Broughton page 1 I. INTRODUCTION The hermit crab Pagurus samuelis (Stimpson, 1857) was observed on the column, tentacles and oral disc of the sea anemones Anthopleura elegantissima (Brandt, 1835) and Anthopleura xanthogrammica (Brandt, 1835). The crab was apparently not stung by the anemone's nematocysts. Symbiotic relationships have been described between anemones and fish (Mariscal, 1966), anemones and shrimp (Dales, 1966), and anemones and crabs (Davenport, 1966; Dales, 1966) in which the fish or crustacean is able to move among the anemone's tentacles without being stung by nematocysts. This investigation examines the behavior of Pagurus samuelis on Anthopleura elegantissima and A. xanthogrammica, the crab's behavior in becoming protected against nematocyst discharge, and possible mechanisms of protection from nematocyst discharge. II. GENERAL MATERIALS AND METHODS All field and lab observations were carried out at Hopkins Marine Station, Pacific Grove, California. A. elegantissima and A. xanthogrammica were collected from the Marine Station reserve and from Pescadero Point on the Monterey Peninsula. Pagurus samuelis, which live between the 0.0 and 1.5 foot tide level in crevices and pools, were collected from the Marine Station reserve. Crabs described as "isolated" were collected from an isolated intertidal pool which Broughton page 2 has no anemones within 5 meters of it. Crabs described as "protected" did not stick to anemone tentacles and did not jerk back upon contact with tentacles. A crab was tested for protection by holding its shell near anemone tentacles and observing the reaction when the crab came out of its shell and encountered the tentacles. Crabs were always handled with blunt forceps. Anemones and crabs were kept in the lab in clear plastic tubs with running sea water. The animals were not fed but remained active and responsive during the course of the study. All night observations were conducted with red light. III. BEHAVIORS OF HERMIT CRABS ASSOCIATED WITH ANEMONES A. Observations Each of the following types of behavior was observed both in the field and in the lab and both during the day and at night. The crab's interaction with the anemone was found to consist of distinct behavioral units which are described below. 1. Nestling. A study was made of the distribution of the crab-anemone relationship during the day in thirteen one meter square quadrats at the 0.5 to 1.5 foot tide level when covered by 1 to 2 feet of quiet water. Of the 262 hermit crabs observed, 17% were sitting quiescently nestled against anemones. Of the 58 solitary A. elegantissima and A. xanthogrammica under water in the quadrats, nine anemones (15.5%) Broughton page 3 had crabs nestled against their columns. Of these nine, six anemones had one crab against them, and the three other anemones had 2, 11 and 12 crabs nestled against their columns. Of the 28 solitary anemones observed out of water, none had crabs nestled against their columns. Of approximately 430 clonal A. elegantissima individuals observed, four hermit crabs were seen nestled against individuals. Fourteen anemone-crab relationships were set up in the lab: five tanks with clonal A. elegantissima, two with a solitary A, elegantissima and seven each with an A. xanthogrammica. Crabs were observed during both the day and night nestling against the clonal A. elegantissima and the solitary A. elegantissima and three of the seven crabs with A. xantho- grammica were observed nestling. 2. Walking on the column. Crabs were observed walking onto the column of anemones and exploring it with their antennae and appendages. Using its chelipeds. a crab often picked up detritus from the column and brought the detritus to its mouth. Crabs sat quiescently on columns only occasionally. In the previously described distribution study, 14.4% of the 262 hermit crabs observed were found walking on the columns of anemones. Of the 58 solitary anemones under water, 8 had at least one crab walking on their column: two anemones had one crab, two had three crabs, and the remaining four anemones had 2, 4, 7 and 8 crabs on their column. Of the 28 anemones observed which were out of water, one anemone had a hermit crab walking on its column; this was the only hermit crab Broughton page 4 observed interacting with an anemone above water in the quadrats, Of the fourteen lab relationships, all crabs but two, one with an A. xanthogrammica and one with an A. elegantissima, were observed walking on the columns of their anemones. All hermit crabs, including the 17 isolated crabs observed in initial interactions with anemones, were able to walk on the columns without being stung, and the anemones