Ecological Impact of Introduced Trout on Native Aquatic Communities in Mountain Lakes
North Cascades National Park Service Complex, WA, USA: Phase III Final Report

Changes in the Behavior of Ambystoma gracile Larvae After the Removal of Fish from a Mountain Lake

Robert L. Hoffman¹, Torrey J. Tyler², William J. Liss², and Gary L. Larson³

---

¹ College of Oceanic and Atmospheric Sciences, 104 Ocean Admin, Oregon State University, Corvallis, OR 97331

² Department of Fisheries and Wildlife, 104 Nash Hall, Oregon State University, Corvallis, OR 97331

³ USGS Forest and Rangeland Ecosystem Science Center, 3200 SW Jefferson Way, Corvallis, OR 97331

---

ABSTRACT

Ambystoma gracile (Baird) larvae were surveyed in two adjacent lakes in North Cascades National Park Service Complex, Washington, USA, between 1994 and 1999. One lake (Upper Panther) was fishless. The other lake (Lower Panther) was inhabited by introduced trout (Oncorhynchus clarki). The objective of this research was to identify any changes in larval salamander behavior after fish removal from Lower Panther was completed in June 1997. We investigated the diel habits and spatial use of larvae in each lake before and after fish removal. Ambystoma gracile in Upper Panther were readily observed during day surveys in immediate shoreline, 2 m, and 5 m transects, and were active day and night during the entire period. Prior to fish removal, larvae in Lower Panther were primarily observed during night surveys and restricted to the immediate shoreline habitat. After fish removal, the percent of larvae observed during day surveys increased significantly as did the percent of larvae observed in 2 m and 5 m transects. Changes in Lower Panther indicate that larval behavior becomes less restricted and secretive after fish are removed. This response is indicative of the ecological release of A. gracile larvae from the threat of predation by introduced trout.

INTRODUCTION

Prey attributes that decrease predation risk include unpalatability, reduction or restriction of activity level, and occupation of habitats that may be ephemeral, stressful for predators, or have complex structure that provides refuge from predation (Sih, 1987). Many amphibian species have noxious and/or toxic skin secretions that reduce their palatability to predators (Duellman and Trueb, 1986). These secretions are less developed in larval amphibians (Formanowicz and Brodie, 1982; Duellman and Trueb, 1986), and, thus, larvae rely on behavioral adaptations to escape threats of predation (Sih, 1987; Kats et al., 1988). In the presence of predators, larvae may alter their use of microhabitats (Kiesecker and Blaustein, 1998) and increase their use of structural components of habitat (e.g., coarse woody debris, rock-talus, and aquatic vegetation) as refugia (Kats et al., 1988; Sih et al., 1988; Tyler et al., 1998b). Larvae have also been shown to become more secretive, shift their diel patterns of behavior (Taylor, 1983; Resetarits, 1995), and increase the tendency to display flight behavior (Taylor 1983).

Laboratory studies have shown that larval ambystomatids increase their use of refuge in the presence of fish. In artificial ponds with fish, larval ambystomatids increase their time spent in refuge (Sih et al., 1992; Jackson and Semlitsch, 1993), and restrict their activity to a narrower range of available substrates (Semlitsch, 1987; Stangel and Semlitsch, 1987; Figiel and Semlitsch, 1990; Tyler et al., 1998b). However, little is known about larval salamander refuge use in response to the presence of fish in natural systems. Taylor (1983) observed lower densities and less activity of larvae in lakes with fish in comparison to lakes without fish. While investigating startle responses of two larval Ambystoma gracile (Baird) populations, he observed that higher numbers of larvae were present in the shallow, nearshore areas of a lake with fish than in a fishless lake. Taylor (1983) speculated that the shallow areas of the fish lake may have offered greater refuge from fish predation.

Another behavioral adaptation of ambystomatid larvae is increased nocturnal activity in the presence of fish. Stangel and Semlitsch (1987) noted decreased diurnal activity in larval Ambystoma talpoideum (Holbrook) in artificial ponds with fish. The addition of predatory fish to stream pools decreased the diurnal activity of larval Ambystoma barbouri Kraus and Petranka (Sih et al., 1992). Several field studies have indicated that A. gracile larvae shift toward increased nocturnal activity when fish are present in lakes (Efford and Mathias, 1969; Neish, 1971; Efford and Tsumura, 1973; Sprules, 1974; Taylor, 1983). In view of these results, the objective of this research was to document the diel distributions of A. gracile larvae in a lake before and after fish removal.

METHODS

Upper Panther (the control lake) and Lower Panther (the treatment lake) are separated by about 4 m and are located at an elevation of 1031 m within the boundaries of the North Cascades National Park Service Complex (NOCA), Washington, USA. Upper Panther is smaller (surface area = 0.1 ha; maximum depth = 3 m) than Lower Panther (surface area = 0.3 ha; maximum depth = 5.8 m). Flocculent organic material is the predominant substrate present in the deeper, offshore area of each lake.

