2 
Fishery Bulletin 107(1 ) 
bluefish. Under a winter photoperiod, as water tem- 
perature decreased, swimming speed correspondingly 
decreased (Olla and Studholme, 1971), but a longer 
study of behavior over a seasonal cycle has not been 
undertaken. 
Bluefish use sustained or powered swimming for 
daily activity and long distance travel. They propel 
themselves by flexing the rear part of the body and 
tail and use their other fins mainly for stability and 
maneuvering. Powered swimming can be used for long 
periods without fatigue, although it has energetic costs 
(Beamish, 1978). For prey capture, bluefish use burst 
swimming at speeds recorded at up to 800-1000 cm/s 
(Olla et al., 1970). A novel and unexpected locomotory 
behavior was witnessed by Studholme and others dur- 
ing an unpublished study of a school of juvenile bluefish 
in the 12- kL research aquarium in 1984-85. At night, 
individual fish rolled onto their sides and, with their 
bodies and fins held still and slightly curved, glided 
downward diagonally, and then ascended. This behavior 
is remarkable in teleost fishes with laterally compressed 
bodies, because these fish are assumed to swim in a 
vertically oriented manner. 
In 1995, a study was designed to expand upon previ- 
ous bluefish research (Olla and Studholme, 1971, 1972; 
Olla et al., 1985) to examine in detail the seasonal 
cycle of activity in bluefish from fall through winter 
and spring. Unfortunately, the documentation of the 
cycle of behavior and gliding in age-0 bluefish from the 
1984-85 study was lost in a fire. Therefore, a school of 
age-0 bluefish was brought to an aquarium of similar 
dimensions to that used earlier. Food consumption, 
swimming speed (as a measure of activity), and mode of 
swimming were recorded to determine their relationship 
to changing temperature and photoperiod. It was hoped 
that the previously observed but undescribed gliding 
behavior would recur. Particular attention was paid to 
changes in behavior that could be related to migration, 
overwintering, and the bioenergetic response. 
Materials and methods 
Source of fish and laboratory conditions 
Spring-spawned age-0 bluefish were captured by hook 
and line from Sandy Hook Bay, NJ, over the period of 
a week. One hundred and five fish were placed in the 
research aquarium on 3 August 1995. The research 
aquarium, located at the National Oceanic and Atmo- 
spheric Administration James J. Howard Marine Sci- 
ences Laboratory, Sandy Hook, Highlands, NJ, held 
121 kL (32,000 gallons), was 10.6 m by 4.5 m, and 
had a water depth of 2 m (Olla et al., 1967). Water 
originated from Sandy Hook Bay and salinity in the 
aquarium varied with ambient conditions (19 to 23 
psu). The aquarium had a flow rate of 400 L /min and 
10% of the water was replaced each week. Heaters and 
chillers were used to maintain the desired temperature 
(±0.5°C). Photoperiod matched that at the latitude of 
North Carolina (35°N ) with a minimum of 9.3 hr of light 
at winter solstice. Light intensity at the bottom was 206 
lux (0.0027 mEinsteins PAR) at midday and 0.17 lux at 
night, which allowed for video recording. 
Water temperature was controlled to approximate 
that in migration and at summering and wintering ar- 
eas, as determined from historical data from the NOAA 
National Data Buoy Center. The temperature was set 
at 20°C from the start of the experiment through 9 
October. Then it was decreased incrementally over 8 wk 
(0.105°C/d) until it reached 15°C, the average minimum 
temperature off North Carolina south of Cape Hatteras. 
The temperature in the aquarium was raised from 15 
to 20°C (0.194°C/d) over the period from 24 May to 23 
June, simulating temperature increases in New Jersey 
coastal waters. 
Feeding and growth 
Bluefish were usually fed each afternoon, 5 d/wk. Live 
mummichogs (Fundulus heteroclitus) were weighed and 
tossed into the middle of the aquarium until the bluefish 
had fed to satiation. In winter, the bluefish did not feed 
as readily and therefore were left undisturbed to feed for 
an hour. Then the uneaten mummichogs were removed. 
A subset of fish was seined from the aquarium, anaes- 
thetized, measured by total length (TL, to ~1 mm), and 
weighed (to ~1 g) every two months. The total weight of 
food consumed each week was divided by the calculated 
weight of all fish in the aquarium that week, and then 
divided by seven, to obtain the consumption rate in g 
food per g body weight per day (BW/d). 
Daily and seasonal rhythms 
Activity was measured by swimming speed. For each 
observation, an actively swimming bluefish was timed 
while it crossed a measured portion of the far wall of 
the aquarium. Only fish close to the wall were timed. 
During the first 5 minutes of each hour of the day, five 
randomly selected fish were timed and their speeds were 
averaged. When live observations could not be made, 
data were taken from videotapes made with a time-lapse 
video recorder and camera placed at a window opposite 
the far wall of the aquarium and set to record for five 
minutes at the start of each hour. Live observations were 
also made for longer periods during day and night. Every 
three or four weeks, or more often, a complete set of 
swimming speeds was recorded, in which 24 five-minute 
segments were observed on each of seven days, for a 
total of seventeen weeks. Speeds were expressed in cm/s 
and body lengths per second (BLl/s). For comparisons of 
day and night activity, a subset of the swimming speed 
observations were used, from 0900 through 1500 and 
2100 through 0300. 
Modes of swimming 
The general behavior of the school, depth of swimming, 
and modes of swimming were noted in the first 5 min- 
