FISHERY BULLETIN: VOL. 79, NO. 3 



70-1 



60- 



E 



I 

 I- 

 O 



z 



LU 



o 



50- 



40 



O E.Bismarck Sea 



□ New Hanover 



• S. New Hanover 



A N. New Hanover 



^ Manus 



(y'©''''^^^ 



Figure 3. — Modal fork lengths of samples of skipjack tuna from all Papua New Guinea areas as a function of time (in months). Modal 



progressions are indicated by lines. 



the Papua New Guinea fishery, observations have 

 shown that vessels continually pursue and fish an 

 individual school during periods of low catch per 

 unit of effort (CPUE), thereby fishing few schools 

 each day. However, during periods of high CPUE, 

 vessels fish a large number of schools, taking fish 

 from each school until they "go off the bite" and 

 then moving on to a fresh school. Size-specificity of 

 the catch, as a consequence of the fishing strategy 

 adopted, would therefore be expected to be greatest 

 during periods of high CPUE, vessels taking only a 

 restricted size range offish from each school (pre- 

 sumably those size classes most vulnerable to this 

 method of fishing), and lowest during periods of 

 low CPUE, during which a wider range of size 

 classes would be represented in the catch. 



Figures 3 and 4 show two periods, November 

 1977 to May 1978 and May 1979 through to the end 

 of sampling in December 1979, during which a 

 wide range of size classes appeared in the samples. 

 These periods coincide with periods of relatively 

 low skipjack tuna CPUE (Table 1). The situation is 

 less clear for yellowfin tuna. However, since skip- 

 jack tuna composed 90% of the total catch on aver- 

 age, it is clear that skipjack tuna abundance would 



determine the adoption of a particular fishing 

 strategy. This would therefore account for the yel- 

 lovi^n tuna size range, varying synchronously 

 with that of skipjack tuna (Figures 3, 4), but inde- 

 pendently of yellowfin tuna abundance (Table 1). 



Observations from the fishing vessels during the 

 course of the study confirmed that yellovdin tuna 

 larger than the maximum size taken by the 

 fishery and schools of very small skipjack tuna, 

 both normally comparatively rare, were common 

 throughout the fishery during late 1977 and from 

 August to November 1979. There is therefore some 

 indication that fishing strategy, determined by low 

 apparent abundance, may not alone account for 

 the appearance of small and large fish in the catch 

 during these periods. 



The results show that the timing of recruitment 

 cannot be determined from the size composition of 

 the landed catch. However, Ueyanagi (1970), 

 Nishikawa et al. (1978), and Naganuma (1979) 

 have demonstrated that skipjack tuna spawn 

 throughout the year in the western Pacific, al- 

 though in different geographical areas depending 

 on season: a situation likely to result in continu- 

 ous recruitment to the equatorial region. The re- 



522 