made no response to the presence of crabs on their columns, 3. Poking tentacles. Crabs were observed poking at tentacles with the tips of their chelipeds and walking legs for periods of between 5 and 30 seconds. The tentacles made no response, just as they made no response to the gentle pokes of a straight pin. 4. Sweeping across tentacles. In encountering tentacles, the crab would first poke the tentacles as previously described and secondly sweep its antennae then several appendages, both chelipeds and walking legs, over the tentacles. In 5 out of 23 observations made of this behavior, the crab went directly to the sweeping behavior without first poking the tentacles. In response to the sweeps, the tentacles would move in towards the mouth slightly such that the tentacles were no longer drooping onto the column or substrate. Within thirty seconds, however, the tentacles would droop back into their original position unless the crab again swept its appendages over the tentacles. With small anemones, continued sweeping resulted in the anemone closing completely. The Broughton page crab would then walk onto the anemone and turn all the way around the anemone, removing detritus from the column, or would shove its cheliped into the gastric cavity as will be described later. 5. Stroking tentacles. Crabs were observed using their chelipeds to stroke tentacles from the base to the tip. The crab would move rapidly from one tentacle to a neighboring tentacle. In response, the tentacle would not contract but would move to where the crab pushed it and then spring back to its original position. 6. Shoving cheliped into gastric cavity. Crabs were observed walking onto the column and tentacles of an anemone less than 5 cm. in diameter, sweeping their appendages across the tentacles until the anemone was closed and then shoving their chelipeds into the gastric cavity of the anemone. The crab would usually bring its cheliped to its mouth upon withdrawing it from the gastric cavity. Upon withdrawal of the cheliped the anemone would begin to open up but would again close if the crab re-inserted its cheliped. This activity was observed to continue for periods of between and 10 minutes. Crabs were observed climbing into the opening of the anemone, almost disappearing from sight, only to climb out within several seconds. Crabs were seen climbing into the mouth opening of already-closed large anemones in only five cases of the over 100 crab-anemone interactions observed in the field. In the twelve lab set-ups with crabs and large anemones this behavior was Broughton page 6 never observed. In the field Mytilus califorianus tissue was fed to a 15 cm. diameter solitary A. elegantissima which was almost closed, and within five seconds three Pagurus samuelis on the column of the anemone ran to the edge of the mouth opening, shoved their chelipeds into the mouth opening, and removed and ate the Mytilus tissue. 7. Eating anemone tissue. In the lab, two hermit crabs from the isolated tidepool and three crabs found in the field on the column of A. xanthogrammica ate excised tentacle, column and pedal disc tissue of both A. xanthogrammica and A, elegantissima. A lab P. samuelis ate a tentacle excised from the lab A. xanthogrammica which it often nestled against and sometimes stroked the tentacles of. The stomach, intestine and feces of isolated. field associated and lab associated crabs were examined three and one-half hours after the crabs were given anemone tissue. Zooxanthellae (algae symbiotic in anemone tissue and mucus) and three to ten undischarged nematocysts were found in each digestive system component, In lab and field observations, chelipeds were never observed to "snip off" tentacles and damaged tentacles were never seen. However, in the lab crabs were observed using their chelipeds to pull at damaged pedal disc tissue of clonal A. elegantissima and damaged column tissue of an A. xanthogrammica. Large chunks of tissue were never seen in the chelipeds but the crabs did bring their chelipeds to their mouths. In the lab and in the field crabs sitting on small closed anemones were observed placing their chelipeds against the base of the anemone and rocking backwards, thereby releasing a small part Broughton page 7 of the pedal disc from the substrate. The crab would then shove its cheliped under the pedal disc and bring the cheliped to its mouth, It was difficult to observe what the crab was holding in its chelipeds, however. The stomach, intestine and feces of a P. samuelis were examined one and one-half hours after it was found stroking the tentacles and climbing into the mouth area of a 5 cm. A. elegantissima. Zooxanthellae, undischarged nematocysts and undischarged spirocysts were found in the crab's intestine. Up to ten nematocysts/spirocysts and 50 zooxanthellae were found in one 450x power microscopic field of the intestinal material. It should be noted here that nematocysts and zooxanthellae are found in both anemone tissue and mucus. 