WEBB: EFFECT OF BOTTOM ON FAST START OF A FLATFISH 



tact would also have been augmented by recoil 

 forces generated by the tail and head tending to 

 displace the fish downwards. 



The motion of the center of mass in the vertical 

 plane also differed between speckled sanddabs ac- 

 celerating from the grid and in the water column 

 (Figure 2C, D). For fish accelerating from the grid 

 there was little vertical movement during fast- 

 start stage 1 because of the presence of that grid. 

 During stage 2, however, the center of mass was 

 accelerated vertically upwards moving 6.0 ±2.4 

 cm in 124 ms. Fish accelerating in the water col- 

 umn showed little vertical motion, except that due 

 to recoil (Figure 2C). 



The resultant motion of the center of mass of the 

 fish showed differences between acceleration in 

 the water column and from the grid, but these 

 differences were less marked than motions in the 

 horizontal and vertical planes (Figure 2E, F). The 

 increase in distance covered with time was ini- 

 tially greater for fish accelerating in the water 

 column. They traveled a total of 3. 7 ±0.6 cm in 63 

 ms compared with 2.9 ±0.4 cm for fish accelerating 

 from the grid. However, once fish pushed against 

 the grid in stage 2, performance improved. By the 

 end of stage 2, they had traveled a cumulative 

 distance of 10.2 ±1.2 cm in 124 ms, greater than 

 that of 8. 7 ±1.3 cm achieved by fish accelerating in 

 the water column. 



Velocities calculated for the center of mass 

 reached maximum values close to the end of stage 

 2. Maximum values were significantly greater for 

 fish accelerating from the grid compared with fish 

 accelerating in the water column (Table 2). Mean 

 acceleration rates would, of course, follow similar 

 trends to velocities. Maximum acceleration rates 

 only showed significantly improved performance 

 for resultant motion of fish accelerating from the 

 grid (Table 2). 



Table 2. — Results of ma.ximum acceleration rates and maxi- 

 mum velocities for Citharichthya stigmaeus accelerating from 

 the gi-id and in the water column. Data ( mean ±2 SE: n = 10) are 

 shown for motion of the center of mass. 



DISCUSSION 



These experiments show that the bottom (grid) 

 does influence fast starts in speckled sanddabs. 

 Fast starts from the bottom were associated with 

 large amplitude motions that would normally 

 cause substantial recoil of the center of mass be- 

 cause the bottom prevents that recoil. Motions of 

 fish accelerating from bottom contact were pre- 

 dominantly vertical compared with horizontal 

 motions offish accelerating in the water column. 

 The relative magnitude of vertical and horizontal 

 displacements of sanddab accelerating from the 

 grid are comparable with those observed in 

 crayfish under similar circumstances ( Webb 1979). 

 For both speckled sanddab and crayfish the 

 marked vertical motion is caused by prolonged 

 contact between the body and the ground. Unfor- 

 tunately, these large vertical displacements in 

 speckled sanddabs preclude any significant hy- 

 drodynamic interaction with the ground after con- 

 tact is lost. 



Withers and Timko (1977) give a succinct expla- 

 nation of this hydrodynamic ground effect, where 

 the ground influences the flow, increasing lift 

 (thrust) and decreasing drag. Hydrodynamic 

 ground effect rapidly declines as the ratio, y/s, 

 increases, where y is the gap between the surface 

 generating thrust and the ground, and s is the 

 span (width) of that surface. Blake (1979) found 

 that 80-90% of the ground effect vanished by the 

 time y/s reached unity in hovering mandarin fish. 

 The mean maximum span of speckled sanddab, 

 located at the center of mass, was 6.7 ±0.2 cm. For 

 this span, y/s = 1 after about 130 ms, 18 ms after 

 the end of physical contact with the bottom (Fig- 

 ure 2D). The caudal fin had a maximum span of 

 2.4 ±0.3 cm, and a value of v/s = lis reached within 

 about 16 ms after the tail loses contact with the 

 bottom (Figure lA). Thus, the vertical accelera- 

 tion rapidly lifts the speckled sanddab out of the 

 hydrodynamic influence of the ground so that 

 there is insufficient time for any interaction to 

 affect performance. However, the absence of this 

 effect during a fast start does not preclude hydro- 

 dynamic ground interactions from improving 

 steady swimming close to the sea floor. 



Thus, the observations on fast starts of speckled 

 sanddab from the bottom show that a single 

 kinematic pattern is used, which sustains bottom 

 contact for most of a fast start, but which prevents 

 development of any significant hydrodynamic 

 ground effect. However, a direct push against a 



275 



