Hannah and Jones: Fishery-induced population changes in Pandatus jordani 



49 



at 16.6mm carapace length. Since 1979, age-1 shrimp 

 are fully recruited to a 38mm trawl in the later months 

 of the season and to smaller gear even sooner. In earlier 

 years, samples were biased toward only larger age-1 

 shrimp, and thus, by comparison, understate the in- 

 crease in growth observed since 1979. 



Our exploratory correlation analysis is inconclusive 

 in differentiating between density-dependent and en- 

 vironmental factors as influences on shrimp growth. 

 While the underlying relationship between the age-1 

 growth index and the environmental variables tested 

 is most likely curvilinear (Ricker 1975), the relatively 

 narrow range of environmental variability being tested 

 in this case warranted the simple linear approximation. 

 The combined age-1 growth index was closely cor- 

 related with our index of shrimp density, CPUE, 

 despite the fact that CPUE is a relatively poor index 

 of density. We showed CPUE to be negatively cor- 

 related with mean size at age over rather large changes 

 in CPUE. Of course, smaller changes in CPUE, not 

 associated with major changes in population density, 

 should be positively correlated with growth, causing 

 CPUE to be a poor index of shrimp density. We also 

 found adjusted sea level at Newport, Oregon in year 

 t- 1 to be closely correlated with the age-1 growth 

 index, indicating that warmer bottom temperatures 

 may have caused improved shrimp growth after 1978. 

 Rothlisberg (1975) showed shrimp growth to be posi- 

 tively correlated with temperature under laboratory 

 conditions. It is possible that elevated sea levels im : 

 prove growth over the normal range observed, but at 



extreme levels such as occurred in 1983, the virtual 

 complete shutdown of coastal upwelling has the reverse 

 effect (Pearcy et al. 1985, Miller et al. 1985). 



The hypothesis that sea surface or bottom tempera- 

 tures (as inferred from sea level data) are controlling 

 shrimp growth will most likely be tested over the next 

 decade or two. In time, the relatively warm ocean con- 

 ditions experienced off the Oregon coast since 1978 will 

 probably be replaced by a colder, upwelling-dominated 

 regime, similar to the early 1970s. The mean increase 

 in length we have measured is equivalent to a 28% in- 

 crease in average weight at age. If sea-bottom tem- 

 peratures are controlling growth and return to lower 

 levels, the drop in fishery yield will be profound. Con- 

 versely, it is unlikely that the shrimp fishery will be 

 substantially reduced, allowing standing stocks of 

 shrimp to rebuild to near virgin levels of the early 

 1970s. Thus, as we see the subsequent trend in mean 

 carapace length at age of pink shrimp, our hypothesis 

 of density-dependent growth will be tested further. 



Charnov et al. (1981) showed that reductions in the 

 population of age-2 and older shrimp (predominantly 

 female) should result in increases in primary females. 

 If Charnov is correct, the trawl fishery, through selec- 

 tive removal of older shrimp, should be causing this 

 effect in the population. The result of accelerated sex 

 change should be higher levels of primary females and 

 a roughly stable sex ratio. Jensen (1965) and Charnov 

 (1980) noted increased levels of young females in 

 Pandalus borealis populations after intensive fishery 

 development. 



We question what these changes in population struc- 

 ture imply for the future productivity of the pink 

 shrimp resource. The evidence for density-dependent 

 growth argues for a harvest-resistant shrimp stock. 

 Our data also support the hypothesis of Charnov et al. 

 (1978) that the population age structure determines the 

 age of sex change in shrimp. As a consequence of in- 

 tensive harvest, the age structure has shifted toward 

 younger shrimp. The percentage of primary females 

 has increased, however, resulting in the maintenance 

 of a sexually balanced breeding population. The capa- 

 city to accelerate sex change in pink shrimp also in- 

 creases the stock's ability to withstand harvest pres- 

 sure, by decreasing the potential for declines in larval 

 production. 



Both catch and effort levels in the pink shrimp fishery 

 are continuing to increase. The preliminary total catch 

 for the states of California, Oregon, and Washington 

 is nearly 36,000 mt in 1989. The large harvests in 

 1987-89 (Fig. 11) appear to be the result of a combina- 

 tion of factors. Improved CPUE in 1987 and 1988 

 (Fig. 2; 1989 data unavailable) indicate some strong 

 year classes of shrimp moving through the fishery. The 

 total harvest levels of age-1 shrimp in these years is 



