772 



Fishery Bulletin 88(4), 1990 



mented by an external source of information. For 

 SRLCA to be effective, the subjectivity required in the 

 selection of input parameters for use with the family 

 of distribution mixture methods (e.g., MacDonald and 

 Pitcher 1979) must instead be applied to interpretation 

 of the output of the procedure. 



These problems are similar to those encountered by 

 others in evaluating the ELEFAN I method of Pauly 

 and David (1981). Testing of ELEFAN I has indicated 

 it produces biased results unless the range of growth 

 parameters considered for testing is relatively narrow, 

 and that it is sensitive to increased variation in length- 

 at-age or increased variation in recruitment timing 

 (Rosenberg and Beddington 1987). Recent work by 

 Morgan (1987), supplementing ELEFAN with age- 

 length data, improved the performance of that method 

 by allowing selection of the appropriate parameters 

 from among several locally optimal solutions. 



We are encouraged, however, that such external in- 

 formation requirements were moderate for SRLCA. 

 For instance, with an expectation of age of the oldest 

 cohort, we would have immediately selected the quater- 

 nary ridge of the pooled analysis as best. Or, had we 

 limited the initial search space for Li„f to within 10% 

 of the CL of the largest animal, and/or 10% of an 

 "average" L,„f given previously derived parameters 

 (e.g., 33 + 2 mm), we again could have proceeded direct- 

 ly to selection of the appropriate score ridge. 



Even given conditions of variable growth and recruit- 

 ment, we were able to inspect the response surface of 

 the test function and, using a moderate degree of sub- 

 jectivity and a stepwise procedure of evaluation, select 

 growth parameters from alternative high-score ridges 

 that resolved the pooled distribution in a manner con- 

 sistent with previous interpretations, and, more impor- 

 tant, in a more satisfactory manner than previously 

 derived von Bertalanffy growth parameters. Overall 

 we found SRLCA to be a simple and generally effec- 

 tive tool for the estimation of growth parameters, and 

 subsequently age frequencies, from length-frequency 

 distributions for Gulf of Maine northern shrimp. 



Acknowledgments 



We thank the ships officers, crew, and scientific staff 

 of the RV Gloria Michelle and the members of the 

 Northern Shrimp Technical Committee for their efforts 

 in collecting the data on which this paper is based. We 

 also acknowledge Dr. Steve Clark, Dr. Wendy Gabriel, 

 the members of the 1989 NAFO Working Group on 

 Progress in Age Determination of Pandalus, and an 

 anonymous reviewer for their valuable comments on 

 the manuscript. 



Citations 



.Vllen, J. .A. 



1959 On the biology of Pnndiilus biirealit; Kr^yer, with refer- 

 ence to a population off the Northumberlami Coa.st. J. Mar. 

 Biol. .Assoc. U.K. ;58:189-2'20. 

 .\pollonio, S., D.K. Stevenson, and E.E. Dunton Jr. 



1986 Effects of temperature on the biology of northern shrimp, 

 Fandahis borealis, in the Gulf of Maine. NOAA Tech. Rep. 

 NMFS 42, 22 p. 

 Basson, M., A. A. Rosenberg, and J.R. Beddinprton 



1988 The accuracy an<i reliability of (wo new nifthods for 

 estimating growth parameters from length-fre(.iuenc.\' data. ,J. 

 Cons. Int. Explor, Mer 44:277-28.5. 

 Blott, A.J., P.J. Diodati, S.H. (lark, D.B. Sampson, and 

 D.F. Schick 



198.'5 Development of a new research trawl for northern 

 shrimp, Piindaliif^ boiritlia. in the western Gulf of Maine. 

 Int. Counc. E.xplor. Sea ICES CM 198.3/8:21, (5 p. 

 Cassie. R.M. 



19.54 Some uses of probability paper in the analysis of size 

 frequency distributions. Aust. .1. Mar. Freshwater Res. b: 

 51.3-522.' 

 Frechette. J., and D.G. Parsons 



198;i Report of a shrimp ageing workshop held at Ste. Foy, 

 Quebec, in May and at Dartmouth. Nova Scotia, in November 

 1981. Northwest Atl. Fish. Organ. Sci. Counc. Stud. ^r. 

 79-100. 

 Hasselblad. V. 



19()6 Estimation of parameters for a mixture of normal dis- 

 tributions. Technometrics 8(,3):431-444. 

 Haynes, E.B., and R.L. Wigley 



1969 Biology of the northern shrimp. Paiidnhis bori'nlis. in the 

 Gulf of Maine. Trans. Am. Fish, Soc. 98:60-76. 

 MacDonald, P.D.M.. and T.J. Pitcher 



1979 Age groups from size frequency data: A versatile and ef- 

 ficient method of analyzing distribution mixtures. J. Fish. Res. 

 Board Can. 36:987-1001. 

 McCrary, J. A. 



1971 Sternal spines as a characteristic for differentiating be- 

 tween females of some Pandalidae. J. Fish. Res. Board Can. 

 28:98-100. 

 Mclnnes, D. 



1986 Interstate fishery management plan for the northern 

 shrimp {Pandalus borealis Kr^yer) fishery in the western Gulf 

 of Maine. Atl. States Mar. Fish. Comm. Fish. Manage. Rep. 

 9, 79 p. 



McNew, R.W., and R.C. Summerfelt 



1978 Evaluation of a maximum-likelihood estimator for anal- 

 ysis of length-frequency distriimtions. Trans. Am. Fhh. Soc. 

 107:703-736. 

 Morgan, G.R. 



1987 Incorporating age data into length-based st<iek assess- 

 ment methods. In Pauly, D., and G.R. Morgan (eds.). Length 

 based methods in fisheries research, p. 137-146. ICLARM 

 Conf. Proc. 13, Int. Cent. Living Aquat. Resourc. Manage,, 

 Manila. Philippines, and Kuwait Inst. Sci. Res., Safat. Kuwait. 



NSTC (Northern .Shrimp Technical Committee) 



1984 Cr-uise results. Gulf nt .Maine northern shrimp survey. 

 Unpubl. rep.. Salem, MA. 16 p. 



1985 (bruise results. Gulf of Maine northern shrimp survey. 

 Unpubl. rep., Salem, MA, 25 p. 



1986 Cruise results, Gulf of Maine northern shrimp survey. 

 I'npulil. rep., Salem. M.'\. 26 p. 



