The fishery for Karluk Lake (Kodiak Island, Alaska) sockeye salmon Oncorhynchus nerka has undergone a period of high production (1920s to mid‐1940s), decline (mid‐1940s to late 1970s), and recovery (late 1970s to present). We evaluated historical spawner–recruit data and recent (1980–1994) limnological data to develop an escapement goal (EG) to provide maximum sustainable yield for sockeye salmon. Ricker analysis of the spawner–recruit data explained only a small proportion of the total variation (r2 = 0.17) in loge recruit/spawner (R/S) and exhibited a significant first‐order autocorrelation (r1 = 0.68; P < 0.0001). Ricker time–polynomial regression exhibited no significant residual autocorrelation, explained 71% of the variation in loge R/S, and indicated that intrinsic productivity (α parameter) has remained fairly constant, whereas system capacity (β parameter) has been temporally unstable. Markov transition probabilities indicated that escapements between 0.8 and 1.0 million fish maximize yield. In comparison, a graphical assessment of actual spawner abundance versus predicted maximum sustained yield indicated a production threshold (about 0.8 million spawners) and revealed a temporal pattern in production related to spawner abundance. These findings suggest a positive‐feedback mechanism related to the size of the escapement, which we hypothesize is primarily due to nutrients from spawner carcasses. Sockeye salmon carcasses produce an estimated 90% of the phosphorus above baseline loading and have accounted for 40% or more of the total phosphorus in Karluk Lake. The historical trend in sedimentary δ15N shows a strong reduction in marine‐derived nutrients since the onset of commercial fishing around 1890, a pattern that roughly corresponds to the decline in sockeye salmon production and system capacity. These findings and other trophic‐level limnological analyses suggest an EG of 0.8–1.0 million fish. Our evaluations also indicated that approximately equal proportions of early‐run (before July 15) and late‐run (after July 15) spawners are needed to optimize sockeye salmon production at Karluk Lake.

Sockeye salmon (Oncorhynchus nerka) have developed into a major run at Frazer Lake, Kodiak Island, Alaska, reaching over 600 000 adults 34 yr after the first egg-plant in 1951. In the last nine years (1978–86), escapements have averaged over 250 000 with a record high of 486 000 adults in 1985. The increase in escapements were inversely related to seasonal macrozooplankton densities (p < 0.05; r2 = 0.42) but directly correlated with increases in the cladoceran to copepod ratio (p < 0.001). In addition, the average body length of the smallest macrozooplanker (Bosmina) has shifted below the observed threshold length of 0.40 mm for juvenile sockeye prey. We also found an inverse relationship between greater parental escapements and age 1. smolt length (p < 0.01; r2 = 0.75). Small-sized smolts produced from a density-dependent rearing environment indicate that spawning capacity equals or exceeds rearing capacity. Juvenile recruits from larger escapements (~400 000) exerted a controlling, density-dependent influence on the secondary trophic level which drastically reduced subsequent adult returns. Moreover, adverse trophic level changes, particularly after the largest escapements, also appeared to reduce the rearing efficiency for lower densities of juveniles, as subsequent smolt sizes from smaller escapements were less than expected. Thus, these findings not only provide a rationale for suggesting that fewer spawners would provide recruits more closely matching the rearing capacity of this lake, but also provide empirical evidence for the potential adverse density-dependent effects of large escapements.

No abstract

No abstract

No abstract