitie (1968, p. 297), referring generally to hermit crabs: "some are able to feed by fanning detritus or plankton out of the water with their modified mouth parts". However, all detailed references to suspension feeding by hermit crabs have been on such genera as Stratiotes and which they Diogenes which have long plumous setae on the second antennae, use to filter the water. This is a report of preliminary observstions and experiments that indicating in addition to being a carnivore and a deposit feeder, P. bernhardus can derive considerable amounts of food by suspension feeding.


Hermit crabs were dredged from various parts of the Oresund and kept in a tank connected to the laboratory sea water system. A salinity of 31%0 and a temperature of 9? C were held constant throughout the course of the experiments conducted in January 1976. The gastropod shells inhabited by hermit crabs were cleaned of epifauna prior to experiments. Larval stages of the brine shrimp, Artemia salina, and cultures of the unicellular alga, Dunaliella marina, were used as food. The algae originated from cultures in the exponential phase of growth, with concentrations ranging from 0.8 to 1.4 million cells per ml1. Before use the culture was centrifuged and the cells resuspended in sea water. Cell concentrations were counted with an improved Neubauer Haemacytometer. The crabs did not get food for a period of 1-3 days prior to the experiments. During the experiments, the animals were kept separately in 400 ml plastic bowls, containing 200 ml of sea water. A single, large animal was kept in a 1-1 glass jar filled with 500 ml sea water. A plastic net, mounted about 10 mm above the bottom, prevented crabs from ingesting any particles which might have settled during the experiments. However, no such settling was observed. The water in the experimental vessels was neither aerated nor stirred. After the experiments, the crabs were killed and their stomach and gut contents were analyzed microscopically. As killing in alcohol sometimes resulted in vomiting and loss of stomach contents (as well as being painful to the animals), the crabs were narcotized with solutions of MgS04 or MgCl2, of which the latter proved to be the most effective. To observe the current flow, droplets of carmine suspension were placed at various positions in the proximity of the anterior part of the animals. Great care was taken to ensure that the droplets were stationary when released from the pipette, until caught by a current originating from the crab. 1 We want to thank Mrs. cand. mag. Grete Moller Christensen for cultures of Dunaliella marina

Filter Feeding in the Hermit Crab, Pagurus bernhardus Results


with Artemia nauplii (Table 1) demonstrate that Pagurus bernhardus Experiments can filter over 300 nauplii from 200 ml sea water within 1 h, very effectively and over 400 nauplii within 2 h. Directly after feeding, the bristles of Artemia were found in the stomach of the crabs. Fecal pellets collected 10 h after feeding also contained such bristles. to 1 mg organic dry weight (PafSeventy to 100 Artemia nauplii correspond h within 2 about 5 food had been ingested, fenh?fer, 1968); mg by animals of about 300 mg dry weight. We conclude that P. bernhardus is capable of feeding Artemia nauplii were caught by grasping movements on zooplankton. of the of the third maxillipeds which have long setae. Food particles are endopods then brushed off from the third maxillipeds of the second by the endopods and are finally transported into the region of the smaller mouth maxillipeds, which move rhythmically. appendages

Table 1. Filtering efficiency of Pagurus bernhardus when feeding on Artemia nauplii. Experiments were performed in 200 ml of sea water. Three hundred Artemia nauplii were offered at start, 150 more added after 1 h Weight of crab g (wet) 1.15 1.28 1.46 9.24 Artemia left from 300 after 1 h Artemia left from 450 after 2 h 28 15 45 0

Table 2. Filtering efficiency of Pagurus bernharduswhen feeding on suspensions of Dunaliella algae for a period of 2 h (figures in brackets : 5 h). Experiments were performed with 200 ml of food suspension (except the last experiment : 500 ml) Weight of crab g (wet) Concentration of algae suspension 106 cells/1 start 0.85 0.85 0.78 1.07 0.67 0.86 0.67 1.02 11.36 351 100 98 98 38 38 9.4 9.4 100 2h (5 h) Cells filtered out 106 cells 2 h (5 h) 6.2 8.0 (9.6) 2.4 4.9 2.5 2.6 0.3 0.9 15.0 (23.0) In % of cells offered Algae found after the end of the experiment

