INTRODUCT Aggressive behavior characteristic of territoriality has been reported for several polychaetes. The polychaete Nereis pelagica exhibits a fighting response which is linked to tube defense (Clark 1959). The polynoid Hesperonoe defends the burrow it shares with Urechis against polychaete invaders. Dimock (1974) found that as a result of aggressive interactions, Arctonoe pulchra distributes itself so that there is only one individual per host. To date, intraspecific aggression in Arctonoe vittata has not been reported. In California waters Arctonoe vittata occurs most commonly in the pallial grooves of the keyhole limpet Diadora aspera, the starfish Dermasterias imbricata, and the giant chiton Gryptochition stelleri. The scent of the host is detected at a distance by receptor organs on the prostomial antennae. Hosts are recognized on contact by sense organs in the palps (Evans 1971). In this paper data are presented which show that adult A. vittata have differing levels of aggression and that groups of these worms, through aggressive interactions, constitute hierarchies. It is seen that removal of the antennae drastic¬ ally curtails the worms' fighting behavior. RIALS AND METHODS MA Arctonoe vittata were obtained from only the giant chiton ptochiton stelleri. The chitons were collected subtidally Aggression in Arctonoe vittata James Weisberger off of the Monterey peninsula, California. Each worm was measured to the nearest millimeter. A series of tests were performed to determine whether Arctonoe has different levels of aggression. The pallial groove of the host proved to be the best environment to con¬ sistently elicit the fighting response. In each test the worms' behavior was observed and the number of bites exchanged in a head-on encounter between two worms was counted. The worms were provoked by gently nudging their posterior ends with blunt forceps. As Arctonoe is sluggish, provocation was necessary so that the worms would quickly move near each other, Test periods for each encounter were five minutes. In the first series of investigations groups of four or five Arctonee were tested together. No individual was used in two different groups. In the remaining studies selected worms were tested against seven others. The same C ptochiton was used throughout the tests. It had a total length of 25.5 cm., with its pallial groove meas- uring 17 cm. long and 0.7 cm. wide. The ( ptochiton was lain on its dorsal side so that its pallial grooves could be more easily seen. Gently flowing seawater bathed the host during tests, submerging the worms. Aggression is quantified in this paper by the number of bites, accomplished by rapid eversion of its armed proboscis, that an Arctonoe delivers to conspecifics in a given time period. James Weisberger Aggression in Arctonoe vittata LSUET Worms were observed in a fingerbowl, between glass plates separated by 8 mm., and in a 7 mm. diameter plastic tube for twenty minute periods in both light and darkness. Only one bite was observed. These same worms would readily attack each other on a host. Thus the host possesses some unidentified factor needed to release the aggressive response, Bites tore off elytra and produced open wounds and torn appendages. Bites almost always were directed at the prostomial region. There was considerable interworm variability as to the duration of bites, some lasting up to 5 seconds. The most bites recorded in a five minute period by one worm was 4 bites, never in rapid succession. When two worms first contacted each other they frequently jerked away. When nearing again, the palps were extended; fighting was most likely to occur if the worms' prostomiums were in contact. Worms "reared up" slightly before they attacked. Conversely, worms that assumed a "head down" posture in which they lowered their prostomial region relative to their opponents almost never were bitten; the other worms just slid by. Only once in 15 trials was a worm attacked when in this position, xperiment 1: To see if the worms exhibited differing levels of aggression, five worms were tested together. Each worm in the group was tested against the others 4 times in the Aggression in Arctonoe vittata James Weisberger space of 8 days. The number of times a worm attacked and the number of times the worm was attacked by others was recorded. Prior to testing, the worms had no injuries. After the tests, the worms were checked for injuries. The results are in Table 1. There is an inverse relationship between the number of times a worm attacked conspecifics and the number of times a worm was bit by them: the more a worm attacked, the less he was attacked, and vice versa (Figure 1). The reason worm E (the circled dot in Figure 1) does not follow the linear relation¬ ship may be because it was very small and would not attack the larger worms. In a different series another 30 mm. long worm exhibited the same pattern of response. Three other groups of four worms apiece were tested for evidence of different levels of aggression. The worms