FISHERY BULLETIN: VOL 77. NO I 



700 

 600 

 500 

 400 

 300 



I 00 - 

 90 

 80 - 

 70 - 

 60 

 50 - 

 40 

 30 



20 



Figure 5. — Changes in the relative abundance of larval Am- 

 modytes spp. in the Northeast Fisheries Center spring ichthyo- 

 plankton surveys from 1974 to 1977 in the area extending from 

 Cape Hatteras northward. Data from Smith, W. G.. and L. Sulli- 

 van. 1978. Annual changes in the distribution and abun- 

 dance of sand lance. Ammodytes spp., on the northeastern 

 continental shelf of the U.S. from the Gulf of Maine to Cape 

 Hatteras. Northeast Fish. Cent.. Sandy Hook Lab., Sandy 

 Hook, NJ 07732. Lab. Ref No. SHL 78-22. 



of individuals. Individuals varied from 7.4 to 24.0 

 cm FL (measurements from bottom-trawl 

 catches), with a mean length of 18.2 cm. Sand 

 lance within a given school were of similar size; 

 slightly larger fish were observed in positions at 

 the head or central "core" of the school, with the 

 smaller individuals occurring at the periphery. 

 This distribution by size within the school was 

 observed in both study areas. Schools were ob- 

 served on the surface, at mid-depth, and near the 

 bottom. 



Inshore school strengths described by 

 Kiihlmann and Karst (1967) for European sand 

 lance species Hyperoplus lanveolatus and Ammo- 

 dytes lancea were commonly 30-100 or 200-300 

 individuals. These smaller schools joined up to 

 form schools of from 500 to > 1 ,000 fish and headed 

 offshore for deeper water in the early morning. We 

 observed schools of this size primarily in the Prov- 

 incetown slope area. However, because the indi- 

 vidual size of the fish and school strengths on 

 Stellwagen Bank were larger, it is unlikely that 

 these schools formed in the Provincetown slope 

 area and moved out to the bank. 



248 



.ScIkkiI Shape 



The shape of sand lance schools, where indi- 

 viduals were not engaged in feeding, was constant 

 in appearance. As a school moved undisturbed 

 through the water it appeared vertically compres- 

 sed, tightly compacted, and bluntly linear from 

 the lateral view (Figure 6). Provincetown slope 

 schools were 1-5 m wide, 0,5-1,5 m high, and 3-20 

 m long; these measurements depended on school 

 strength. This school form, where the height- 

 width-length ratio was approximately 1 :3: 10 ( hav- 

 ing more individuals situated ahead, alongside, 

 and behind than above or below), is called a strat- 

 ified school iWahlert and Wahlert 1963). This 

 school formation was, in general, independent of 

 .school strength. The "nearest-neighbor" distance 

 between fish was approximately '/■!-% body length 

 (BL) (Figure 6), This distance became greater 

 along the school's flanks. The "nearest-neighbor" 

 distance decreased to '4 BL when the school exhib- 

 ited a fright reaction to divers. The fishes leading 

 the school and ones along the flanks usually swam 

 the deepest. School shapes described by 

 Kiihlmann and Karst (1967) were similar to the 

 measurements reported in this study, but a sig- 

 nificant difference appeared in the school height 

 and length measurements. Kiihlmann and Karst 

 (1967) listed their school height as 15-50 cm, and 

 their school length as s40 m. Sand lance schools 

 encountered in our study were more than double 

 the height and shorter in length. The European 

 study took place in water depths of 1-6 m. and in 

 this relatively shallow water, there may be a ten- 

 dency for a school to flatten out and increase its 

 length, 



Moxement 



The swimming motion of sand lance is sinusoi- 

 dal in form and eellike in appearance from the 

 dorsal and ventral views. Sidewise undulations 

 begin at the head and run along the body toward 

 the tail (Figure 7). Schools swimming undis- 

 turbed, and not engaged in feeding, maintain an 

 estimated speed of 30-50 cnVs. Schools exhibiting 

 feeding behavior usually swim at about half the 

 speed of undisturbed schools, or about 15-25 cm,s, 

 and spread out to a little over double the normal 

 schooling distances so that the nearest neighbor is 

 approximately l-l'/2 BL away. Smaller schooling 

 groups were observed to swim faster than larger 

 schools. When approached by divers, schools ac- 