Both lakes have shorelines consisting of bedrock, talus, and woody debris, although Lower Panther contains more woody debris than Upper Panther. Lower Panther has additional shoreline areas of overhanging vegetation and undercut banks. Both lakes have a history of cutthroat trout (Oncorhynchus clarki) introductions. Beginning in 1990, removal of fish from the lakes by angling and gillnetting was initiated. Gillnets were set extending from the lake shoreline to the center of the lake. Each gillnet was 42 m long and 2 m deep with four sections of monofilament panels of 12.5 mm, 18.5 mm, 25 mm, and 33 mm mesh. The last trout were removed from Upper Panther in July 1992. In Lower Panther, trout were stocked twice following fish removal efforts. In September 1990, trout were stocked at a density of 750 fry/ha in an attempt to introduce a known number of fish into the lake. In 1994, an unauthorized stocking of trout occurred at an unknown density. Using mark-recapture methods we estimated that there were 320 trout/ha in Lower Panther in 1992 (Gresswell et al., 1997), and 250 trout/ha in 1996 (Torrey Tyler, unpublished data). Fish have not been collected or observed in Lower Panther since seven fish were removed in June 1997.

From 1994 through 1999, densities of A. gracile larvae in both lakes were estimated from snorkel surveys (Tyler et al., 1998a). In Lower Panther, surveys were conducted along four 25 m segments randomly selected along the lake's perimeter. In Upper Panther, the entire lake perimeter (100m) was snorkeled. All surveys were conducted parallel to the shoreline. The same shoreline segments were sampled on all subsequent sampling visits.

Three types of surveys were performed during each sampling visit (Tyler et al., 1998a). Search surveys were conducted near to (i.e., within approximately 1 m) and along the immediate shoreline of each lake. Substrate materials (e.g., talus, woody debris, organic detritus, and aquatic vegetation) present in this nearshore area of each lake created a relatively complex structured habitat potentially useful to larvae as refuge. Larvae were often obscured from the snorkeler's view by these materials. Therefore, during search surveys, snorkelers searched through these substrate materials for larvae and recorded the number of larvae observed. Search surveys were only conducted during mid-afternoon. Two-meter surveys were conducted approximately 2 m from shore where water depth was typically >1 m. During two-meter surveys, snorkelers counted the number of larvae they could observe without disturbing nearshore substrates within an area extending from their 2 m offshore position toward the shore. Five-meter surveys were conducted approximately 5 m from shore where water depth was typically >2 m. During five-meter surveys, snorkerlers counted salamanders without disturbing substrates within an area extending approximately 1.5 m to each side of the snorkeler's longitudinal axis. Two-meter and 5 m surveys were performed once during mid-afternoon, and again 30 min after sunset with the aid of a handheld dive-light. Since 2 m and 5 m surveys were conducted without disturbing substrate materials, these surveys only censused larvae observed in the open and not obscured from the snorkeler's view by substrate materials. The sequence of surveys proceeded from 5 m to 2 m to search. Sampling was conducted systematically so as to avoid recounting larvae potentially flushed from shoreline refugia and moving out to deeper areas of the lake.

Salamander surveys were typically conducted twice a year during the period from mid-June to late August. Both lakes were sampled only once in 1994 due to a forest fire. Equipment failure prevented night surveys during the August 1996 sampling. Nine day surveys and eight night surveys were completed for each lake.

Two-meter and 5 m surveys were combined to investigate differences between day and night larval relative densities in each lake when fish were present in Lower Panther (i.e., 6/94 - 6/97) and after fish were no longer collected or observed in Lower Panther (i.e., 7/97 - 8/99). The combined survey counts were expressed as the percent of all larvae that were observed during the day versus at night. Paired T-tests were performed to test for significant (i.e., p < 0.05) within and between lake differences in the percent of larvae observed during day surveys before and after fish were removed from Lower Panther. In each lake there were four dates before fish removal when both day and night surveys were completed (i.e., n = 4) and four day and night survey dates after fish removal (i.e., n = 4).

To investigate differences in the number of larvae observed in the immediate shoreline area of each lake before and after fish removal, survey counts were expressed as the percentage of total larvae (i.e., search + 2 m + 5 m surveys) observed per day search surveys. The rational for comparing search surveys with combined 2 m and 5 m surveys was as follows: search surveys expressed the number of larvae associated with refuge in each lake, while 2 m and 5 m surveys represented the number of larvae in the open and typically not near refuge. Mann-Whitney tests were used to test for significant within lake differences (i.e., p < 0.05) in the percent of larvae observed during day search surveys before (i.e., 6/94 - 6/97) and after (i.e., 7/97 - 8/99) the removal of fish from Lower Panther. Paired T-tests were used to examine between lake differences. There were five day survey dates prior to fish removal (i.e., n = 5) and four day survey dates after fish removal (i.e., n = 4) for each lake.