8. Protection from threat. Threatened crabs on anemones remained on their anemone, backing up around the anemone in the lab and wedging themselves between the anemone and a rock in the field. In ten lab tanks with an anemone and crab in each tank, the crab was removed from the tank to test for protection against nematocysts and then replaced in the tank after a few minutes. In six of the ten tanks the crab immediately ran to sit beside the anemone or on the base of the anemone's column. In five of these six tanks the crab had not been observed near the anemone for several days previous to this encounter. Broughton page 8 B. Discussion The interaction between hermit crabs and sea anemones may provide several benefits to both parties. By nestling, crabs may be protected against dessication and predators. Anemones may also receive some protection from dessication by having shells against their columns. Crabs are probably obtaining nutrition from anemones in the form of detritus from the column and tentacles, other animal tissue from the gastric cavity and either anemone mucus or tissue. It is still not clear whether hermit crabs remove tissue from whole anemones. Besides the other observations already presented, crabs move among tentacles such that their mouth parts would be in a position to bite off tips of tentacles, but it is very difficult to see just what the mouth parts are doing. The anemone may benefit from the crab's behavior by having excess mucus and detritus removed, but, of course, the removal of tissue and/or food would be deleterious. When the removed crabs ran to their anemones after being replaced in their tanks, they could have been seeking protection or going to the anemone after being "reminded" of its presence by contact with anemones in the test for protection. With large anemones the crab engages in sweeping behavior when encountering tentacles while walking along the column of the anemone. The tentacles move so that the crab is able to walk freely along the column area formerly under the drooped tentacles. By sweeping the tentacles of small anemones the whole anemone closes Broughton page 9 so the crab is free to walk on the anemone, pick up detritus, shove its cheliped into the gastric cavity and walk into the mouth opening of the anemone. It is not clear why the crab strokes tentacles; the chelipeds were seldom brought to the mouth during this behavior so the crab is probably not eating. This behavior would fit, however, as would the other crab behaviors, into the theory which will be discussed later of anemone mucus as the crab's protection. IV. INITIATION OF CRABS TO ANEMONES A. Observations One isolated Pagurus samuelis was placed in each of seven tanks, each of which had one or more already-settled anemones. Five tanks had one A, xanthogrammica, one tank had one solitary A. elegantissima and one tank had twelve clonal A. elegantissima 1 to 2 cm. in diameter. One crab was never seen encountering its A. xanthogrammica even after two and one-half weeks in the tank, and so will not be included in the initiation behavior data. The initiation behavior leading to protection was observed to consist of the following sequential steps: 1. Exploring. The crabs spent between 17 seconds and 10 minutes exploring the tank, eating detritus and trying to climb the walls, before encountering an anemone. Broughton page 10 2. Retreating. During the exploring behavior the five crabs with A. xanthogrammica each came within a few cms, of their anemone and then quickly walked backwards away from the anemone. The crab which eventually had the greatest number of interactions with its anemone did this three times, The two crabs which had an intermediate number of interactions each retreated from their anemone twice, and the two crabs which had the fewest interactions did this once each. 3. Contacting the column. With the four A. xanthogrammica the first encounter consisted of touching the column with antennae and appendages. The crab was not stung and the anemone did not react. 