INTRODUCTION

Larval ambystomatid salamanders are important vertebrates native to montane lakes of western North America (Dodson 1970, 1974; Dodson and Dodson 1971; Sprules 1972), and function as top carnivores in fishless lakes (Taylor 1983). Since the late 1880's, trout (Oncorhynchus sp.) and char (Salvelinus sp.) have been introduced to many high-elevation lakes in the western USA that were once fishless (Bahls 1992). Several survey studies have provided evidence that trout and char can eliminate ambystomatid salamanders from or reduce their abundance in montane lakes (Burger 1950, Blair 1951, Sprules 1974, Taylor 1983).

A combination of field and laboratory studies demonstrate the effects of fish on native ambystomatids of eastern and midwestern USA. Several field studies show that fish, particularly predatory fish, can reduce or eliminate larval ambystomatids from ponds, lakes, and streams (Thompson et al. 1980, Petranka 1983, Sexton and Phillips 1986, Semlitsch 1988, Sih et al. 1992). Predatory fish can also inhibit growth (Semlitsch 1987, Figiel and Semlitsch 1990) and reduce survival of larval ambystomatids in controlled experiments (Semlitsch 1987, Sih et al. 1988). Furthermore, fish can influence larval ambystomatid behavior. In the presence of fish, larvae may increase nocturnal activity (Sih et al. 1992), restrict activity to benthic substrates (Semlitsch 1987, Stangel and Semlitsch 1987, Figiel and Semlitsch 1990, Jackson and Semlitsch 1993, Sih and Kats 1994), or increase time spent in refuge (Sih et al. 1988, 1992).

Rainbow trout (O. mykiss) regularly prey upon perennibranciate northwestern salamanders (Ambystoma gracile) in Marion Lake, British Columbia, Canada (Efford and Mathias 1969). Fish predation by this and other salmonids is likely responsible for lower densities of A. gracile in lakes with fish versus no fish (Sprules 1974, Taylor 1983). Gilled A. gracile may behaviorally adapt to fish by increasing nocturnal activity (Taylor 1983, 1984).

Effects of fish on larval long-toed salamanders (A. macrodactylum) and A. gracile have been inferred from comparison of abundance and behavior of larvae in fish versus fishless lakes. For example, in mountain lakes of northern Washington, USA, densities of A. macrodactylum are significantly lower in lakes with stocked trout than in fish lakes (Tyler et al. 1998). We are unaware of any experimental studies that have documented the effects of fish on A. macrodactylum and A. gracile larvae. The objective of this research was to quantify the impact of trout on survivorship, growth, and refuge use of larval A. macrodactylum and A. gracile in controlled laboratory experiments. Specifically, we tested the null hypotheses that survival, growth, and refuge use of larvae did not differ between artificial ponds with fish and without fish.