Zusammenfassung: This set of computer programs for stock assessment uses MS FORTRAN 77 version 4.1* for all the programs. The programs will compile under version 3.2 of MS FORTRAN 77, however the Open statements with file types “Unknown” must be changed to “New” or “Old” depending on the expected status of the file. The “shell” called TUNE is used to build the appropriate batch file for submission of the multiple runs used for calibration. Each of these programs is an interactive stand alone module, each can be used independently of the others. This report gives an outline of how and where these programs fit into an analytical process. A set of disks with the source code and the executable (*.EXE) files is available from the editor or the programs can be copied from the ICCAT Secretariat in Madrid; an Annex to this report containing a listing of the source code is also available. A glossary of terms commonly used in the assessment process is presented in Appendix I.

Zusammenfassung: A recently introduced multivariate state-space time-series modeling algorithm, system-theoretic time-series, can be used to create a particularly robust state-space representation. Combining the model approximation methodology with time-series modeling of the residuals in accurate models of short-run dynamics. The approximate structural modeling methodology is illustrated with single species models of yellowfin sole (Limanda aspera ) biomass, and multispecies models of Pacific halibut (Hippoglossus stenolepis ), yellowfin sole, and red king crab (Paralithodes camtschatica ) biomasses in the eastern Bering Sea. A final application of the approximate structural/system-theoretic time-series model includes walleye pollock (Theragra chalcogramma ), Pacific halibut, Pacific cod (Gadus macrocephalus ), yellowfin sole, and red king crab. (Prepared in cooperation with Alaska Univ., Fairbanks.)

Zusammenfassung: An adjustment method for age-length keys is presented that incorporates yearly variability in cohort abundances. The method is based on two models (one for growth and the other for standard deviations of length versus age) and involves an iterative algorithm. The algorithm rapidly converges to stable results that yield, under some conditions, maximum likelihood estimates. It is also used to estimate parameters of the length standard deviation model. The method is applied to yellowfin tuna (Thunnus albacares) catches in the eastern Atlantic from 1975 to 1989.