Terceiro and Idoine: SRLCA application to Pand3lus borealis survey data 



765 



Table 1 



Von Bertalaiiffv growth parameter [jairs ami SRLCA scoi'e 

 (S) for 1984 northern shrimp survey length-frequency distribu- 

 tion: primary and secondary ridge crests. 



L,„( (mm) 



K 



Primary ridge crest 



Secondary ridge crest 



'High score parameters of ridge crest, on hnundary of ex- 

 plored space. 



'High score parameters of ridge crest, nonboundary max- 

 imum. 

 'Parameters selected for final evaluation. 



group 3, and animals >25 mm defined as age 4-t- 

 (NSTC 1984). SRLCA scores exhibited a broad, pri- 

 mary (highest scores) ridge of scores within the ex- 

 plored parameter space ranging from a high score on 

 the border of explored parameter space at Linf = 39 

 min, K = 0.50 to L,nf = 50 mm, K = 0.30, with a non- 

 boundary local maximum at L,„i = 41 mm, K = 0.43 

 (Table 1, Fig. 3). Based on parameters in this ridge, 

 the length frequency was interpreted as a single domi- 

 nant mode (age-group 1), with the distribution to the 

 right of the mode classified as two older age-groups, 

 for a total of three cohorts present. A secondary ridge 

 of scores, from Li„f=33 mm, K = 0.50 to L|„f=38 

 mm, K = 0.30, with a nonboundary local maximum at 

 L,„f = 36 mm, K = 0.37 (Table 1, Fig. 3), classified four 

 ages and defined the dominant mode as age 2, but did 

 not interpret the small modes at 22.5 and 25.5 mm in 

 accordance with prior assumptions (assumed age 

 groups 3 and 4, respectively). However, this secondary 

 ridge most closely matched the n priori interpretation, 

 and represented the best performance of SRLCA for 

 the 1984 data (Fig. 4). A tertiary ridge (a spur of the 

 secondary) ranged from a nonboundary maximum at 

 L,„f = 31 mm, K = 0.46 to L,„,-= 34 mm, K = 0.30, and 

 interpreted the length frequency in nearly the same 

 manner as parameters from the secondary ridge. 



1 985 distribution Previous vvoi'k suggested that the 

 first two modes of this distribution. 13-18 mm and 

 20-24 mm, should be interpreted as age groups 1 (1984 

 YC) and 3 (the strong 1982 YC), thus confronting 

 SRLCA with a "missing cohort" situation (the weak 



Figure 3 



Response surface of the SRLCA score function for the noi-thern 

 shrimp survey 1984 length-frequency distribution. 



Figure 4 



Northern shrimp survey 1984 length-frequency distribution (histo- 

 gram) and pattern of SRLCA scores (solid curve) for growth param- 

 eters selected for final evaluation (L,„, = 36 mm. K = 0.37). 



1983 YC; Fig. 2). Sex determinations further showed 

 that shrimp in the 20-24 mm mode were not the ex- 

 pected first-year females, but mostly mature males that 

 seemed not to have undergone transition during the 

 previous winter (NSTC 1985). Inspection of the posi- 

 tion of this age-3 mode, relative to the modal lengths 



