Effects of the insecticide sevin and its first hydrolytic product, 1-naphthol,on some early developmental stages of the bay mussel Mytilus edulis

Six developmental stages of the bay mussel Mutilus edulis, from fertilization to 32 h after fertilization, and also unfertilized eggs, were exposed for 1 h to different concentrations of the insecticide Sevin and its first hydrolytic product, 1-naphthol. After exposure, the larvae or eggs were separated from the pesticide solution and returned to clean water. At 48 h after fertilization, the numbers of normal and abnormal larvae were determined and 1-h EC₅₀ values (the effective concentrations that caused anomalous development of 50% of the test animals) were calculated. The most sensitive developmental stage was the one that occurred shortly after fertilization at the time of appearance of the first polar body. Thereafter, sensitivity decreased as age increased. The EC₅₀ of Sevin for the first polar body stage and 32-h stage were 5.3 and 24.0 mg/l of Sevin, respectively. The EC₅₀ of l-naphthol for the first polar body stage was 5.2 mg/l. Effects of the toxicants on development were characterized by disjunction of blastomeres, a reduction in the rate of development, and asynchronous and unaligned cleavages.     

Onsets of schooling behavior and social transmission in chub mackerel Scomber japonicus

Grouping behavior has various types of antipredator functions. Some of these functions require social transmission of information, such as the many-eyes effect, whereas others do not, such as the dilution and confusion effects. Functions of grouping behavior would enhance with social transmission among group members. We investigated and compared the onsets of schooling behavior and social transmission of information in chub mackerel Scomber japonicus. Onset of schooling behavior was observed in rearing tanks by calculating the degree of parallel swimming. Onset of social transmission was examined by using visual cues from conspecifics. A group of five individuals was put in each of three experimental chambers from which they could see a group of conspecifics in the neighboring chamber. A weak electric stimulus was given to one of these chambers, and information transfer among individuals was observed. We found that social transmission by visual cues started on 30 days posthatching (25.1 mm in standard length), which was 2 weeks after the onset of schooling behavior. The late onset of social transmission relative to schooling behavior might be attributed to different predation pressure with development, or by underdevelopment of optic tectum, as the volume of the optic tectum did not increase just after the onset of schooling behavior.     

Schooling behavior of Antarctic krill (Euphausia superba) in laboratory aquaria: Reactions to chemical and visual stimuli

Antarctic krill,Euphausia superba, often exhibit abnormal behavior in laboratory aquaria, usually hovering in a stationary position, unresponsive to most external stimuli. In the austral summer of 1985–1986 at Palmer Station on Anvers Island, Antarctica, we provided laboratory conditions which inducedE. superba to school in large aquaria. Captive krill swam horizontally and exhibited the full spectrum of behaviors normally displayed while schooling at sea. Schooling krill avoided visually contrasting stimuli, with avoidance distances correlated with stimulus size. Schools responded in qualitatively different ways to presentations of food, chemical compounds, and abrupt increases in light intensity. We describe the conditions necessary for aquarium schooling and discuss the importance of an appropriate social environment for displays of escape, avoidance, and feeding behaviors and of positional preference within the school.     

Temporal patterns in three-dimensional structure and activity of schools of the Atlantic silverside Menidia menidia

Temporal patterns in the structure and activity of schools of the Atlantic silverside Menidia menidia were investigated under laboratory conditions using a new computerized video technique for three-dimensional analysis. Fish were collected by seining at Middle Bridge, Pettaquamscutt River, South Kingstown, Rhode Island, USA. Approximately once a week from May to December, 1979, the behavior of a school was videotape-recorded in the laboratory at 30 min intervals from sunrise to sunset under a temperature and photoperiod regime simulating ambient field conditions. Four measurements of structure and activity were computed: (1) swimming speed (SPEED); (2) distance to nearest neighbor (NN); (3) mean direction of travel (DOT); (4) depth in the tank (DEPTH). Results demonstrated a strong seasonal rhythm in the behavior of schools, but were inconclusive in showing periodicities associated with daily, tidal or lunar cycles. NN and DEPTH were significantly different between spring, summer, fall and winter while SPEED changed significantly across 3 seasons (summer and winter were not significantly different). DOT was bimodally distributed, with peaks in spring and fall. Changes in these variables corresponded to shifts in the behavior of silversides from active, synchronized schools in spring and fall to inactive mills in summer and inactive non-schooling groups in winter. Changes in schooling behavior were only partly related to changes in temperature and photoperiod. Patterns of silverside schooling behavior correlated with their reproductive and migratory cycles and may represent an anti-predator strategy.     