NCSS 2000 (Hintze, 1998) was used to calculate Paired T-test and Mann-Whitney test results.

RESULTS

The percent of larvae observed during day versus night surveys in Lower Panther was quite low (i.e., 1% - 6%, average = 4%; Figure 1) when fish were present in the lake (i.e., 6/94 - 6/97; survey_dates = 4); whereas the percent of larvae observed during day surveys after fish were removed (i.e., 7/97 - 8/99; survey_dates = 4) increased significantly to 29% - 48% (average = 41%)(Paired T-test, p = 0.002). In Upper Panther, typically 30% - 50% (average = 40%) of the total number of larvae counted during day and night surveys were observed during the day (Figure 1). Furthermore, no significant difference in the relative percentage of larvae observed in Upper Panther during the day occurred before (survey_dates = 4) versus after (survey_dates = 4) fish were removed from Lower Panther (Paired T-test, p = 0.46). Comparisons between lakes showed that the relative percentage of larvae observed during the day in Lower Panther (survey_dates = 4) was significantly lower than in Upper Panther (survey_dates = 4) prior to fish removal from Lower Panther (Paired T-test, p = 0.002); whereas after fish removal the two lakes did not significantly differ (Paired T-test, p = 0.71; survey_dates = 4 in both lakes) (Figure 1).

Day search surveys revealed within and between lake differences in the proportion of A. gracile larvae observed in the immediate shoreline areas of each lake. There was a significant decrease in the proportion of larvae observed during the day in the immediate shoreline area of Lower Panther after all fish had been removed from the lake (i.e., by June 1997)(Mann-Whitney test, p = 0.01). Prior to fish removal (i.e., 6/94 - 6/97; survey_dates = 5) the percent of larvae observed was above 80% (range = 80% - 99%; average = 87%) and then dropped below 50% (range = 14% - 49%; average = 30%) after fish were removed (i.e., 7/97 - 8/99; survey_dates = 4)(Figure 2). In fishless Upper Panther, the percent of larvae observed in day search surveys ranged from 16% - 52% (average = 33%), and was typically < 40% of all larvae observed (i.e., search + 2 m + 5 m surveys)(Figure 2). The percent of larvae observed in Upper Panther search surveys conducted prior to when fish were removed from Lower Panther (range = 16% - 44%, average = 28%; survey_dates = 5) did not differ significantly from the percentage observed after fish were removed from Lower Panther (range = 26% - 52%, average = 39%; survey_dates = 4; Mann-Whitney test, p = 0.11). Between lake comparisons showed that the percentage of larvae observed in the immediate shoreline area of each lake differed before but not after fish removal from Lower Panther. Prior to fish removal a significantly greater mean percentage of larvae were observed in Lower Panther search surveys than in Upper Panther search surveys (Paired T-test, p = 0.002; survey_dates = 5 for each lake) (Figure 2). After fish removal the average percentage of larvae observed in search surveys did not significantly differ between lakes (Paired T-test, p = 0.12; survey_dates = 4 for each lake) (Figure 2).

DISCUSSION

This study provides experimental evidence based on whole-lake manipulations that indicates that A. gracile larval behavior differs between lakes with and without fish. This study also documents the response of A. gracile larvae to the removal of fish from a lake. Prior to the removal of all trout from Lower Panther, A. gracile larvae tended to be primarily active at night. Larvae were mostly restricted to the shallow, shoreline area of the lake that contained substrates that created complex habitat structure. This structure provided refuge for larvae from predation by trout. During this same period, salamander larvae in fishless Upper Panther were active day and night and were observed throughout the lake. In this context, the secretive and restricted behavior of A. gracile larvae in Lower Panther can be seen as indicative of an anti-predator response to the presence of fish in the lake.

Previous studies have observed that A. gracile larvae in response to the threat of predation by trout may decrease their diurnal activity and restrict the number and types of habitats they occupy in a lake. Efford and Mathias (1969) attributed the secretive and wary behavior of A. gracile larvae in Marion Lake, British Columbia, Canada, to predation by trout. In ponds and lakes containing fish in British Columbia and Oregon, A. gracile larvae are almost exclusively nocturnal (Efford and Mathias 1969; Sprules 1974; Taylor, 1983). Liss et al. (1995) found that although A. gracile larvae could be detected during day surveys in some NOCA lakes with fish, most of these larvae were observed in refuges (e.g., woody debris, rock-talus, undercut banks). At one NOCA lake containing fish, counts of larvae at night were ten times greater than day counts. Taylor (1983) reported that in three lakes with fish in the Oregon Cascade Mountains, A. gracile larvae were detected only in nearshore refugia where the water was <1.3 m deep. Tyler et al. (1998b) determined that A. gracile larvae in experimental ponds with fish utilized a narrower range of available habitats (e.g., rock and wood) than did larvae in fishless controls. In contrast to the nocturnal and secretive behavior of larvae in lakes with fish, A. gracile larvae in fishless lakes tend to be readily detected during the day, more active throughout a lake, and do not appear overly wary or secretive in behavior (Sprules 1974; Taylor 1983).

