A novel data logger incorporating a 3-axis gyroscope, a 3-axis accelerometer, and a 3-axis magnetometer was developed and externally attached to Japanese amberjacks, Seriola quinqueradiata, to investigate the possibility of using this device for monitoring the movement performance of the faststart behaviour of fish in the field. Triggered escape behaviours were measured simultaneously by the data logger (500 Hz) and a high-speed camera (200 Hz) in a tank. By using a gyroscope, the data logger accurately reconstructed the gravity-based acceleration, 3-dimensional attitude, dynamic acceleration, and angular velocity of fish during the fast-start movement, which was impossible by previous methods using only an accelerometer and a magnetometer; these variables can therefore be used to assess the distance

The dead-reckoning technique is a useful method for obtaining 3-D movement data of aquatic animals. However, such positional data include an accumulative error. Understanding the source of the error is important for proper data interpretation. In order to determine whether ocean currents affect dive paths calculated by dead-reckoning, as has previously been hypothesized, we examined the directions of the estimated positions relative to the known real points (error direction) and the relationship between the error direction and the current direction. 3-D dive paths of emperor penguins Aptenodytes forsteri diving at isolated dive holes in eastern McMurdo Sound were reconstructed by dead-reckoning, and the net error and error direction were calculated. The net error correlated positively with the dive duration. The error directions were not distributed uniformly, and the mean error direction tended to be north of the starting point of dives. Because there was a southwardflowing current in eastern McMurdo Sound, the ocean current was likely to affect the calculated dive paths. Therefore, the method of error correction generally used, in which the net error divided by the dive duration is applied to each estimated position, is realistically appropriate, provided that the current does not change significantly during a dive.

Animal tracking provides integral spatiotemporal information that contributes to the growing field of movement ecology.AT is one of the main approaches to track the movements of aquatic animals.

In order to establish techniques for the stock enhancement of red tilefish Branchiostegus japonicus, it is important to understand the behavioral characteristic of both hatchery-reared and wild fish. Four hatchery-reared and six wild fish were released and tracked using acoustic biotelemetry in Maizuru Bay, Kyoto, Japan, were also referred to for analysis. The released fish moved around in relatively large areas within about 10 days after the release, and then three hatcheryreared and two wild fish settled and stayed within limited areas. Fish of both origins showed strong site fidelity and diel activity patterns, i.e. they were detectable in the day but not at night, probably due to hiding behavior in their burrows at night. However, some discrepancies between fish of both origins were detected by comparing their track terms and activity patterns in detail. These discrepancies are probably attributable to the differences in environmental conditions that the fish had experienced before release.KEY WORDS: acoustic biotelemetry, Branchiostegus japonicus, diel activity pattern, hatcheryreared fish, red tilefish, site fidelity, stock enhancement.

Wild red tilefish Branchiostegus japonicus were tracked and/or monitored for up to nearly 3 months in Western Wakasa Bay and Maizuru Bay, Japan, using ultrasonic telemetry to examine their movements and burrow fidelity patterns. In Western Wakasa Bay, it is guessed that fish released in relatively shallow waters, about 30 m deep, actively moved into more suitable deeper habitats after their release. In Maizuru Bay, the fish did not undertake large-scale movements and showed burrow fidelity about 1 month after their release. That is, the fish primarily remained in the burrow, which was constructed in the sea-bottom, during the night and moved outside the burrow during the day. The marked circadian rhythm from Fourier analysis indicated a clear diel movement pattern of the red tilefish.KEY WORDS: burrow fidelity, diel movement, red tilefish, ultrasonic telemetry.