CHEMORECEPTION AND BEHAVIOR 253 should be due mainly to the different slopes. The spike frequency discrimination could be very useful for orienting and tracking in a (and rise time) during this portion of each slope was 3.1 Hz (1800 fluid environment, because so much physical information about the ms) for the Low slope, 5.5 Hz (980 ms) for the Medium slope, and plume and odor source is present in the distribution of slopes in the 70 Hz (160 ms) for the High slope. Once the concentration stopped eddy field (4). In this study, single cells could discriminate be- rising, cells adapted to the constant background within a couple of tween a range of odor onset slopes with rise times similar to those seconds. We note that the lack of a response to the second slope measured in laboratory studies of jet plumes (1, 4). In general, rise in the Medium slope (middle panel between 7 and 8 seconds) animals may have different chemoreceptor cells “tuned” to partic- of Figure 1 may be due to the large fluctuations in the output of the ular ranges of pulse characteristics, such as frequency, height, piston pumps preceding this rise. These periodic fluctuations length, and slope (1, 2, 3). (some of which exceed 5 /J,IVQwould cause cumulative adaptation Thanks to D. Mellon and G. Gomez for helpful discussions. in the chemoreceptor cell such that the overall response is de- Supported by NSF grant IBN-9723542 to JA and a B.U.M.P. pressed (5). Because spike adaptation can occur rapidly, beginning Humes alumni Award to EZ. within 500 ms of the stimulus presentation, a steeper onset slope may minimize adaptation and provide the highest frequency for a given concentration. We will not discuss here the intracellular Literature Cited signal transduction pathways that may be involved in the excita- 1. Atema, J. 1996. Bid. Bull. 191: 129-138. tion and adaptation phenomena observed. 2. Borroni, P., and J. Atema. 1988. J. Comp. Physiol. A. 164: 67-74. These preliminary results demonstrate the feasibility of deliver- 3. Gomez, G., and J. Atema. 19%. J. Exp. Biol. 199: 1771-1779. ing measured concentration slopes and provide initial evidence 4. Moore, P., and J. Atema. 1991. Biol. Bull. 181: 408-418. that chemoreceptor cells can act as “slope detectors.” Odor slope 5. Voigt, R., and J. Atema. 1990. J. Comp. Physiol. A. 166: 865-874. Reference: Biol. Bull. 197: 253-254. (October 1999) Individual Recognition and Memory in Homurus americanus Male-Female Interactions Cristin Berkey’ and Jelle Aterna (Boston University Marine Program, Woods Hole, Massachusetts 02543) Individual recognition and memory have been studied in Ho- show a definite loser and winner after the first 20-min period was rnarus americanus, the American lobster. It has previously been disqualified and the fight was not included in the data analysis. demonstrated, through lobster boxing matches, that males can The fights were scored using a pre-established scale of agonistic recognize individuals. When a male encounters the same male it levels (1) with some additions made for differences that arose as a lost to previously, it avoids a new fight. The same animal will fight result of the fights being between males and females. An agonistic on subsequent occasions if presented with an unfamiliar opponent level was assigned to each animal every 5 s. The levels ranged (1). The ability to recognize a previous opponent lasts for up to a from -2 to 5, with -2 demonstrating fleeing behavior and 5 week, and it is based on olfactory recollection of the opponents’ demonstrating claw snapping or claw ripping at the opponent. An urine (2). The same results have been obtained in female-female overall agonistic scale was calculated for each animal in each fight encounters (3). We therefore hypothesize that male-female pairs of by summing all of the agonistic values. A fight was determined to Homarus americanus can also recognize individuals, as demon- be over when one animal ceased to show an agonistic level above strated by shorter periods of fighting on the second encounter with one. In this way, the duration of each fight was scored. To prevent familiar individuals. bias during analysis of the behavioral tapes, the scorer was blind to Males and females were paired by carapace length, with no pair whether the tapes were from a first or second fight. differing by more than 3 mm. It was also ensured that in all In group A, 10 pairs of fights were carried out. Significantly, in matched pairs, the animals had not previously met. Animals were 9 out of 10 instances, no fighting occurred on the second encoun- kept in isolation for 24 h before their first boxing match and for the ter; fighting did occur in the tenth case. In group B, 13 pairs of 24 h between the first and second matches. All boxing matches fights were carried out. Two sets were disqualified because there took place in a 240-I glass aquarium and were videotaped. One was no clear winner after the first fight. In 3 out of 11 instances, no male and one female were placed in the tank and separated by a fighting occurred on the second encounter. In 8 instances, fighting removable, opaque divider. The lobsters were allowed to acclimate occurred. The fraction of times in which fighting occurred on the for 10 min before the divider was removed, and then allowed to second encounter is significantly different between the two groups interact for 20 min. There were two sets of fights, group A and (P < 0.005 .x? = 8.416 df = 1). group B. In group A, the same pair of animals fought twice. In When a fight was between familiar opponents, the aggression group B, the animals were rotated so that each animal met a new level of the loser was significantly lower in the second boxing opponent in the second fight. In both sets, the first and second match than in the first boxing match (P < 0.05 paired t test). fights were separated by 24 h. Any pair of animals that did not However, for unfamiliar opponents, the aggression level of the loser was not significantly