in each group were tested against each other twice in two days. The results are shown graphically in Figure 2. Linear relation¬ ships are evident. Experiment 2: To test the effectiveness of the observed "hierarchies", all five worms from Group 1 were put in the pallial groove of one Cryptochiton. Four other Cryptochitons without worms were placed in the tank. They were left over¬ night. The most aggressive, worm C, remained in the pallial groove. Worm E, the smallest one, was found in the groove of the mantle of the host, apparently not large enough to evoke full aggression. James Weisberger Aggression in Arctonoe vittata All others were found in new hosts. When this was repeated, the most aggressive, worm C, was the only worm on the host. All 4 others had found new hosts. Arctonoe sometimes switches to new hosts spontaneously. Throughout the six weeks of investigation, 7 out of 27 worms changed hosts. Experiment 3: Most of the injuries sustained during fights were the removal of prostomial sensory appendages. To see which sensory appendages contribute to Arctonoe's level of aggression the following tests were done. In the first test 4 new worms were tested against each other twice. The two most aggressive ones were anesthetized in an isotonic solution of MgClo and seawater and both palps were excised. After a day's recovery, all four worms were tested against each other twice. There was no change in the worms' behavior when the palps were removed, (Figure 3). No control was done in the following tests because the shock of anesthesia and the operation did not seem to affect the two worms whose palps were removed (i, e. reduce their aggression). Next, four new worms were tested against each other once, Under anesthesia, both palps, both sets of tentacular cirri, and the three prostomial antennae were removed from the most aggressive. After a day recovery the worms were tested once. The most aggressive worm did not bite, (Figure 4). Under anesthesia, the antennae were removed from the most aggressive of four new worms. After a day recovery the worms Aggression in Arctonoe vittata James Weisberger were tested twice. The antennae removal affected the worm's aggression level, changing it from an average of 4.5 attacks per test to an average of 0.5 attacks, (Figure 5). The an- tennae seem instrumental in maintaining aggressive behavior. These results were confirmed by testing the most aggressive of 8 new worms against the other seven. After antennae removal and recovery, it again was tested. The results in Table 2 indeed show that antennae removal results in significantly reduced aggression. Using a 2X2 contingency table, a "P" value of less than .005 was achieved. ISCUSSION Arctonoe vittata is an obligate commensal. Apparently, being isolated from a host produces abnormal behavior, a manifestation of this being the conspicuous lack of intra¬ specific aggression. This is consistent with Dimock (1969). who found that Arctonoe would bite a sea star while on its host but at no time did an isolated worm bite a sea star. Intraspecific aggression seems to be a way to effect uniform distribution of large Arctonoe on a host population. This is shown in the experiment in which five worms were placed on the same host. In once case only the most aggressive worm was left. In another run, the most aggressive and the smallest were left. It could be hypothesized that Arctonoe, being an obligate commensal, defends the pallial groove of its host Aggression in Arctonoe vittata James Weisberger from others, as the host provides both protection and food, Perhaps the amount of food available to Arctonoe on the host is not very abundant. If this is the case, then a high aggress¬ ive level would be of selective advantage to the worm in com¬ petition for this scarce resource. Dimock (1974) states that intraspecific aggression within Arctonoe results in the density which occurs in the field, namely, one large worm per host, The word "large" is important here. As shown in Table 1 and Figure 1, the smallest worm does not follow the linear relationship. This is because it would not attack the larger worms of 60+ mm. length. Out of a total population of worms in the study, worm E was the smallest in both length and width, Dimock (1974) also states that very small worms of Arctono pulchra were never observed to bite the largest worms. Un- fortunately, he gives no measurements. There is no consistent slope to the attack vs. attacked lines of Figure 2. However, the number of worms available was limited (27 total) and more samples would presumably have in¬ dicated if there is one standard slope of the lines. These lines point to a "hierarchy" between worms in a group, much like a pecking order, with one worm on top and others below him. The fact that the most aggressive worm stayed on the host while the others left in experiment 2 points to the aggressor's dominance. There is nothing ritualistic about Arctonoe's aggressive behavior; injuries are serious and sometimes lead to death. Aggression