4. Contacting the tentacles. Three of the four crabs associated with A, xanthogrammica were seen to contact A, xanthogrammica tentacles with their appendages and antennae. In this first contact with tentacles, the antenna or appen- dage stuck to the tentacle, released and then was placed back on the tentacle. This was repeated two to ten times. For crabs with A. ele- gantissima, their first encounter with anemones was to touch their antennae to the tentacles, which was repeated up to twenty times. After this encounter the appendage or antenna stuck fewer times for shorter, less severe contacts—the crab did not jerk back as much as in previous encounters. Broughton page 11 5. Sweeping across tentacles. The crab with the clonal A. elegantissima immediately began to sweep its antennae across tentacles after its antennae were protected. Three other crabs, the one with the solitary A. elegan- tissima and two with A, xanthogrammica, were observed sweeping tentacles while walking on the columns of their anemones several days after their initial encounter. 6. Nestling Nestling was observed with clonal A. elegantissima, solitary A, elegantissima and two A. xanthogrammica the day after the initial encounters. Three weeks after their initial encounters the crabs were held against the tentacles of their own anemones to test for protection. Of the five total crabs with A. xanthogrammica, the crab with the most tentacle contact during the three weeks was protected against nematocysts, the crab with less tentacle contact was protected after one encounter with the tentacles, and the three crabs with one or no encounters during the three weeks were not protected. The crab with the solitary A. elegantissima was stung upon contact with the tentacles four weeks after the initial encounter, but the crab had not been observed near the anemone for the previous eight days. The crab with the clonal A. elegantissima was protected. Broughton page 12 B. Discussion The behavior of an isolated crab with clonal A. elegantissima was similar to the behavior, previously described, of crabs on clonal A. elegantissima in the field. The only difference in behavior was the initial encounter with the anemone in which the antennae of the isolated crab stuck to the tentacles. The fact that at the end of one month the crab with the solitary A. elegantissima was not protected even though it had been at one point indicates that the protection against nematocysts can be lost. This will be discussed further in the next two sections. From the initiation set-ups with A. xanthogrammica the following pattern of sequential steps leading to full protection against nematocyst discharge can be elucidated: the crab retreats from the anemone, later pokes at the column, walks on the column. touches the tentacles and sticks several times, touches the tentacles and sticks fewer times, and finally is able to touch, sweep and stroke the tentacles without being stung. The retreating behavior indicates that perhaps in the initial stages the crab perceives the anemone as something to be avoided, but then changes this perception. However, the fact that one crab was never seen to interact with its anemone, and another crab was seen interacting only three times in four weeks suggests that perhaps not all crabs seek to or are able to initiate relationships with anemones. Broughton page 13 V. DEGREES OF PROTECTION A. Observations A crab found on the column of an anemone may or may not be protected against nematocyst discharge by the same individual anemone; of 15 Pagurus samuelis found on A. xanthogrammica columns, 8 were protected against nematocyst discharge. Of 14 isolated crabs tested, none were found to be protected against the tentacles of either A. xanthogrammica or A, elegantissima. A crab found on the column of A. xanthogrammica was not protected against the tentacles of solitary A. elegantissima in all nine cases tested and was not protected against clonal A, elegantissima in 8 out of 11 cases tested. However, all seven test crabs found on the column or tentacles of solitary A. elegantissima were protected against A, xanthogrammica. In seven cases tested, five crabs found on the tentacles of clonal A. elegantissima were protected against solitary A, elegantissima and A. xanthogrammica. Hermit crabs have not been observed on anemones other than A. elegantissima and A, xanthogrammica, but various anemone species were tested in the lab for response to crabs. Tealia lofotensis and Tealia crassicornis tentacles stuck very strongly to both isolated crabs and crabs associated with A. xanthogrammica. Corynactis californica tentacles stuck to the appendages of an isolated crab but did not stick to crabs associated with A. xanthogrammica. Metridium senile tentacles did not stick to either isolated crabs, associated crabs or my finger. Broughton page 14 Hermit crabs found associated with anemones were isolated from their anemones for various lengths of time and then their protection against nematocysts was