172 



Fishery Bulletin 



1990 



that month. Monthly values are joined by line segments 

 in the figures: segments with steep slopes are associ- 

 ated with a relatively large proportion of total regional 

 effort. Conversely, if little or no effort took place within 

 a month, the corresponding segment will be nearly 

 horizontal. 



The interpretation of the cumulative time density 

 curves is related to the pattern of hake migration. For 

 example, effort started to accumulate in region 1 (Fig. 

 6a) around April or May, when fish first arrived in this 

 area on the northward leg of their migration. The S- 

 shaped 1978 curve indicated that there was a single 

 peak in the deployment of effort during winter months 

 (May-August). In contrast, in 1979 most of the effort 

 was deployed between March and June, with a second- 

 ary peak in September-October. 



In region 4 (Fig. 6b), effort accumulated rapidly dur- 

 ing February, March, and April of 1978, as the hake 

 were migrating northwards along the shelf break. 

 Afterwards, effort decreased greatly. In 1979, the ini- 

 tial effort was much like the previous year's, but it 

 tapered off until June, July, and, particularly, August 

 (winter). Effort in winter months was not observed in 

 1978. 



In region 8 (Fig. 6c), the effort patterns were re- 

 markably similar for both years. The curves were flat 

 between March and September, as both the fish and 

 the fleet were absent from the region. During October, 

 November, and December the spawners arrived in the 

 area and fishing effort accumulated rapidly. 



Interannual differences between cumulative time 

 density patterns were statistically analyzed using the 

 Kolmogorov-Smirnov (K-S) two-sample test. Because 

 of the large N values (the annual total number of hours 

 trawled in a region), all the K-S tests were significant 

 (see Rugolo 1984), including those for regions not 

 shown. As a consequence, the discussion will focus on 

 regions 1 and 4, where the patterns appeared to differ 

 most between the two years. 



From the patterns in regions 1 and 4, it appears that 

 the fishing fleet (and, thus, hake) was located further 

 south in 1979 than in 1978. The spatial statistics of 

 fishing effort also reflect a more southerly deployment 

 of effort in 1979. The latitudinal means of fishing effort 

 for July and August 1978 were, respectively, 36.4°S 

 and 36.6° S; the corresponding values for 1979 were 

 38.2°S and 40.2°S. That is, the "center of mass" 

 of fishing effort in the winter of 1979 was located 

 2-4 degrees of latitude further south along the shelf 

 break than in 1978. The changes in hake distribution 

 may reflect interannual differences in environmental 

 conditions. 



The northernmost extension of the hake feeding 

 migration, reached during June- August, is associated 

 with the Brazil-Malvinas confluence. The confluence is 



C 80 



o 

 o 



Q. 60 

 O 



> 



TO 



3 



E 



D 



o 



Buenos Aires 

 Slope North 



C 80- 

 O 



8. 60 

 O 



0) 

 > 



— 20 



E 



D 



O 



c 

 o 



g_60i 



o 

 qI 



> 



3 



5 

 o 



2 i A b6789 



Month 



Patagonian 

 Slope North 



1979 



23456789 



Month 



234 56769 



Month 



Figure 6 



( 'uinulative time densitifs for (.1 ) Bueno.s Aires sliipe iKjrth 

 (region 1), (/?) Patagonian slope north (region 4) and 

 ((') southern spawning grounds (region 8). Location of 

 regions is shown in f^igure 1. 



an effective barrier for adult hake of commercial size: 

 aggregations of hake found north of it are mainly 

 formed by juveniles (Rojo and Silvosa 1969). Olson 

 etal. (1988) showed considerable fluctuations in the 