Schooling tendency and foraging benefit in the ocean surgeonfish

Summary Field observations of individuallytagged ocean surgeonfish (Acanthurus bahianus) indicated that schooling provided a foraging benefit: an increase in relative forging time. The magnitude of this increase differed among individuals, primarily due to time budgets of the fish when they were not in schools. The proportion of nonschooling time devoted to foraging was positively correlated with fish size, while the proportion of schooling time that was spent forging was similar for all individuals. Schooling tendency (percentage of time in schools) was positively related to magnitude of foraging benefit and negatively related to fish size. When fish size was held constant by partial correlation, schooling tendency remained positively related to foraging benefit. These results provide evidence that fish which derive a greater foraging benefit from school membership spend more time in schools.     

Effects of temperature and body mass on metabolic rates of sprat, Sprattus sprattus L.

Sprat, Sprattus sprattus L., is a small schooling clupeid forming large stocks in several ecosystems. Despite its high trophodynamic impact, little is known about its energy consumption rates. As a central component of a bioenergetic budget, metabolic rates of sprat from 3.11 to 9.71 g wet weight (WW) were measured at nine different temperatures (T) ranging from 9 to 21°C using a computer-controlled intermittent-flow respirometer. Routine metabolism (R _R) was related to T (°C) and WW (g) by R _R = 0.074 WW^1.077 e^{0.080 T}. Standard metabolic rates (R _S) as calculated from the 10% percentiles of the repeated measurements were on average 12% lower and still influenced by continuous swimming activity: R _S = 0.069 WW^1.073 e^{0.078 T}. We interpret the deviation of the scaling exponent b from typically found exponents of b ~ 0.8 as a consequence of permanently elevated activity level. The high permanent swimming activities also indicated that the concept of standard metabolism may not be meaningful in schooling planktivorous fish. These results suggest that generally in bioenergetic models for clupeid schooling fish the activity multipliers should be chosen very conservatively.     

Predator avoidance and school size in a cyprinodontid fish,the banded killifish (Fundulus diaphanus Lesueur)

Summary Solitary and schooling banded killifish (Fundulus diaphanus Lesueur) were attacked with a fish predator model in the littoral zone of a lake and their flight reaction distance to the model recorded. Reaction distance of solitary fish did not differ significantly from that of schooling fish, which was statistically constant over a wide range of observed school sizes. The results fo not support the proposed early predator warning function of social groups, which predicts an asymptotically increasing flight reaction distance with increasing group size. However, fish in larger and safer groups may detect an approaching predator sooner but delay their flight response, thereby permitting their individual members to spend more time assessing the nature of the threat before escaping. Such a strategy would reduce costly false alarms. Relative variation in recorded reaction distances was correlated negatively with school size, suggesting that schooling reduces variability in latency time to escape predators. This may be evidence for the above behavioural strategy. Further, the speed of transmission of the flight response within the school greatly exceeded the speed of approach of the predator model. This phenomenon enables schooling individuals to initiate avoidance behaviour before the approaching predator can be detected and before it reaches them.     

Escape behaviour of solitary herring (Clupea harengus ) and comparisons with schooling individuals

The escape behaviour of solitary herring (Clupea harengus L.) startled by a sound stimulus was observed by means of high-speed video-filming. The results were compared with data from a previous study on the escape behaviour of schooling herring. Escape responses were divided into “away responses” and “towards responses” according to the orientation of the C-bend of the body relative to the stimulus. The proportion of away responses was smaller for solitary than for schooling herring. In solitary herring, the subsequent escape trajectories of fish making initial away responses showed a bimodal pattern of distribution, with modes at 130 and 180° from the stimulus. Trajectories following towards responses, however, were mainly within the semicircle directed at the stimulus, and their pattern of distribution differed from that of away responses. This result contrasts with observations on schooling herring, whose trajectories following both initial away and towards responses are directed away from the stimulus. In addition, we measured the response latency, defined as the interval of time between stimulus presentation and the first detectable movement of the fish. Solitary herring showed a higher proportion of short-latency responses (latency <50 ms) than schooling herring. Different behaviours appear to be exhibited by herring depending on whether they are solitary or within a school. We hypothesize that schooling may raise the threshold for initiation of fast escape responses, giving longer latencies and slower responses which are more appropriate in their directionality and reduce the possibility of collisions with neighbours. In addition, we suggest that schooling behaviour enhances the directionality and co-ordination of the escape response of the whole school, possibly increasing the probability of surviving a predator attack. Received: 2 October 1996 / Accepted: 17 October 1996     