tested. The protection was found to last from 45 minutes to three days. B. Discussion Apparently there are different degrees of protection against nematocyst discharge, as evidenced by the varied times which crabs remain protected and the fact that not all crabs found associated with anemones are protected. In addition, there are species differences between the protection acquired from an A, elegantissima and an A. xanthogrammica. Possible explanations for these phenomena will be discussed in the next section. VI. POSSIBLE MECHANISMS OF PROTECTION AGAINST NEMATOCYST DISCHARGE A. Observations Two isolated crabs which were stung by A. xanthogrammica were allowed to walk on the column for two minutes. The crabs were then stung in only 4 and 6 out of 8 encounters with tentacles of the anemone and the stings were apparently less severe since the crab was able to touch the tentacle for a few seconds before jerking back and the crab did not jerk back as far as before. After rubbing my finger along the column of an anemone 15 to 20 times, the finger no longer stuck to the tentacles of the anemone Broughton page 15 but a control finger did stick. A finger protected against A. xantho- grammica was also protected against other A. xanthogrammica, but not against A. elegantissima. A finger protected against A. elegantissima was also protected against other A. elegantissima and slightly protected against A. xanthogrammica (the stings felt less severe than the stings on the control finger). Paint brush bristles stuck to the tentacles of a lab A. xantho- grammica, and still stuck after being brushed in the detritus on the floor of the tank around the anemone. However, after being brushed along the column ten times the bristles stuck in only five out of ten and then in one out of ten encounters with the tentacles. B. Discussion These observations and those of initial interactions indicate that protection is acquired by contact with the anemone. There is no direct evidence that anemone mucus is the factor which provides protection to the crab, but mucus is a likely substance to be trans¬ ferred from the sea anemone to a crab, finger, or paint brush. Perhaps these items are protected in only a fraction of the same group of encounters because of incomplete coating with mucus. The idea of a mucus protection also fits the observations of different degrees of protection: crabs may have acquired different amounts of mucus and therefore be protected to different degrees. Also, the protection could be lost by a loss of mucus, and the more mucus a crab had and/or the less active a crab was, the longer the protection would last. The Broughton page 16 species differences in protection against different species could be explained by species-specific mucus: A. elegantissima is still stimulated despite the presence of A. xanthogrammica mucus but A. xanthogrammica is stimulated less in the presence of A. elegan- tissima mucus. The initiation of hermit crabs to anemones is similar to the initiation of anemone-fish to anemones as described by Mariscal (1966): the fish is stung upon contact with the anemone but continues to touch tentacles and eventually is not stung. Schlichter (1972) has shown that the anemone-fish becomes protected from nematocyst discharge by acquiring anemone mucus. He proposes that this mucus contains substances which de-sensitize the anemone's sensory inputs so that the anemone does not sting itself or nearby objects with which the tentacles have continuous contact. The acquisition of mucus may also play a significant role in the acclimation of Pagurus samuelis to Anthopleura elegantissima and A. xanthogrammica. Since an inanimate object is able to become acclimated to an anemone, it is clear that a theory explaining the crab's protection need not include a "crab factor": we need not further examine the hypothesis that the crab raises the threshold for nematocyst discharge by anesthetic or recognition by the anemone. The development of immunity to nematocysts after ingestion of anemone tissue would require a longer acclimation period than has been observed; also, three crabs fed anemone tissue for one week were still stung by tentacles, Broughton page 17 If the protection is strictly a matter of anemone mucus we may ask why more animals are not associated with anemones in order to reap the benefits of protection from predators and a ready food source. In order to obtain anemone mucus without being captured by the anemone, an animal must be able to jerk itself quickly and force- fully from the tentacles when it is initially stung. Hermit crabs and fish are capable of this. Because hermit crabs are scavengers, they are able to