2 h (5 h)

in stomach many few many few plenty many plenty plenty

in midgut

321 60 (52) 86 74 25 25 7.8 5.0 70? (54)

39.8 (47.8) 12.2 24.9 33.3 34.4 17.0 46.0 30.0 (40.0)

plenty many plenty none none

densely packed densely packed


S.A. Gerlach et al.

with Dunaliella algae (Table 2) give evidence that a reasonably Experiments of the suspended algae can be filtered out of the water within large percentage 2 h. Filtering efficiency ranges from 9 to 46%, with a mean of 26%. Since the methods applied for counting algae were rather crude (single samples, no much to too stirring prior sampling), emphasis cannot be laid on these figures, decrease in the number of cells which only serve to indicate that a significant is clearly shown in Table 2, in suspension did occur. This general tendency both in low concentrations of 10 million cells per liter, and in very high concentrations of 350 million cells per liter. Filter feeding is not restricted to small crabs : a specimen 11 g weighing caused a 30% reduction of algae in 2 h. Measurements taken in 5-h experiments showed a continuous uptake of algae. Large quantities of Dunaliella cells were found in the stomachs of all experimental crabs, except for some which vomited prior to dying. We have no proof that algae were digested in the diverticula of the gut. After 2 h of feeding only one animal had large quantities of algae in the midgut ; however the midgut of crabs allowed to feed for 5 h was filled with both whole algae and fragments of algae. We do not know, again, to what extent food is utilized in the midgut of P. bernhardus; in this species the non-cuticularized midgut is rather long with other 18 million cells of (Jackson, 1913). Paguridae Roughly compared

exopods of second and third maxillipeds

first antenna

respiratory currents

endopod of third maxilliped

endopod of second maxilliped

Fig. 1. Schema of antennae and mouth parts in Pagurus [modified from Roberts (1968), Fig. 4: Pagurus longicarpus], and water currents observed during filter feeding activity

Filter Feeding in the Hermit Crab, Pagurus bernhardus


8 ?p?) have a dry weight of 1 mg (Winter, Dunaliella 1969); one (diameter hermit crab of about 200 mg dry weight could take 0.14 mg food dry weight of about 40 million cells per liter, and 0.02-0.05 from a concentration mg food from a concentration of about 10 million cells per liter, within 2 h. From this calculation does not provide it seems evident that filter feeding on phytoplankton a large quantity of food per unit of time, but that algae might be considered as an additional source of food. water current is provided During filter feeding activity, a nearly constant the of and of the the second third which normally by flagella exopods maxillipeds beat in unison. The flagella of the right and left side occasionally work together in short bursts of activity, but mostly those of one side are working. The current thus created comes directly from in front of the animal when flagella of both sides work together, and from the opposite laterofrontal area, when only In all cases a current of 1-2 cm/s flows those of the one side are working. upward and slightly frontally away from the animal (Fig. 1). Tests with carmine revealed that the incoming current apparently is composed of two layers. The upper layer curves upward in front of the animal and it to the first antennae ; seems that its main purpose is to bring chemical information they repeatedly dip into the current and then are brushed off by the endopods The lower layer of the current curves upward directly of the third maxillipeds. beneath the mouth area and passes across the setae of the third and second thus reaching the direct proximity of the smaller mouth appendages. maxillipeds, Large quantities of carmine particles, which have a size range similar to Dunaliella algae, were found in both the stomach and gut of the crabs after the selective while filter It seems that P. bernhardus is not particularly experiments. feeding on smaller particles. current enters between shell and crab on the left the respiratory Normally side of the animal. The water is expelled from the gill chambers by the paddlelike The exhalant currents join the outward water motion of the scaphognathites. flow created by the action of the macilliped flagella.