Individual differences in the mechanisms underlying the onset and maintenance of paternal behavior and the inhibition of infanticide in the monogamous biparental California mouse,Peromyscus californicus

Virgin male Peromyscus californicus tend to behave infanticidally or nonparentally towards pups, whereas virtually all males exhibit parental behavior following birth of their own young. Most males (65–75%) living with their pregnant partner attacked or ignored unfamiliar pups and did not become parental (behave parentally) prior to birth of their young. However, a significant minority of males (34%) living with their partner became parental after just 24 h of postcopulatory cohabitation with the female and remained parental throughout their mate's pregnancy. Males that were infanticidal before the birth of their young became parental only after their young were born. The presence of the mother was necessary for the postpartum maintenance of paternal behavior and the inhibition of infanticide in males that were infanticidal prepartum. In contrast, males that were paternal prior to birth of their young continued to be paternal after birth, even in the absence of postpartum contact with the mother. Thus, different mechanisms are involved in the inhibition of infanticide and the onset and maintenance of paternal behavior. Correspondence to: D.J. Gubernick     

Differential attraction of females to dominant males: Olfactory discrimination and mating preference in the brown lemming (Lemmus trimucronatus)

Summary This study investigated differential attraction of estrous brown lemmings (Lemmus trimucronatus) to conspecific males recently exposed to each other for a 10-min agonistic encounter. In tests conducted 5 min, 1 h and 24 h after agonistic encounters, females preferred the oder of dominant males to that of defeated males when both odors were presented simultaneously in a Y-maze olfactometer. Defeat in an agonistic encounter did not reduce the propensity of male lemmings to initiate sexual behavior. In one-male, one-female tests conducted 5 min after agonistic encounters, dominant males achieved higher mount and thrust scores while defeated males obtained higher scores for attempted mounts. The sexual behavior of dominant and defeated males did not differ significantly in similar tests conducted 1 h and 24 h later. In contrast, females readily mated with dominant males and tended to avoid defeated males in two-male tethering tests conducted 5 min after agonistic encounters. In these tests, females still showed a preference for dominant males 1 h and 24 h after male agonistic encounters.     

Patterns of the electric organ discharge during courtship and spawning in the mormyrid fish,Pollimyrus isidori

Summary Pollimyrus isidori's electric organ discharge (EOD) is of the pulse type. Patterns of EOD intervals were investigated prior to, during and following spawning behaviors as related with overt behaviors, and with the sound production by the nestbuilding male. Prior to the time of reproduction, isolated and socially interacting fish (n=15) showed characteristic discharge interval patterns for resting, swimming, probing, hovering and hiding activities. Males (n=8) and females (n=6) did not differ in their mean EOD repetition rates during resting (11.6±2.5 Hz), nor Short Bursts/min (less than 20 intervals of 8–13 ms). In interacting fish Long Bursts (greater than 20 intervals of 8–13 ms, lasting for more than 300 ms) were observed only during the attack and bite sequence. A pursuing fish displayed a rapid alternation of Long Bursts with Discharge Breaks (300–1000 ms silence) during the chase behavior. Avoidance behavior which followed from several attacks was correlated with a Medium Uniform Rate (8–12 Hz) normally lasting for 20 to 60 s, or a Discharge Arrest (silence greater than 1 s) in the submissive fish. The nocturnal courtship behavior began soon after dark (1900 h). Spawning typically started 2 to 5 h after dark, continuing for 2 to 6 h until about 0200 h. During courtship and spawning the female's brief visits (15–25 s) to the male's territory recurred every 30–60 s. At all other times the female was aggressively excluded from the nest region. Courtship and spawning behaviors are described along with the electrical displays identified from 19 spawnings in three fish pairs (from a total of 37 spawnings in 4 males and 4 females). Just prior to the onset of courtship behavior, with male territorial aggression beginning to decline, females switched from a Medium Sporadic Rate pattern (resting and hiding patterns; 13 Hz) to a Medium Uniform Rate pattern (6–8 Hz) while still in their hiding area. Females continued to display this uniform rate throughout the courship and spawning period, including the courtship and spawning bouts when Discharge Breaks or Arrests also occurred. This persistance distinguishes the courtship pattern from the similar avoidance pattern (see above). The male courtship and spawning EOD pattern was similar to the female's and unique for a territorial male. He switched from a High Sporadic Rate (swimming EOD pattern; about 18 Hz) to a regularized Medium Uniform Rate (about 9 to 11 Hz) only during courtship and spawning bouts, including 1–3 EOD Breaks during Vent-to-Vent coupling (average interval: 272±71 ms, n=37). No sooner had the female left the spawning site than he resumed displaying a High Sporadic Rate. This temporal correlation of reproductive behaviors with electrical displays suggests their instrumental role in mutual acceptance of mates. Males showed their sex-specific type of EOD phase-locking, the Preferred Latency Response, only during the first few hours of entry of a fish in their tank. Two females with EOD waveform features more typical of males also spawned repeatedly; waveform does not appear to be critical. Males stopped their nocturnal sound production for the later part of courtship and the whole spawning period. Except for infrequent attacks on the female between spawning bouts, the male did not resume singing until the end of spawning when all eggs were shed (around 0200 h); from this time on the male sang until dawn. The sequencing of the three acoustic elements (moans, grunts, growls) are described. A catalogue of discharge patterns correlated with overt behaviors (Tables 1, 2), and an integrated summary time table of P. isidori's complex reproductive behavior are presented.     