exploit detritus on the anemone and partially digested food in the anemone as food sources. Pagurus samuelis lives in crevices and pools of the 0.0 to 1.5 foot tide level, a habitat which the habitats of Anthopleura elegantissima and A. xanthogrammica include. P. samuelis is therefore more likely to encounter these anemones than P. granosimanus and P. hirsutiusculus which occur lower in the intertidal, or shore crabs which occur higher. Occasionally these two other pagurid species were observed interacting with A, elegantissima and A. xantho- grammica and appeared to behave like P. samuelis. The main factor preventing them from more abundant occurrence with anemones is probably habitat exclusion. VII. CRABS AS AN ANEMONE FOOD SOURCE A. Observations P. samuelis were placed on the tentacles of sixteen solitary A. elegantissima to test the response of the anemones to hermit crabs as food. The A. elegantissima ate two isolated crabs, two crabs found Broughton page 18 on A, xanthogrammica, two crabs found on clonal A. elegantissima and a crab which was climbing on the column and sweeping the tentacles of a solitary A. elegantissima several days previous to the experiment. An isolated crab was in the gastric cavity of an A. elegantissima for 20 minutes then escaped or was released, covered with mucus and missing one antenna. The solitary A. elegantissima did not eat two crabs found on clonal A. elegantissima, three crabs found near solitary A, elegantissima and three crabs found on the tentacles of solitary A, elegantissima. Seven A. xanthogrammica were tested for their response to hermit crabs as food. One crab found on the column of A. xanthogrammica but isolated 45 minutes was not dropped on the tentacles but was captured as it began to interact with a lab A. xanthogrammica, and was eaten. A crab from the isolated tidepool was taken into the gastric cavity of an A. xanthogrammica 15 minutes then released. A crab found on the tentacles of a solitary A. elegantissima and a crab from clonal A. elegantissima were both eaten by A. xanthogrammica. Two crabs found on the column and one crab from the oral disc of A. xantho¬ grammica were not eaten. The remains of a Pagurus granosimanus were found in the gastric cavity of a Tealia lofotensis in the lab. The crab and anemone had been in the same tank for a week prior to the crab being eaten. B. Discussion Generally, solitary A. elegantissima will eat isolated crabs and crabs associated with A, xanthogrammica but not crabs associated with Broughton page 19 solitary A. elegantissima. Crabs found on clonal A. elegantissima may or may not be eaten. A. xanthogrammica generally will not eat crabs found associated with other A. xanthogrammica. This agrees with the other observations of species-specific protection and varying degrees of protection. The one occasion in the lab in which a crab was captured when the crab approached the anemone to interact with it indicates that anemones are able to catch hermit crabs as food. VIII. SUMMARY The hermit crab Pagurus samuelis is able to develop a relation- ship with the sea anemones Anthopleura elegantissima and A. xanthogrammica such that the crab is able to interact with the column, tentacles and oral disc of the anemone without being stung by the nematocysts. The crab steals the anemone's food from its gastric cavity, picks up detritus from the column and tentacles, and eats either anemone tissue or mucus, The protection against nematocysts is probably due to the acquisition of anemone mucus on the crab's appendages so that the anemone responds to the crab as it would respond to its own tissue. ACKNOWLEDGEMENTS I wish to express my thanks and appreciation to Chuck Baxter, D. P. Abbott, I. A. Abbott and F. A. Fuhrman for their aid and encouragement in the pursuit of elucidation of the anemone-hermit crab relationship. e Broughton page 20 LITERATURE CITED Dales, R. P. 1966. Symbiosis in marine organisms, p. 299-326. In Henry, Mark S. (ed.), Symbiosis. Vol. I. Academic Press, New York. 478 p. Davenport, D. 1966. Anlysis of behavior in symbioses, p. 381-429. In Henry, Mark S. (ed.), Symbiosis. Vol. I. Academic Press, New York. 478 p. Mariscal, R. N. 1966. A field and experimental study of the symbiotic association of fishes and sea anemones. Doctoral Dissertation. University of California, Berkeley. Mariscal, R. N. 1972. Behavior of symbiotic fishes and sea anemones, p. 327-360. In Winn, H. E. and B. L. Olla (eds.), Behavior of Marine Animals. Vol. 2: Vertebrates. Plenum Press, New York, 503 p. Schlichter, D. 1972. Chemische tarnung: die stoffliche grundlage der anpassung von Anemonenfischen an Riffanemonen. Mar. Biol. (Berl.). 12(2): 137-150.