lower in the second boxing match than ’ Tufts University, Medford, Massachusetts. in the first (paired t test). The aggression level of the winner was 254 REPORTS PROM THE MBL GENERAL SCIENTIFIC MEETINGS not significantly different between the two fights in either group A determines the end of the fight: in male-female pairs, females or group B (paired t test). typically lost. We observed two behaviors that had not previously been seen in The fact that second encounters were much shorter, to the point either male-male (1) or female-female (3) matches. One of these that no fighting occurred, supports the hypothesis that male and was mounting behavior, which typically occurs just prior to mating female lobsters can recognize each other as individuals. The results (4): the male climbs on the back of the female and attempts to turn also demonstrated differences, such as nonreactive contact and her onto her back using his first two pairs of walking legs and his mounting behavior, in the way male and female lobsters interact third maxillipeds. In our encounters, however, this mounting be- within the boxing match situation, compared to male-male and havior never progressed to actual mating. We observed male- female-female interactions. It is possible that these differences female mounting behavior in five separate encounters, two of parallel the natural interactions between male and female lobsters. which involved the same pair of animals. A second form of Females are allowed to enter male shelters, whereas other males behavior not seen in same-sex boxing matches was nonreactive are not (5). Therefore, males and females must be capable of being contact. This included any situation in which the two animals were near to each other without demonstrating aggressive behavior. in physical contact but did not seem to be reacting to each other. Financial support from NSF-REU (OCE-9605099 site award The average time spent in contact across the 42 fights analyzed to Boston University). We thank Leslie McLaughlin, Jennifer was 179 s per 20-min fight. This value had no obvious correlation Walters, and Dr. Frank Grass0 for their assistance with this project. with previous exposure to the opponent. In addition to the differ- ences already mentioned, fight duration in first encounters was significantly shorter in male-female fights than in male-male fights Literature Cited (P < 0.05, t test). The average fight duration was 250 s (SE ? 50 s) 1. Karavauich, C., and J. Atema. 1998. Anim. Behav. 56: 1553-1560. in male-female pairs, but 457 (SE + 73 s) in male-male pairs (1). 2. Karavanich, C., and J. Atema. 1998. Behaviour 135: 719-730. However, there was no significant difference in fight duration 3. Atema, J., T. Breithaupt, A. LeVay, J. Morrison, M. Mallidis, and between male-female and female-female first fights: the average M. Edattukaran. 1999. Chem. Senses (in press). fight duration in female-female pairs was 206 s (SE ? 73 s) (3). 4. Atema, J., and D. G. Engstrom. 1971. Nature 232: 261-263. The similarity in fight duration between female-female and male- 5. Bushmann, P., and J. Atema. 1997. Can. J. Fish. Aquat. Sci. 54: female pairs is consistent with the observation that the loser 647-656. Reference: Bid. Bull. 197: 254-255. (October 1999) Urinary Protein Concentration in Connection with Ago&tic Interactions in Homuncs americanus Leslie C. McLaughlin’, Jennifer Walters2, Jelle Atema, and Norman Wainwright (Boston University Marine Program, Marine Biological Laboratory, Woods Hole, Massachusetts 02543) Homarus americanus, the American lobster, relies heavily on Twenty-three wild-caught female lobsters (79-83 mm carapace chemical signals as a source of information (1). Urine in particular length) were used. They were kept in groups of 5 to 15 in 400-l has notable behavioral effects on these animals, suggesting that it communal tanks with flow-through seawater and artificial shelters is an important source of chemical information, specifically in made from PVC pipe. All lobsters were healthy and fed squid social communication. Lobsters store urine and, in different be- twice a week. Lobsters were isolated in individual aquaria 48 h havioral contexts, release it from two paired, frontally facing before their first agonistic encounters with unfamiliar lobsters, nephropores into their gill current (2, 3). Previous studies (4, 5, 6) following Karavanich and Atema (4). Isolated lobsters were fed a have shown that lobsters recognize conspecifics as individuals for small piece of squid at least 1 h before each interaction. All at least 24 h and up to a week after one 20-min staged agonistic agonistic interactions took place in a 240-l glass aquarium that was interaction. Both urine release and antennular chemoreception are rinsed and filled with unfiltered seawater immediately before each necessary for this recognition (7). Evidently lobster urine carries a lobster interaction. Two female lobsters unfamiliar to one another chemical signal that allows for individual recognition among an- were placed in the aquarium on either side of a removable, opaque imals; the challenge is to identify the chemical composition of the divider and were allowed 10 min to acclimate to their new sur- signal. One candidate is the proteinaceous content of the nephro- roundings. To initiate the encounter, the plastic divider was raised pore gland (8). We set out to study protein concentration in the and the lobsters were allowed to interact for 20 mitt, after which urine of lobsters subjected to two agonistic encounters separated they were returned to their respective isolation tanks. Twenty-four by 24 h. hours later this procedure was repeated: about half (n = 13) of the animals were exposed to the same lobster they had encountered the ’ Brown University, Providence, Rhode Island. day before; the other half (n = 10) were paired with an unfamiliar * Oregon State University, Corvallis, Oregon. lobster. All encounters were videotaped for later analysis.