Life history of Gonatus onyx (Cephalopoda: Teuthoidea): ontogenetic changes in habitat, behavior and physiology

In the present study 670 individuals of Gonatus onyx (Cephalopoda: Teuthoidea) were observed in Monterey Bay, California from a remotely operated vehicle. The vertical distribution of this species was bimodal, with peaks at 400 and 800 m depth during the day and 300 and 500 m during the night. The bimodal distribution reflects a life stage shift between younger, schooling juveniles living in shallower water and older, solitary adults which live deeper. Ontogenetic changes in behavior associated with this life stage shift are reflected in the physiology of the organisms as well. Both aerobic and anaerobic metabolism, as estimated using mitochondrial and glycolytic enzymes, decline with increasing body mass, suggesting reduced locomotory capacity in deeper-living adults. Oxygen consumption rates were also determined in relation to oxygen partial pressure. Oxygen consumption regulation was similar between juvenile and adult squids. The critical oxygen partial pressures (29 to 30 mmHg) correspond precisely to the oxygen concentrations found at the depth of maximal abundance for day and night populations of juveniles and adults, respectively. Behavioral and physiological changes with ontogeny of G. onyx are believed to result from reduced visual predator/prey interactions in the light-limited deep sea. Received: 7 February 1999 / Accepted: 26 January 2000     

Stereophotography for the field biologist: measurement of lengths and three-dimensional positions of free-swimming sharks

A stereophotographic technique for determining size and relative position of free-swimming sharks is described and illustrated for schooling scalloped hammerhead sharks (Sphyrna lewini). The method yields total length and nearest-neighbor, interindividual distance; and each of these dimensions is expressed as a function of the shark's distance into a school on the vertical and horizontal planes. Stereopairs of photographs were taken by an aligned, beam-mounted pair of cameras (Nikonos III). The scale to determine the length of a shark from the paired photographic images was obtained from the horizontal displacement between the images. Displacement was correlated with optical axis separation from photographs of a scaled staff at known distances from the camera. Image dimensions on the photographs were measured by projecting a scale onto the stage through a camera lucida. The precision of repeated measurements of a 50 cm section of a scaled staff at increasing distances from the cameras of 2, 4, and 8 m was ±5.0%. Lengths of the sharks ranged from 109 to 371 cm, with a median of 178 cm, for 3 offshore sites in the Gulf of California during July and August 1979. At one site, El Bajo Gorda, lengths increased with both distance from the camera and distance into the group; in contrast, the interindividual distances (head-to-head) did not vary with distance into the group and possessed a median of 232 cm.     

