CRAB SEASON YEARS (JANUARY - AUGUST ) 



_ U) _ O) 



in in iO 



CRAB SEASON YEARS (DECEMBER - AUGUST) 



1950 1955 I960 1965 1970 



UPWELLING YEARS (MAY - SEPTEMBER ) 



Figure 3. — Upwelling index values compared with crab 

 catch in Washington. Crab catches are plotted opposite 

 summer upwelling Vi yr earlier. 



The northern California upwelling-crab data 

 (Figure 5) have paralleleci Oregon's since the 

 summer of 1959. The upwelling data for the 

 summers of 1949-55 would match crab catch 

 better if the crab data were lagged Vi yr (i.e., 

 crab data shifted to the right 1 yr) instead of 

 IV2 yr. The high crab catches of the 1958-59 and 

 1960-61 seasons may have resulted from exploi- 

 tation of newly discovered crabbing areas, as 

 suggested earlier. 



In Washington State (Figure 3) trends are 

 not apparent in the 1949-59 data set, but from 

 the 1960-61 fishing season until 1971-72 the 

 trend is similar to Oregon's trend. 



Each upwelling-crab catch data set can be 

 tested statistically for associations between the 

 magnitude of the upwelling index and crab 

 catch. Scatter diagrams of upwelling index vs. 

 crab catch were prepared for lagtimes of V2 to 

 4V2 yr for Washington and Oregon, and V2 to 



CRAB SEASON YEARS (DECEMBER -AUGUST) 



500 



X 



UJ 



o 



z 



a. 



3 



I 

 u 



H 

 < 



CQ 

 < 



a. 



1950 1955 I960 1965 1970 



UPWELLING YEARS (APRIL-SEPTEMBER) 



Figure 4. — Upwelling index values compared with crab 

 catch in Oregon. Crab catch, in millions of pounds, is 

 plotted opposite upwelling l'/2 yr previous. 



1950 1955 I960 1965 



UPWELLING YEARS (MARCH-SEPTEMBER) 



Figure 5. — Upwelling index values compared with crab 

 catch in northern California. Crab catch is lagged IV2 yr. 



31/2 yr for northern California. (California crabs 

 mature 1 yr earlier than those from more nor- 

 therly points (Poole and Gotshall, 1965).) 



Figure 6 is Oregon's set of upwelling-crab 

 catch scatter diagrams for the indicated lag 

 times. (The Washington and northern California 

 scatter diagrams are not illustrated but results 

 of statistical tests performed upon the scatter 

 are summarized in Table 1.) The median values 

 of the upwelling index and crab catch partition 

 each scatter diagram into quadrants. The first 

 and third quadrants (upper right and lower 

 left) represent strong upwelling-good crab 

 catch and weak upwelling-poor crab catch, re- 

 spectively. The second and fourth quadrants 

 contain data points with the inverse relation- 

 ship. 



A number of statistics can be applied to these 

 scatter diagrams to test the null hypothesis of 

 independence, i.e., no relations between up- 

 welling and crab catch. One of the tests is the 

 corner test (Tate and Clelland, 1957:67). This 

 test is useful in exploring general associations 

 in a scatter diagram. Results of the test (Table 

 1) indicate that at the 5% level, a significant 

 association exists between upwelling and crab 

 catch V2 yr later in Washington, Wi and 2V2 

 yr later in Oregon, and V2, IV2, and 2^/2 yr 

 later in northern California. Such a close time 

 link between an environmental factor (upwell- 

 ing) and its hypothesized effect upon a fishery 

 implies that the fishery is dependent upon sur- 

 vival of older individuals, rather than juvenile 

 individuals. 



906 



