FISHERY BULLETIN: VOL. 73, NO. 4 



concern to fisheries biologists. A possible explana- 

 tion is that crab catch is merely following the 

 cyclic upwelling. However, the auto-correlation of 

 upwelling (Figure 4) does not indicate significant 

 periodicity and the weak cycles suggested have 

 different periods from those for the catches. This 

 result may be due to the fact that the data used by 

 Bakun (1973) were taken from four different 

 sources covering the time periods January 1946- 

 March 1955, April 1955-December 1959, January 

 1960-June 1962, and July 1962- December 1971. The 

 data within each time period may differ by a mul- 

 tiplicative factor from the others but are consis- 

 tent within the time period. This would affect the 

 auto-correlation function more for longer than for 

 shorter lags. The total effect of possible inconsis- 

 tencies in the data upon the auto-correlation 

 function is difficult to assess since the magnitudes 

 of the multiplicative factors are not known. 



A second possible explanation is that years of 

 good catch indicate that a higher than normal 

 percentage of the population is being caught thus 

 depleting the adult reproductive stock. This 

 explanation, however, is not commensurate with 

 the widely held belief that almost all legal males 

 are caught every year. 



A third possible cause of the cyclic variations in 

 abundance is predation. The observed cycles may 

 be the result of interaction of crabs with a preda- 

 tor species. However, there is no a priori reason to 

 expect that the period of the cycles would be 

 slightly greater than twice the generation time. 

 Also, the large number of predator species sug- 

 gests that predation would maintain a fairly con- 

 stant level unless all predator species were cycling 

 in abundance. 



A fourth possible explanation involves inter- 

 actions between members of the population rather 

 than external causes as dominant factors. Den- 

 sity-dependent factors which limit reproduction or 

 growth and survival of the younger members of a 

 population as the numbers of adults increase may 

 cause cycles in abundance of adults. Ricker (1954) 

 provides a discussion of density-dependent 

 recruitment in fisheries and a simulation of time 

 variation of abundance based on several different 

 density-dependent recruitment relationships. One 

 of the forms of compensatory mortality which 

 could cause density dependent recruitment is can- 

 nibalism (Ricker 1954). Butler (1954) reported 

 finding small Dungeness crabs (about 1 cm) among 

 the stomach contents of older Dungeness crabs. 

 For certain stock-reproduction curves the 



simulated populations oscillated in numbers. The 

 requirements for permanent oscillation were a 

 rapid decrease in reproduction (or survival of the 

 young) beyond the peak of the stock-recruitment 

 curve and mixing of generations in the breeding 

 populations. The period of the oscillations was 

 twice the mean length of time from parental to 

 filial eggs. 



Our results show a period of approximately 9 yr 

 in California and 12 yr in Washington and Oregon. 

 This would indicate a generation time of 4.5 yr in 

 California and 6 yr in Washington and Oregon. 

 Poole and Gotshall (1965) stated that mating oc- 

 curs in California after 2.5 yr and males enter the 

 fishery at 3.5 or 4.5 yr. Cleaver (1949) reported that 

 in Washington maturity occurs at 3 yr and legal 

 size occurs at the end of the fourth year. McKay 

 and Weymouth (1935) reported that female crabs 

 in southern British Columbia probably reach sex- 

 ual maturity during the fourth or fifth year. The 

 longer generation times in the northerly waters 

 are commensurate with our results, although the 

 period of the cycles indicated by our results is 

 greater than would be expected based on Ricker's 

 simulation and estimates of age at maturity. This 

 disparity may be due to inaccuracies in the es- 

 timates or in our results. One possibility is an 

 inaccurate estimate of w^hen, in the life cycle, a 

 major portion of reproduction occurs. If upwelling 

 increases growth rate, the determination may 

 have been made in good years, thereby yielding 

 shorter time periods. A second possible explana- 

 tion is inaccurate determination of the period of 

 the cycle in our analysis. The length of the data 

 (two or three cycles) is not adequate for accurate 

 determination of the period of the cycle, especially 

 when external influences appear to have caused 

 large variations. Examination of the raw data 

 reveals a reasonable explanation and also demon- 

 strates correlations implied in this paper. In 

 Figure 1, catch for northern California goes from a 

 peak at 1950-51 or 1951-52 through a low point to a 

 peak in 1958-59, a 7- or 8-yr cycle. It then passes 

 through a low point and 7 yr later is again at a high 

 point. However, instead of beginning to decrease, 

 it rises even higher for the next several years 

 before dropping. Figure 2 indicates that this is 

 also the time when upwelling index is at a very 

 high value. That the crab population reached a 

 high density in 1965, but did not suffer the ex- 

 pected decline because the upwelling had provided 

 enough food resources to sustain a larger popula- 

 tion is a plausible explanation. 



906 



