Wilson and Seki: Biology and population characteristics of Squalus mitsukuni 



853 



Total length (TL) was determined to the nearest 1 

 mm by placing the shark on its side and measuring 

 from the anterior tip of the snout to the posterior 

 edge of the upper lobe of the caudal fin in the "natu- 

 ral" upright position. For some specimens, body 

 length (STL) was measured from the anterior tip of 

 the snout to the posterior edge of the upper lobe of 

 the caudal fin after the lobe was depressed to a posi- 

 tion in line with the body axis; fork length (FL) was 

 also measured from the snout to the fork or middle 

 of the caudal fin. Simple linear regression relation- 

 ships among the three length measurements were 

 determined to facilitate comparisons with other pub- 

 lished studies. The relationships are TL= 1.08 x LF + 

 1.67 (^=0.99, rc=342); TL = 0.95 x STL + 0.24 (^=0.99, 

 /i=212); STL=1.01 x FL + 6.31 (rM).91, rc=463). 



A nonlinear estimation procedure (Wilkinson, 

 1988) was used to fit individual length and weight 

 (WT; in grams) data to the relationship, WT = a x 

 TL b . Weight-length relationships between sexes were 

 evaluated with analysis of covariance (ANCOVA) on 

 the log-transformed data. 



Catch per unit of effort (CPUE) for S. mitsukurii 

 was determined by using only specimens taken by 

 bottom longline for several reasons. The bottom 

 trawl, handline, and gillnet gear did not have fish- 

 ing effort comparable with the bottom longline, and 

 the former data were too limited to construct indi- 

 vidual time series for each gear. Furthermore, the 

 vertical longline gear was used to define the vertical 

 distribution of pelagic armorhead above the summit; 

 therefore only a few hooks of uncertain number were 

 on the summit and available to S. mitsukurii. 



Finally, specimens from gear other than bottom 

 longline accounted for only 8% of the total numbers 

 caught. Therefore, it was unlikely that the exclusion 

 of these data from the CPUE calculations would sig- 

 nificantly bias our results. Fishing effort on S. 

 mitsukurii was expressed in number of hooks set, 

 minus hooks that caught pelagic armorhead. Pelagic 

 armorhead respond more quickly to the bottom 

 longline than do S. mitsukurii (Somerton 3 ). Thus, 

 excluding those hooks occupied by pelagic armorhead, 

 although not removing all bias from the species-gear 

 interaction, likely improved estimates of fishing ef- 

 fort on S. mitsukurii over the unadjusted value. 



An estimate of the initial exploitable biomass of S. 

 mitsukurii was made with the Leslie model (Ricker, 

 1975) in which a linear function was fitted to CPUE 

 and cumulative catch data from all cruises. The func- 

 tion has a slope equal to catchability (q) and inter- 

 cept equal to the product of q and the initial exploit- 

 able biomass. The model assumes that changes in 

 CPUE over time are due to fishing and that other 

 sources of losses and additions to the population are 

 relatively minor or in balance (e.g. natural mortal- 

 ity and recruitment). The 95% confidence limit on 

 the abundance estimate was calculated by using the 

 method of Polovina ( 1986). 



First and second dorsal spines were collected dur- 

 ing the summer of 1986 for ageing S. mitsukurii with 

 procedures outlined by McFarlane and Beamish 

 (1987). All age results were based on the second 



3 Somerton, D. National Marine Fisheries Service, Seattle, WA 

 98115. Personal commun., 1992. 



