Scale Methods for Discriminating between Great Lakes Stocks of Wild and Hatchery Rainbow Trout, with a Measure of Natural Recruitment in Lake Ontario

Abstract

Growth characteristics of scales were used to discriminate between rainbow trout Oncorhynchus mykiss of hatchery and wild origin and to provide a measure of natural recruitment in Lake Ontario. Scale samples were collected in 1971, 1974, and 1989–1992 from 20 sources that included three hatcheries and Lakes Ontario, Huron, and Superior and their tributaries. Spacing of the circuli about the first annulus has been shown to separate wild and hatchery rainbow trout. However, criteria for recognizing and locating the first annulus are lacking, which leads to subjectivity and general imprecision. We developed methods that use the first two scale checks (breaks in the configuration of the circuli) regardless of designation, incorporate differences in the thickness of the circuli, and detect regenerated nuclei so that these scales could be eliminated from the analysis. The dichotomous key we developed provides specific criteria that makes it easy to locate a diagnostic check and discriminate origin of the fish. Relative differences between both spacing and thickness of circuli on either side of the check indicate fish origin. The ratio (mean ± 95% confidence interval) between circulus spacing was 1.11 ± 0.03 for hatchery and 1.52 ± 0.04 for wild; separation criterion (SC) = 1.31; mean error rate (MER) = 14.6%. The ratio between circulus thickness (thick = 7.8 ± 0.3 μm; thin = 5.2 ± 0.2 μm) was 1.07 ± 0.06 for hatchery and 1.50 ± 0.05 for wild; SC = 1.29; MER = 9.4%. When both criteria were combined, discrimination improved (MER = 8.9%; wild identified as hatchery = 10.5%, hatchery identified as wild = 7.2%). This scale method provides an accurate, cost-effective way of recognizing hatchery rainbow trout that have been stocked unmarked. Wild rainbow trout were remarkably abundant in unclipped samples from Lake Ontario (44 ± 6%) and composed the majority of the spawning population (65 ± 10%), whereas hatchery fish were much more abundant in angled samples (70 ± 6%).

Notes

¹ Casselman, J. M., and K. A. Scott. 1991. CSAGES, unpublished software. Ontario Ministry of Natural Resources, Glenora Fisheries Station, Rural Route 4, Picton, Ontario K0K 2T0, Canada.