118 



Fishery Bulletin 101(1) 



Analysis of covariance («=0.05) was used to determine if 

 there were significant differences in the weights-at-length 

 (FL) relationships between sexes for P multidens. Length 

 and weight data were transformed to natural logarithms 

 to satisfy assumptions of normality and homogeneity. 

 Multiple comparisons were performed with Tukey's hon- 

 estly significant difference (HSD) test. Trends in mean 

 length and weight offish over time were assessed by using 

 analysis of variance (a=0.05). 



Otolith preparation and analysis 



Otolith removal, measurement, and preparation followed 

 the procedures and protocols described in Newman et al. 

 (1996), Newman et al. (2000b), and Newman and Dunk 

 (2002). All age estimates were based on the analysis of 

 thin transverse sections of otoliths. These thin sections 

 were examined under a dissecting microscope at 10-30x 

 magnification with reflected light on a black background. 

 The otoliths from eight juvenile P. multidens (80-140 

 mm FL) were examined for daily bands with a different 

 technique. One sagitta per fish was embedded in epoxy 

 resin and a thick transverse section (>500 pm) was cut. 

 The section was then ground and polished from each side 

 to a level near the core (perpendicular to the long axis of 

 the otolith) by hand with ebony paper (1000 grade) and 

 lapping film (9 and 3 pm). A polished thin transverse sec- 

 tion approximately 100 pm thick was produced. The sec- 

 tion was then examined with a compound microscope. 



Age validation 



Marginal increment analysis, routinely used to validate fish 

 age, relies on the assumption that if a translucent zone is 

 laid down once a year, there should be a clear pattern of peri- 

 odic growth on the edge of the otolith during the year. Mar- 

 ginal increment analysis is appropriate only if all fish in the 

 population lay down the translucent zone at the same time. 

 Thus, an annulus consists of a single opaque and a single 

 translucent cycle within a r2-month period. The opaque 

 zone is believed to form during periods of slow growth. 



Marginal increment analysis usually implies measure- 

 ment of marginal growth and hence many researchers 

 have measured the width of the edge of the otolith sec- 

 tion over an annual cycle. This measurement approach 

 has an advantage in that it should be possible to plot 

 growth of the edge over time to validate that only a single 

 translucent mark is laid down each year. However, in P. 

 multidens, it can be difficult to determine a consistent 

 location to measure on the otolith because of the inherent 

 variability of their otoliths; hence this technique was not 

 used in the present study 



Edge type analysis was adopted for the marginal incre- 

 ment analysis of P. multidens and edge types were clas- 

 sified according to Pearson (1996) as either translucent, 

 narrow opaque (opaque area less than half of the previous 

 opaque zone), or wide opaque (opaque area greater than 

 half of the previous opaque zone). Sectioned otoliths offish 

 of all ages were examined under a dissecting microscope 

 with reflected light on a black background. 



Age determination 



Because the peak spawning period off! multidens occurs 

 in late March, all fish were assigned a birth date of 1 

 April to assure proper year-class identification. Ages were 

 assigned from counts of annual growth increments con- 

 sisting of alternating opaque and translucent rings from 

 sectioned otoliths (opaque rings were counted). Annual 

 growth increments were counted in the ventral lobe of the 

 otolith from the primordium to the proximal surface, as 

 close as was practicable to the ventral margin of the sulcus 

 acousticus. Annual growth increments were counted with- 

 out reference to fish length or date of capture. Each otolith 

 section was examined on four separate occasions. When 

 the counts differed, otolith sections were re-examined. In 

 most cases that required resolution, the fourth and final 

 count was used for analysis of age and growth because 

 by this time considerable experience had been gained in 

 the interpretation of the otolith structure. Otoliths with 

 structural irregularities (such as unusual calcification, 

 deterioration of the ventral lobe, or poorly defined annual 

 growth increments) were considered indecipherable and 

 were excluded from analysis offish age. 



Counts were compared and the precision of age esti- 

 mates were calculated with the average percent error 

 (APE) of Beamish and Fournier (1981). Greater precision 

 is achieved as the APE is minimized. The relationship be- 

 tween fish length (FL) and age and otolith dimensions was 

 assessed with linear regression techniques. 



Timing of translucent zone formation in P. multidens 

 and mean sea surface temperatures (SST was assumed to 

 reflect the temperature at depth) were compared by scal- 

 ing values from the two data sets. The scaling process al- 

 lowed direct comparison of each series and any time lags 

 of one in relation to the other. Using the scaling score = 1 



- ((maximum data value - data value) -^ range), where (in 

 the month of November) mean SST for the month was 29°C, 

 the maximum for the year (data set) was 29.7 and the range 

 of the data values was 3.7, we calculated the scaled SST = 1 



- ((29.7-29) -=■ 3.7) = 0.81; in addition the scaled '7c frequency 

 of otoliths with translucent edge types = 1 - ((67-20) -^ 67) 

 = 0.30. 



Growth and mortality models 



The von Bertalanffy growth function (VBGF) was fitted to 

 estimates of length-at-age with nonlinear least squares esti- 

 mation procedures. The VBGF is defined by the equation 



L, = L„[l- exp[-K{t - 1^)]], 



where Lf = mean length offish of age t; 

 L , = asymptotic mean length; 

 /f = is a rate constant that determines the rate at 



which Lf approaches Lj, 

 t = age of the fish; and 



^11 = the hypothetical age at which the mean 

 length is zero if it had always grown in a 

 manner described by the VBGF. 