Effects of nitrate on the diurnal vertical migration,carbon to nitrogen ratio,and the photosynthetic capacity of the dinoflagellate Gymnodinium splendens

A non-thecate dinoflagellate, Gymnodinium splendens, was studied in a 12 d laboratory experiment in 2.0x0.25 m containers in which light, temperature, and nutrients could be manipulated. Under a 12 h light: 12 h dark cycle, the dinoflagellates exhibited diurnal vertical migrations, swimming downward before the dark period began and upward before the end of the dark period. This vertical migration probably involved geotaxis and a diel rhythm, as well as light-mediated behavior. The vertical distribution of nitrate affected the behavior and physiology of the dinoflagellate. When nitrate was present throughout the container, the organisms resembled those in exponential batch culture both in C:N ratios and photosynthetic capacity (P_max); moreover, they migrated to the surface during the day. In contrast, when nitrate was depleted, C:N ratios increased, P_max decreased, and the organisms formed a subsurface layer at a depth corresponding to the light level at which photosynthesis saturated. When nitrate was present only at the bottom of the tank, C:N ratios of the population decreased until similar to those of nutrient-saturated cells and P_max increased; however, the dinoflagellates behaved the same as nutrient-depleted cells, forming a subsurface layer during the light period. Field measurements revealed a migratory subsurface chlorophyll maximum layer dominated by G. splendens. It was just above the nitracline during the day, and in the nitracline during the night, which concurs with our laboratory observations.     

Active swimming in meiobenthic copepods of seagrass beds: geographic comparisons of abundances and reproductive characteristics

The entry of meiobenthic copepods from sediments or seagrass blades into the water column and reproductive characteristics of actively migrating fauna were investigated from 1981–1986 in a temperate intertidal Zostera capricorni seagrass bed in Pautahanui Inlet, New Zealand and in a subtidal Thalassia testudinum bed in Tampa Bay, Florida, USA. Emergence of copepods in New Zealand varied over a tidalcycle, while in Florida a distinct diel periodicity was displayed. Selected copepod species in New Zealand had similar numbers emerging from sediments and/or blades over a 6 h period as the common copepods actively migrating from sediments in Florida. Daily abundances of emerging copepods (24 h) in Tampa Bay, Florida, were substantially greater than those in New Zealand, where migration is linked to tidal cover. In Z. capricorni meadows in New Zealand, sex ratios of copepods in sediments and on blades were dominated by females; males dominated water-column samples. In T. testudinum meadows in Tampa Bay, sex ratios of males to females, although of a lowermagnitude than in Z. capricorni beds, were higher in trap than in sediment samples. Differences in sex ratios, the availability in emergence traps of females of appropriate stage for mating, and observations on clasping in live samples from traps suggest that swimming behavior in copepods may be partly linked to prenuptial courtship. Meiobenthic copepods may use the water column as an important habitat for reproductive behavior.     

Do activity costs determine foraging tactics for an arctic seabird?

How energy costs affect foraging decisions is poorly understood for marine animals. To provide data relevant to this topic, we examined the relationship between activity levels and foraging behavior by attaching activity recorders to 29 chick-rearing wing-propelled diving birds (thick-billed murres, Uria lomvia) in 1999–2000. We connected the activity during the final dive bout with the prey item we observed being fed to the chicks. After accounting for changes in activity level with depth, activity was highest during the final dive of a dive bout, reflecting maneuvring during prey capture. Pelagic prey items, especially invertebrates (amphipods), were associated with higher depth-corrected activity, leading to shorter dives for a given depth (presumably due to higher oxygen consumption rates) and, thus, shorter search times (lower bottom time for a given depth). Pelagic prey items were likely captured during active pursuit, with the birds actively seeking and pursuing schooling mid-water prey. In contrast, benthic prey involved low activity and extended search times, suggesting that the birds slowly glided along the bottom in search for prey hidden in the sediments or rocks. We concluded that activity levels are important in determining the foraging tactics of marine predators. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Caribbean mangroves and seagrass beds as daytime feeding habitats for juvenile French grunts, Haemulon flavolineatum

