Spawning Stock (millions) 



Figure 2.— Stock-recuit relation for sockeye salmon in the Karluk River basin. 

 Squares are the running geometric mean (by 9) of stock and recruit estimates 

 for the 1922-77 broods. The curved, solid line was described by R = 1.83 

 (lO*^) + 7.73 P + 1.29(10"^) p2 - 5.58(10-^2)P^ where, i? = recruits and 

 P = stock. Ages of fish in the escapement (1922-36 from Barnaby 1944; 

 1937-69 from the Northwest and Alaska Fisheries Center, National Marine 

 Fisheries Service, Auke Bay, AK; 1980-85 from Alaska Department of Fish 

 and Game, Kodiak, AK) were used to estimate the recruits produced by each 

 brood. The diagonal lines show how the replacement line changes as the ex- 

 ploitation rate increases from to 0.3. 



lapsed" into the lower of the two. Fishing can cause 

 such a collapse (Peterman 1977) and recovery 

 becomes impossible unless exploitation rates are 

 reduced to levels substantially lower than the rate 

 that caused the collapse. 



Multiple equilibria in an exploited population can 

 be caused by depensatory mortality— the loss of a 

 relatively greater fraction of the population when 

 it is small than when it is large (Neave 1953). Several 

 functional responses (Ricker 1954; Rolling 1973) 

 have been used to describe relations between prey 

 density and predation rate— one of which (Type III 

 relation) can produce multiple equilibria in the stock- 

 recruitment curve of a prey population (Peterman 

 1977). The Type III or S-shaped functional response 

 is characteristic of predators that consume a small 

 fraction of the prey at low prey population density; 

 as prey population density increases, however, the 

 predators rapidly increase the fraction consumed 

 through learning or aggregation. The concave. Type 

 II functional response holds when the fraction con- 

 sumed is high at low prey density. 



The apparent potential for stock collapse, as 

 depicted in the Karluk sockeye stock-recruitment 

 curve, could be the consequence of Type III preda- 

 tion mortality. A preliminary survey of the food 

 habits of fish in the system showed that coho salmon. 



0. kisutch, and two chars— the Dolly Varden, 

 Salvelinics malma, and Arctic char, S. alpinus— 

 were predators of juvenile sockeye salmon. We set 

 out to determine whether the functional responses 

 for coho salmon and for char were of Type II or Type 

 III. Our approach was to describe the relation be- 

 tween the number of prey eaten per predator and 

 the index of prey abundance provided by the annual 

 counts of adult sockeye salmon that entered the 

 system for spawning. Unfortunately, the study had 

 to be terminated after five years because of a man- 

 agement decision to enhance the productivity of 

 Karluk Lake with commercial fertilizer; we could 

 not eliminate the possibility that the effects of fer- 

 tilization would confound predation responses. We 

 describe the data that were accumulated during five 

 field seasons and our tentative conclusions concern- 

 ing the role of predation mortality in the dynamics 

 of these sockeye salmon. 



Methods 



Sampling sites were established at locations 

 around the littoral zone of Karluk Lake at the outlets 

 of spawning streams and at beach spawning areas 

 (Fig. 1) in 1982. The Karluk River was sampled from 

 the outlet at the lake to about 100 m downstream. 



613 