in Arctonoe vittata James Weisberger Table 1 shows the injuries sustained by the first group after 4 trials. In all cases one of the palps are missing, yet clearly there are aggressive worms here. The most aggressive. averaging 5 attacks per test, was missing both palps. This data substantiates the results of the experiment in which the palps were excised and no change in aggression was observed. It is curious that the palps play no critical role in aggression as they are conspicuously employed during worm interactions. The least aggressive worm, averaging only 1 attack per test. was missing an antenna. This supports the results of the exper¬ iment in which the antennae were removed and a significant change in behavior was noted. Perhaps the reason that worms usually attack the prostomial region is to bite off the opponent's antennae, thereby reducing chances of retaliation. Evans (1971) states that the antennae are used in distance chemoreception to detect the host and that the palps are used in contact chemoreception to recognize the host. It could be hypothesized that an Arctonoe without antennae cannot detect his opponent easily and therefore is not quickly aroused to aggressive behavior. Exactly how the antennae mediate aggression is not yet known. Aggression in Arctonoe vittata James Weisberger SUMMARY 1. Adult Arctonoe vittata were tested in head-on aggressive encounters in the pallial groove of a Cryptochiton host. 2. Arctonoe was found to have intraspecific differences with respect to levels of aggression. 3. The most aggressive worm in a group successfully defends a host against others, evicting conspecifics, 4. Removal of the palps had no effect of the worms aggression, but antennae removal significantly reduces aggressive behavior. Aggression in Arcto ACKNOWLEDGEMENT I wish to thank Dr. Robin Burnett for his su help in this research. Aggression in Arctonoe vittata James Weisberger LITERATURE CITED 1. Clark, R. B. 1959. The tubicolous habit and the fighting reactions of the polychaete Nereis pelagica. Animal Behav. 7: 85-90. 2. Dimock, R. V. and J. C. Dimock 1969. A possible "defense" response in a commensal polychaete. Velige 12: 65-68. 3. Dimock, R. V. 1974. Intraspecific aggression and the distribution of a symbiotic polychaete on its hosts. pp. 29-44 In Vernberg, W. B. (ed.). ymbiosis in the Sea. Univ. S. Carolina Press, Columbia, S. C. 4. Evans, S. M. 1971. Behavior in polychaetes. Quart. Rev. Biol. 16: 379-405. 5. MacGinitie, G. E., and N. MacGinitie 1968. Natural listory of Marine Animals. McGraw-Hill, New York. 11 Aggression in Arctonoe James Weisberger ittata TABLE 1: A ratio of the number of attacks to the numbe a worm was attacked by others. Worm E is the smallest. effect of its size upon aggression is discussed in the Aggression in Arctonoe vittata TABLE 1 F NINRI E NOFIDI ER AVEl AVERAGE NUMBER OF TIMES OF TIMES WORM WORM ATTACKE WORM WAS AT 4.5 1.0 1.5 5.0 1.75 2.75 3.0 1.0 2.0 James Weisberger JURIE missing: front elytra 1 palp, 1 set of cirri, 1 an¬ tenna missing: front elytra 2 palps 1 set of cirri missing: front elytra 2 palps missing: front elytra 2 palps 1 set of cirri missing: front elytra 1 palp 13 LENC ITH WIDT 65 mm 5.5 mm 62 mm 6.5 mm 50 mm 5.0 mm 47 mm 5 mn Aggression in Arctonoe vittata James Weisberger TABLE 2: Antennae from the most aggressive worm were exci It was tested against 7 other worms before and after r Values show a significant decrease in aggression and an accompanying rise in aggression toward it by other worms Using a 2X2 contingency table, a "P" value of less than was obtained. Aggression in Arctonoe vittata TABLE 2 BEF James Weisber Aggression in Arctonoe vittata James Weisberger FIGURE 1: Average times a worm attacked vs. th attacked by others. The circled value is the smalli 6- —.— FIGURE 1 â- — â 1 —— ——. — — . —— — — — — —.—— ââ ..... — —— AVERAGE NUMBER OF TIMES WORM WAS ATTACKED Aggression in Arctonoe vittata James Weisberger FIGURE 2: Relationships between levels of aggression in different groups. Group 1 (without worm E) -2 Group 2— O, Group 3 -X, and Group 1 -O. No uniform slope is seen. 18 o FIGURE 2 — AVERAGE NUMBER OF TIMES WORM WAS AITACKED Aggression in Arctonoe vittata James Weisberger FIGURE 3: Before and after relationships in the palps excision experiment. ° — before palp removal; O - after removal. The circled values represent before and after values for the worms that underwent palp removal. 2 — FIGURE 3 AVERAGE NUMBER OF TIMES WORM WAS ATTACKED Aggression in Arctonoe vittata James Weisberger URE 4: Before and after values when the palps, cirri, and antennae have been removed. — before removal; O — after removal. The values for the worm that had the sensory appendages removed are circled. 22 FIGURE 4 0 AVERAGE NUMBER OF TIMES WORM WAS ATTACKED Jmes Weisberger Aggression in Arctonoe vittata 1 5: Before and after values when the antennae w — before removal; O- after removal. The for the worm whose antennae were removed are circled. 9 FIGURE 5 AVERAGE NUMBER OF TIMES WORM WAS ATTACKED