Multivariate Morphometric Variability in Pacific Salmon: Technical Demonstration

Abstract

A computer oriented approach to the collection and analysis of morphometric characteristics in juvenile chinook salmon (Oncorhynchus tshawytscha) is described. A three-step data collection and storage method is used whereby X–Y coordinate data for relevant morphological features on a body form are collected with a digitizing pad and used to calculate morphometric characters. To test this method, I calculated two morphometric data sets, a conventional and a truss network, and compared them by multivariate analysis in a preliminary study of growth and development in one hatchery stock of fish, and in a survey of population differences in three naturally occurring populations of chinook salmon. Technical advantages of using a digitizing pad for collecting morphometric data are demonstrated. Hatchery-reared chinook salmon showed marked changes in body shape during the period of spring smoltification when marked changes in condition factor occurred. Multivariate differences were discerned among the three Oregon coastal stocks. Truss data provided more specific information concerning shape changes in the study of early development and produced greater between-group differences in the geographic survey. The results of these preliminary analyses can be applied to problems of identifying smolt status in hatchery fish and stock origin in mixed-stock fisheries.