200 



_ l5 ° 

 cc 

 uj 

 m 



2 



5 ioo 



0- 



Z 



50- 



csss 



I I R BOREALIS 

 ESSSP.GONIURUS 

 ^^ P.HYPSINOTUS 



■i OTHER SHRIMP 



ES 



ESS 



0.6 22.9 45.7 68.6 



POT DEPTH (M.) 



91.4 



Figure 4. — Species composition of individual pot catches 

 from a surface-to-bottom series of shrimp pots fished 

 1 day in 300 feet (91.4 m.) of water in Kachemak Bay, 

 Alaska. 



Crangonidae and Pandalopsis dispar were 

 readily taken in pots. Although P. dispar 

 and species of Crangonidae were common 

 in trawl catches, no _P. dispar and only a few 

 Crangonidae were taken in pots during the 

 3 years we sampled in Kachemak Bay. Ap- 

 parently they did not enter baited pots. 



The fishing of surface -to-bottom pots was 

 applied in two ways in our shrimp investiga- 

 tions in Kachemak Bay. First, 1-day sets 

 were made at each of three sampling sites 

 twice each month. This sampling provided 

 information on seasonal changes in species 

 composition and size of shrimp by depth 

 and site. Figure 4 illustrates the catch from 

 a typical 1-day set. Second, two strings of 

 pots occasionally were set about 400 feet 

 (121.9 m.) apart in 300 feet (91.4 m.)of water, 

 and each string was pulled and reset (after 

 removing the catch) every 3 hours through- 

 out a 24-hour period. This sampling supplied 

 information on the diel changes in vertical 

 distribution (vertical migration) of shrimp. 

 Catches at various levels showed the pro- 

 gression of the vertical migration. Compari- 

 son of the catches in the two strings for 

 each 3-hour period indicated the variation 

 between sampling areas in the abundance of 

 shrimp at particular levels. Figure 5 illus- 

 trates the type of data obtained for one string of 

 pots during one set of this type (catches of only 

 one species, Pandalus borealis , are shown). 



1200 



1300 



ISOO 



1800 



BAR WIDTH INDICATES 

 NUMBER CAUGHT 



■ SO 



■^ 100 



■■■■■■200 



2IO0 2400 



2400 0300 



TIME PERIODS 



0400 

 0900 



0(00 

 1200 



Figure 5. — Catches of Pandalus borealis for 3-hour fishing 

 periods throughout a 24-hour period in pots of a surface- 

 to-bottom set, Kachemak Bay, Alaska, April 16-17, 

 1966. The set consisted of five pots, one suspended at 

 each of the following approximate depths: 2 feet (0.6 m.) 

 (surface), 75 feet (22.9 m.), 150 feet (45.7 m.), 225 feet 

 (68.6 m.), and 300 feet (91.4 m.) (bottom). 



We did not use sets of surface-to-bottom 

 pots specifically for studying shrimp in un- 

 trawlable areas, but comparison of catches 

 of shrimp in pots and trawls made the values 

 of this use apparent. The need for a method 

 of studying the shrimp populations in un- 

 trawlable areas is indicated by the observa- 

 tion of Hjort and Ruud (1938) that most of 

 our knowledge of shrimp stocks is from 

 trawling, whereas in the same paper they 

 mentioned an apparent increase in shrimp 

 abundance near "the rocky sides of the prawn 

 grounds ." 



Representative data from catches of shrimp 

 taken by bottom trawling, surface-to-bottom 

 pot fishing, and bottom pot fishing in Kachemak 

 Bay are presented in figure 6. The relative 

 abundance of the species taken in surface - 

 to-bottom pot fishing is nearer that of the 

 trawl catches than is the relative abundance 

 in the catches in bottom pots only. Although 

 the three methods generally take the same 

 species, the trawl and surface-to-bottom pots 

 catch primarily P_. borealis , and secondarily 

 P. goniurus , whereas the bottom pots catch 

 mostly P. hypsinotus and species of the 

 family Hippolytidae. 



If we assume that the trawl catches are 

 our best indication of the species com- 

 position of the shrimp population at the 

 point fished, the surface-to-bottom pot fish- 

 ing gives a better indication of the species 

 composition than does bottom pot fishing. 

 This difference in the catches of the two 

 methods of pot fishing should be con- 

 sidered in studies of shrimp in untrawlable 

 areas. 



