LONGITUDE 



o 

 o 



^ 



Q.3i3tjujuja.3 



MONTH 



Figure 4. --Es t imated average apparent abundance of albacore (in numbers of fisin 

 caught per hundred hool<s) in IS'*? and in I96I for Marsden squares 90-95 and 

 126-130. Circles denote slopes that were significant at the 5-percent level. 



ing that there was a real reduction in CPUE 

 over the study period. 



In order to express the decline in abundance 

 for the entire longline fishery we averaged the 

 slopes and intercepts over the fishing area 

 shown in figure 4 for those months (January, 

 February, and March) for which the data were 

 most abundant. Averaging was accomplished 

 by weighting each slope and intercept by the 

 inverse of its variance. The weighted averages 

 along with the statistics used to compute them 

 are set forth in table 1. We note that the 

 weighted average intercept is at its peak in 

 February, having an average value of 4.13 

 albacore per hundred hooks. The weighted 

 average slopes are remarkably similar, espe- 

 cially considering the many possible sources of 

 variation. There is an average loss of about 

 0.2 fish per hundred hooks per year. Thus the 

 February CPUE of 4.13 albacore per hundred 

 hooks would be reduced, after 13 years, to a 

 value of 1.53 fish per hundred hooks, indicating 

 a nearly 60-percent reduction in apparent abun- 

 dance and implying a corresponding decrease 

 in actual abundance. If we treat the February 

 CPUE as a simple index and assume that the 

 trends in apparent abundance actually reflect 

 trends in actual abundance, then a reduction of 

 0.2 fish per hundred hooks per year in apparent 

 abundance would provide us with a crude esti- 

 mate of the increase in the per-head rate of 

 death of O.l per year. This increase places the 

 total instantaneous mortality coefficient of the 



albacore in the longline fishery at a minimum 

 value of 1.3 at the end of the 13-year study 

 period. If we apply this mortality coefficient 

 to the yield isopleths computed by Suda (Kami- 

 mura, 1966: p. 765) and if our assumptions are 

 reasonable, either a reduction in fishing inten- 

 sity or an increase in recruitment size would 

 probably produce an increased yield per recruit 

 on the longline fishing grounds. 



It total mortality did, in fact, increase on the 

 albacore longline grounds, during the study 

 period, then, all other things being equal, we 

 should expect a decrease in the average age 

 and consequently in the average size of alba- 

 core taken on the longline ground. Such a re- 

 duction in average age was indicated by Suda 

 (1959) and can also be seen qualitatively in 

 another paper by Suda (1963b: p. 1250). On the 

 otlier hand, we have computed the average 

 lengths of the albacore for the North American 

 (the California fishery), the longline, and the 

 pole-and-line fisheries from data given by 

 Clemens and Craig (1965) and Suda (1963a). We 

 have plotted these data in figure 5 and note that 

 a decline in average length for these fisheries 

 is not apparent. We cannot, at present, recon- 

 cile the decline in one set of the Japanese data 

 and the lack of a decline in the other. 



Deviations from the Long-term Trend 



In this section we consider the deviations 

 from the regressions of average CPUE on year 

 for the Marsden squares in the albacore area. 



