FISHERY BULLETIN: VOL. 79, NO. 2 



Procedures 



Individual speckled sanddabs were starved for 

 24 h. Each fish was lightly anesthetized (MS-222)2 

 and a white thread was sewn above the center of 

 mass of the stretched-straight body calculated 

 from previous measurements on a subsample of 

 fish. This reference facilitated later film analysis. 

 Each fish was then placed in an observation arena 

 and allowed to recover for 24 h overnight. 



The observation arena consisted of a chamber 45 

 cm long, 30 cm wide, and 40 cm deep. A fast start is 

 a short duration stereotyped activity during 

 which fish travel short distances but accelerate at 

 high rates to reach large speeds (Eaton et al. 1977; 

 Webb 1978). The size of the arena would not inter- 

 fere with fast-start activity but would prevent sus- 

 tained high-speed swimming. This study is con- 

 cerned only with the former activity. 



The observation chamber had a false bottom 

 made from a wire grid (2.5 cm squares made from 

 0.08 cm diameter wire). A solid bottom was located 

 beneath the grid and could be set 0, 1, 3, or 6 cm 

 beneath. Comparison of performance during fast 

 starts with the grid set at various distances above 

 the solid bottom was expected to show the impor- 

 tance of the hydrodynamic ground effect as differ- 

 ences in performances once physical contact with 

 the grid had ended. Vertical reinforcers prevented 

 the grid from acting as a spring when a force was 

 applied by an accelerating fish. The observation 

 chamber also allowed room for fish to swim in the 

 water column for short distances permitting ob- 

 servation of fast starts in the water column remote 

 from all possible ground interaction. 



The bottom was located in one of the four posi- 

 tions beneath the grid while the sanddab became 

 accustomed to the chamber overnight. Next morn- 

 ing a 1.0 V/cm d.c. electric shock was used to ini- 

 tiate a fast start either from the grid or when the 

 sanddab was temporarily stationary in the water 

 column clear of the grid and bottom. All experi- 

 ments were performed at 14° C. 



Fast starts were recorded on movie film at 250 

 frames/s. A 45° mirror above the observation 

 chamber allowed simultaneous observation of top 

 and bottom views. A floating lid prevented surface 

 ripples distorting the top-view image. 



Each fish was killed after an experiment. The 

 body outline was traced on paper. Mass, total 



^Reference to trade names does not imply endorsement bv the 

 National Marine Fisheries Service, NOAA. 



272 



length, wetted surface area, location of the center 

 of mass, and muscle mass were measured as de- 

 scribed by Webb (1977). 



There were no differences in these morphomet- 

 ric characteristics among various groups of speck- 

 led sanddabs. Combined values were (mean ±2 

 SE); mass, 28.96±3.78 g; length, 13.7 ±0.6 cm; the 

 center of mass, located 5.2 ±0.2 cm from the nose; 

 wetted surface area, 120 ±10 cm^; muscle mass, 

 free of skin and scales, 11.78 ±2.28 g. 



Movie film was analyzed frame by frame to ob- 

 serve kinematics and to measure performance. 

 Details are given by Webb (1978). Briefly, perfor- 

 mance was measured for the motion of the center 

 of mass of the stretched-straight body. This point 

 approximates the instantaneous center of mass of 

 the flexing body which is the point about which 

 propulsive forces act. Coordinates for motion of the 

 center of mass were measured in the horizontal 

 and vertical planes. The resultant motion of the 

 center of mass was calculated. Velocities and ac- 

 celeration rates for these three motions (hori- 

 zontal, vertical, and resultant) were calculated 

 using moving point linear regression methods. 

 Performance parameters were compared using the 

 ^-test, and significant differences between means 

 are declared at the 5% level. 



RESULTS 



Kinematics 



There were no differences in kinematics of fish 

 accelerating from the grid set at various distances 

 above the solid bottom. Fast-start kinematics for 

 acceleration from the grids (Figure 1) showed the 

 normal three stages originally described by Weihs 

 (1973). During stage 1, the body was bent into a 

 U-posture (0-80 ms in Figure lA) comparable to 

 the C-posture of other teleosts when viewed from 

 above (Eaton et al. 1977; Webb 1978). The center of 

 mass showed some recoil (i.e., lateral movement in 

 the opposite direction to the tail, and normal to the 

 fish axis) towards the grid because the fish adopted 

 a posture with the body raised by the median fins 

 before a fast start (Stickney et al. 1973). The body 

 was always bent away from the bottom. 



During stage 2 (80-120 ms in Figure lA), the 

 body was bent in the direction opposite to that of 

 stage 1 as the body curvature traveled caudally 

 along the body. Some point of the body remained in 

 contact with the grid almost to the end of stage 2. 



