UC San Diego Library – Scripps Collection
Fish Bulletin No. 120. Estimating Absolute Age Composition of California Salmon Landings
- Author(s): Kutkuhn, Joseph H
- et al.
Drawing valid inferences about parent populations is directly dependent upon the precision or reliability of estimated parameters. Self-evident as this statement is, however, situations frequently arise wherein unwarranted conclusions may be drawn from sample information whose reliability is imperfectly known. Thus it is clear that wherever possible, estimates of any sort should be supplemented with some measure of their precision. This implies, of course, that assumptions underlying any application of modern sampling theory have been reasonably satisfied.
The present paper represents an attempt to assess estimates that permit delineating the absolute age structure of fish catches, assuming the appropriate attributes are observed in a manner that at least approximates probability sampling. Such procedure further provides a basis for modifying sampling designs so future estimates derived therefrom will possess some a priori degree of precision imposed in accordance with available manpower and monetary resources.
Although specifically concerned with determining age composition of California landings of king salmon, Oncorhynchus tshawytscha (Walbaum), the techniques discussed herein could be applied to other species whose catches are handled in similar fashion. Moreover, they would be appropriate where dealing with any of several attributes commonly descriptive of fish populations. Besides the number of scale annuli, constituting the attribute of interest in the present paper, these include: (1) characteristics defining each kind of fish subject to capture; or, perhaps, for a given kind of fish, characteristics that would identify a member of a particular subpopulation, aggregations of which may comprise a commonly fished stock; (2) any artificially applied "mark"; (3) a particular class of lengths or weights; and, (4) sex.
Designed to provide a basis for predicting year-to-year king salmon yield potential, the parent study from which part of my data were selected evolved during the recovery phase of a large-scale salmon marking program. In addition to sampling the California commercial salmon landings for marked fish, it was necessary to sample concurrently for other attributes. During the program's early stages, it was further decided that still others capable of yielding potentially valuable information, viz., length and age, could be sampled incidentally with little added effort and expense. To supplement those secured from the above source, I have used additional data collected during a concomitant study of the ocean sport salmon catch at various points to help develop the methods discussed.