Discussion movements The normal respiratory current of Pagurus is created by paddlelike But there are at of the second maxillae. of the scaphognathites, appendages least five more ways for a hermit crab to create water currents: 1) grasping of the of the third maxillipeds, of the endopods movements 2) the beating of the second and third maxillipeds, 3) the beating flagella of the exopods and of the flagella of the first maxillipeds, 4) the beating of the left pleopods, currents the described Brock shell. in the of abdomen the movements (1926) 5) since as orientational, made by the flagella of the second and third maxillipeds first antennae. for the stimuli chemical they provide According However, the function of these water currents may be many-sided. material and fecal pellets out of range to Roberts (1968) they expel indigestible that water one might conclude of the mouth parts, and from our experiments role currents generated play a significant by the flagella of the maxillipeds


S.A. Gerlach et al.

in filter feeding. However, we cannot exclude the possibility that P. bernhardus is capable of filter feeding by using more than one type of current. Possibly even the normal respiratory current may be filtered through the smaller mouth parts. Artemia nauplii are caught by grasping movements of the endopods of the third maxillipeds ; it seems that this is the same kind of movement as that displayed when hermit crabs and other Anomura sort light (Galathea, Lopholithodes) out of sediment. organic particles suspended In Porcellana movements have been described as filter grasping, horizontal In the of the third cannot, however, Pagurus feeding activity. activity maxillipeds explain the filtering efficiency with minute particles like Dunaliella algae, because the denticulate setae on the endopods of the third maxillipeds are rather coarse. Like Eremita some hermit crabs {Diogenes Stratiotes brevirostris, setosus) have long plumous setae in two rows on the long second antennae. Usually these species feed from the sediment like other hermit crabs, but if food particles are in suspension, the second antennae begin rapid to and fro movements (cast-net feeding, Boltt, 1961 ; Greenwood, interspersed with cleaning movements the antennal filter mechanism is rather coarse, in Diogenes 1972). However, brevirostris the plumous setae are 3 mm long and set 0.2 mm apart. Pagurus has only simple setae on the second antennae, and they are shorter than the width of the segments. Roberts that an auxiliary (1968) therefore concludes mode of feeding with the antennae is not possible in Pagurus. Zoea larvae ?? Pagurus longicarpus, reared in the laboratory, capture nauplii of Artemia as well as post-trochophores of Arenicola, using the maxillipeds when encounters behavior. Plankton happen during normal swimming algae have been found in the gut of these zoea larvae too, but details of ingestion could not be observed (Roberts, 1974). It is difficult to see what the smaller mouth parts really do during feeding the endites of the appendages make backward and forward activity. Apparently in such a way that, as the second maxillae movements move forward, the first maxillipeds and the first maxillae move backward (Roberts, 1974). Coarse sand grains are brushed free from epigrowth with spines and setae on the coxal and basal endites. Material to be discarded is moved forward until it is cast away by the flagella. In P. pollicaris setae form a band long plumous on the coxal endites of the second maxillae, and in Pagurus longicarpus such setae are found on the first maxillae, too. These setae may serve to prevent material loosened from sand grains from escaping the buccal area (Roberts, We think that on small 1968). suspension feeding particles like Dunaliella algae is effected by a water current passing between first and second maxillae. This would mean that one filtering mechanism is used 1) for sorting food particles from sand grains, 2) for sorting food particles from suspended and sediment, further observations 3) for sorting food particles from suspension. However, are necessary to obtain a real understanding of the filter feeding apparatus of Pagurus bernhardus. From our experiments it is evident that a hermit crab in a few hours can collect enough zooplankton to fill stomach and gut, provided the zooplankton