Caribbean seagrass beds are important feeding habitats for so-called nocturnally active zoobenthivorous fish, but the extent to which these fishes use mangroves and seagrass beds as feeding habitats during daytime remains unclear. We hypothesised three feeding strategies: (1) fishes feed opportunistically in mangroves or seagrass beds throughout the day and feed predominantly in seagrass beds during night-time; (2) fishes start feeding in mangroves or seagrass beds during daytime just prior to nocturnal feeding in seagrass beds; (3) after nocturnal feeding in seagrass beds, fishes complete feeding in mangroves or seagrass beds during the morning. We studied the effect of habitat type, fish size, social mode and time of day on resting and feeding behaviour of large juvenile (5–10 cm) and sub-adult (10–15 cm) Haemulon flavolineatum in mangroves and seagrass beds during daytime. Sub-adults occurred in mangroves only, spent most time on resting, and showed rare opportunistic feeding events (concordant with strategy 1), regardless of their social mode (solitary or schooling). In contrast, large juveniles were present in both habitat types and solitary fishes mainly foraged, while schooling fishes mainly rested. Exceptions were small juveniles (±5 cm) in seagrass beds which foraged intensively while schooling. Large juveniles showed more feeding activity in seagrass beds than in mangroves. In both habitat types, they showed benthic feeding, whereas pelagic feeding was observed almost exclusively in the seagrass beds. In both habitat types, their feeding activity was highest during 8:00–10:30 hours (concordant with strategy 3), and for seagrass fishes, it was also high during 17:30–18:30 hours (concordant with strategy 2). The study shows that both mangroves and seagrass beds provide daytime feeding habitats for some life-stages of H. flavolineatum, which is generally considered a nocturnal feeder.     

Kin preference behavior in Bufo boreas tadpoles

Summary The ontogeny of sibling recognition behavior was studied in the laboratory in tadpoles of the western toad (Bufo boreas boreas) to test the hypothesis that Bufo tadpoles associate with siblings and to compare this behavior with two species of anurans previously studied. Tadpoles reared exclusively with sibs demonstrated a preference to associate with sibs over non-sibs both early and late in development but tadpoles reared with sibs and non-sibs (mixed rearing groups) exhibited no preference. Larvae that developed a preference for sibs after being reared with them for 75 days lost this preference following exposure to a mixed group for 2 to 6 days. Additionally, larvae reared in a mixed group did not develop a preference for the familiar mixed group nor was a preference exhibited by individual larvae that were reared exclusively with 5 non-sibs. These results suggest that larvae learn by prior association to discriminate sibs from non-sibs and that preferences are rapidly modifiable following exposure to non-sibs. However, the latter two experiments suggest that social preferences are not totally labile and are not based entirely on familiarity. Field observations of larval activity, aggregation behavior, and dispersal patterns in B. boreas indicate that tadpoles mix with non-sibs from early larval stages throughout development. Because even short term exposure to nonsibs resulted in a loss of sib preference in our experiments, we suggest that a kin selection interpretation of B. boreas aggregation behavior is not parsimonious. Results differ with those of other studies of sib recognition in larval anurans.     

Periodic behavior and residence time of yellowfin and bigeye tuna associated with fish aggregating devices around Okinawa Islands,as identified with automated listening stations

In order to better understand the associative behavior of yellowfin tuna (Thunnus albacares) and bigeye tuna (T. obesus) with anchored fish aggregating devices (FADs), we conducted long-term monitoring of these tuna tagged with ultrasonic transmitters, using automated receivers deployed on seven FADs around the Okinawa Islands. Current and surface water temperature were also monitored by data loggers attached to the stations as a way to examine the influence of these factors on the associative behavior of tuna with FADs. We monitored a total of 52 yellowfin and 11 bigeye tuna at monitoring FADs for a period of 2.5 years. We found that the majority of tuna remained continuously at the monitoring stations for a certain period (max.=55 days) without day-scale (>24 h) absences, until they left the stations completely. The residence time at a single FAD was estimated to be about 7 days, as the half-life for both yellowfin and bigeye tuna. No inter-specific differences were seen, though there was a significant difference in residence time between two size classes: the residence time of the larger size class was shorter than that of the smaller size class. We also found there was a periodicity of approximately 24 h and regularity of associative behavior, estimated based on the fluctuation pattern of the detection rate and of short-term (<24 h) absences. In particular, absences of several hours occurring once a day with high temporal precision were considered to be excursions within several nautical miles from the FADs. These results indicate that tuna express periodic behavior in relation to the FADs and can locate the FADs precisely enough to return to them after a certain time. No relationship was seen between associative behavior and abiotic oceanographic conditions. Therefore, the biological environment (prey availability, the presence of predators, etc.) and the internal state of the individual (hunger, etc.) may be more important than abiotic environmental cues for inducing changes in associative behavior and/or the departure from FADs. In addition, the strong association of tuna with a single FAD and the relatively prolonged residence time observed in the present study may relate to the vigorous activity of FAD fisheries in Okinawa and their utilization of a large amount of bait.Communicated by T. Ikeda, Hakodate