Morphometric Analysis of Tule Perch (Hysterocarpus traski) Populations in Three Isolated Drainages

Abstract

Morphometric analysis of tule perch from six sites in three isolated drainages was conducted to assess the validity of previous systematic work. Nine variables were subjected to principal component and discriminant analysis. Principal component analysis indicated that groupings by drainage were real. Classification by discriminant functions among drainages was 96% accurate. Classification among collection sites was less accurate (88% overall), but most misclassified cases were assigned to another site in the same drainage. Ecophenotypic variation was apparent but the degree of overlap between drainages was minimal and indicated that the within-drainage differences were less than the among-drainage differences. Our analysis supports the view that populations in separate drainages are sufficiently distinct to warrant subspecific designation. T HE tule perch (Hysterocarpus traski) is the only exclusively freshwater member of the surfperch family (Embiotocidae) and is confined to three drainages in central California. It is unique among the freshwater fauna of California in that it is the only native viviparous fish. One population is largely confined to two lakes in the Clear Lake drainage basin, another is confined to the Russian River, a large coastal stream, and the third is found in a wide variety of habitats in the Sacramento-San Joaquin drainage including small tributary streams, main river channels, impoundments, and sloughs in the Sacramento-San Joaquin Estuary. Hopkirk (1962, 1973) recognized the populations in each drainage as morphological subspecies but Hubbs (1974) questioned Hopkirk's findings, mostly on the basis of his methodol