Filter Feeding in the Hermit Crab, Pagurus bernhardus


in retaining is high enough. The efficiency is concentration phytoplankton of algae which correspond to naturally occurring poorer, and from suspensions concentrations (9 and 38 million cells per liter) less than 0.15 mg phytoplankton food (dry weight) was gained by crabs of 0.7-1 g wet weight, within 2 h. Organisms which exclusively depend on phytoplankton, like bivalves, have much higher efficiencies under similar laboratory A 1.5 g (wet weight) mussel, conditions. Modiolus modiolus, at 12? C and with a food concentration of 20 million Chlamydomonas cells per liter, filters out nearly 10 million cells per hour, corresponding to about 0.8 mg dry weight. The large 11 g hermit crab of our experiment which, in 5 h, filtered algae of 1.2 mg dry weight out of the water (containing 100 million cells per liter) might be compared with an 11 g Modiolus modiolus which filtered about 3.5 mg algae in 1 h (Winter, 1969). Feeding on plankton seems to be of minor for P. bernhardus, but it might algae only importance is very dense. be important when other food is scarce, or when phytoplankton is that it is still an how valuable the contribution However, open question P. could reared from of be come Zoea larvae might longicarpus phytoplankton. on a diet of Artemia or zooplankton, but a diet of algae w^s not sufficient for complete development, although there was evidence of longer survival compared with starved animals (Roberts, 1974). There are more decapod crustaceans lobster with filter feeding capacities. Jatzke (1970) reports on a 6-month-old but without access (Homarus gammarus) which lived under in situ conditions, to larger food items. During 17 months it grew like other lobsters which were fed in the laboratory. of a benthic The hermit crab, P. bernhardus, provides one more example animal's ability to rely on more than one feeding type, so that it may use optimally any food source becoming available to it in the course of the seasons.

References Boltt, R.E. : Antennary feeding of the hermit crab Diogenes brevirostris Stimpson. Nature (Lond.) 192, 1099-1100(1961) Brightwell, L.R.: Some experiments with the common hermit crab (Eupagurus bernhardus Linn.), and transparent univalve shells. Proc. zool. Soc. Lond. 121, 279-283 (1951) Brock, F.: Das Verhalten des Einsiedlerkrebses Pagurus arrosor Herbst w?hrend der Suche und Aufnahme der Nahrung. Z. Morph. ?kol. Tiere 6, 415-552 (1926) Greenwood, J.G. : The mouth parts and feeding behaviour of two species of hermit crabs. J. nat. Hist. 6, 325-337 (1972) Jackson, H.G.: Eupagurus. Proc. Trans. Lpool biol. Soc. 27, 495-573 (1913); and in: L.M.B.C. Memoirs on British Marine Plants and Animals (W.A. Herdman, ed.), Vol. 21, pp. 1-79. London: Liverpool Marine Biological Committee 1913 Jatzke, P. : The Trichterkreisel, an in situ device for cultivating marine animals in tidal currents. Helgol?nder wiss. Meeresunters. 20, 685-690 (1970) MacGinitie, G.E.: Natural history of marine animals, 2nd ed., pp. 1-523. New York: McGraw Hill 1968 Orton, J.H. : On the mode of feeding of the hermit crab, Eupagurus bernhardus, and some other Decapoda. J. mar. biol. Ass. UK 14, 909-921 (1927) Paffenh?fer, G.-A. : Nahrungsaufnahme, Stoffumsatz und Energiehaushalt des marinen Hydroidpolypen Clava multicornis. Helgol?nder wiss. Meeresunters. 18, 1-44 (1968)


S.A. Gerlach et al.

Roberts, M.H. : Functional morphology of mouth parts of the hermit crabs, Pagurus longicarpus and Pagurus pollicaris. Chesapeake Sc. 9, 9-20 (1968) Roberts, M.H. : Larval development of Pagurus longicarpus Say reared in the laboratory V. Effect of diet on survival and molting. Biol. Bull. mar. biol. Lab., Woods Hole 146, 67-77 (1974) Thorson, G. : Some factors influencing the recruitment and establishment of marine benthic communities. Neth. J. Sea Res. 3, 267-293 (1966) Winter, J.E. : ?ber den Einflu? der Nahrungskonzentration und anderer Faktoren auf Filtrierleistung und Nahrungsausnutzung der Muscheln ?rctica isl?ndica and Modiolus modiolus. Mar. Biol. 4, 87-135 (1969) Yonge, CM.: The sea shore, pp. 1-311. London: Collins 1949

Received March 7, 1976