DISCUSSION

Several studies have used controlled experiments to demonstrate fish predation on eastern ambystomatid larvae (Petranka 1983, Semlitsch 1988, Sih et al. 1992), and numerous researchers have reported that the presence of fish reduces the abundances of larval ambystomatids (Burger 1950, Blair 1951, Sexton and Phillips 1986, Semlitsch 1988).

In the present study, survivorship of A. macrodactylum and A. gracile larvae was significantly lower in artificial ponds with trout than in fishless ponds. These findings support hypotheses from field studies that larval abundance of both species is reduced in the presence of trout (Sprules 1974; Taylor 1983, 1984; Tyler et al. 1998). Trout predation is likely responsible for reduced larval survival in artificial ponds. However, reduced foraging opportunities due to the threat of predation may have resulted in starvation of larvae, thus, influencing larval survival.

At the termination of the experiments, larval A. macrodactylum and A. gracile occurred in a narrower range of habitats in artificial ponds with trout than in ponds without trout. The ability to detect predators and select habitats providing refugia from predation has been documented for larvae of several eastern ambystomatid species (Kats et al. 1988, Sih et al. 1992, Sih and Kats 1994), including increased refuge use (Sih et al. 1988) and decreased activity levels in the water column (Stangel and Semlitsch 1987, Figiel and Semlitsch 1990). In the present study, A. macrodactylum in ponds with trout were observed only in rock substrates while larval A. gracile in ponds with trout were found in both rock and wood substrates. Larvae of both species in fishless ponds were observed in all habitats, including open areas. Of the substrates provided, rock and woody material likely provided larvae with the greatest protection from predation.

Larval size of both A. macrodactylum and A. gracile was influenced by trout; both species exhibited greater SVLs in control ponds than in fish ponds at the conclusion of experiments. However, only A. gracile had greater body mass in fishless ponds than in ponds with trout. Semlitsch (1987) and Figiel and Semlitsch (1990) observed reduced growth in SVL of larval A. maculatum when raised in tanks with a predatory sunfish.

Reduced foraging opportunities associated with increased refuge use or competition between larvae and trout for a limited food resource could have reduced the growth of larval A. macrodactylum and A. gracile (Semlitsch 1987, Sih et al. 1988, Figiel and Semlitsch 1990). However, we were unable to determine which of these mechanisms was more important.

Our findings support inferences from field studies that trout can reduce survival of A. macrodactylum and A. gracile larvae. Our findings also indicate that trout can affect larval habitat selection and growth rates. Slower growth rates may reduce salamander survival either through increased length of the larval period and, thus, increased susceptibility to mortality factors in the aquatic environment (e.g., predation, lake desiccation, or lake freeze). Slower growth rates may also decrease larval size at metamorphosis, thus increasing susceptibility to terrestrial threats (e.g., desiccation, amphibian requirements of osmoregulation and thermal regulation).

ACKNOWLEDGMENTS

We are grateful to the former and present members of our scientific advisory panel: S. Loeb, S. Dodson, R. Hughes, W. Neill, W. J. O'Brien, J. Petranka, W. Platts, and H. B. Shaffer. This research would not have been possible without the co-operation and logistical support provided by the personnel of North Cascades National Park Service Complex. This research was funded by the National Park Service and the USGS-Forest and Rangeland Ecosystem Science Center.