After all of the fish were removed from Lower Panther by June 1997, the behavior of A. gracile larvae changed and became strikingly similar to the larval behavior in fishless Upper Panther (Figs. 1 and 2). In essence, upon the removal of fish from Lower Panther the proportion of larvae observed in the lake during day surveys increased as did the proportion of larvae observed in the transects (i.e., 2 m and 5 m) further from the shoreline of the lake. To our knowledge, the present study is the first to indicate that A. gracile larval behavior can shift following the removal of fish from a lake. The change in the level of diurnal activity and location of larvae in Lower Panther might be explained as a form of ecological release, and more specifically as a release from predation. In general, ecological release is expressed when prey exhibit density compensation and/or habitat expansion upon reduction of the level of interspecific competition or predation within an ecosystem (MacArthur et al., 1972; Cox and Ricklefs, 1977; Ricklefs, 1979). The suppressed species can become more abundant, more active, and less restricted in the habitats it is able to occupy or exploit. For instance, sea urchin sizes and densities were higher on Kenyan reefs where the exploitation of predatory fish was high as compared to reefs where exploitation was low (McClanahan and Muthiga 1988). Furthermore, species richness and total abundance of young-of-the-year non-piscivorus fishes were greater on predator removal coral reefs than on control reefs (Caley 1993). In a study examining ecological release in amphibians, the relative abundance of green frogs (Rana clamitans Latreille) in Point Pelee National Park, Ontario, Canada, increased four-fold after the extirpation of the green frogs' potential competitor and predator, the bullfrog (Rana catesbeiana Shaw; Hecnar and M'Closkey 1997).

The shift in A. gracile larval behavior in Lower Panther occurred relatively soon after the last fish were removed from the lake. Ambystoma gracile larvae can co-exist with introduced trout in mountain lakes albeit at lower densities than in lakes without trout (Liss et al., 1995; Robert Hoffman and Torrey Tyler, personal observations). They survive in lakes with trout by becoming secretive and wary, and by restricting their diurnal activity to lake habitats that provide refuge from predation. Being able to coexist with introduced trout enhances this species' ability to reestablish potentially threatened populations upon the removal or extinction of fish from a lake. This outcome is certainly encouraging as resource managers attempt to deal with issues related to declining amphibian populations in the western United States. Future research should include additional amphibian species, especially those that may be less able to coexist with introduced fish (e.g., Ambystoma macrodactylum Baird; see Tyler et al., 1998a), so that we may better understand the full potential of amphibian population recovery associated with the removal of introduced fish from mountain lakes.

ACKNOWLEDGMENTS

This research was funded by the United States Geological Survey, Biological Resources Division. The authors would like to thank Mike Collopy and the USGS Forest and Rangeland Ecosystem Science Center for support of this project; Bruce Freet, Reed Glesne, and park personnel of North Cascades National Park Service Complex; and Elisabeth Deimling and Gregg Lomnicky for their assistance in completing the field work associated with this study.

REFERENCES

Caley, M.J. 1993. Predation, recruitment and the dynamics of communities of coral-reef fishes. Marine Biology 117:33-43.

Cox, G.W. and R.E. Ricklefs. 1977. Species diversity and ecological release in Caribbean land bird faunas. Oikos 28:113-122.

Duellman, W.E., and L. Trueb. 1986. Biology of Amphibians. The John Hopkins University Press, Baltimore, Maryland, U.S.A. 670 p.

Efford, I.E., and J.A. Mathias. 1969. A comparison of two salamander populations in Marion Lake, British Columbia. Copeia 1969:723-735.

Efford, I.E., and K. Tsumura. 1973. A comparison of the food of salamanders and fish in Marion Lake, British Columbia. Transactions of the American Fisheries Society 102:33-47.

Figiel, C.R., and R.D. Semlitsch. 1990. Population variation in survival and metamorphosis of larval salamanders (Ambystoma maculatum) in the presence and absence of fish predation. Copeia 1990:818-826.

Formanowicz Jr., D. R. and E. D. Brodie, Jr. 1982. Relative palatabilities of members of a larval amphibian community. Copeia 1982:91-97.

Gresswell, R.E., W.J. Liss, G.A. Lomnicky, E.K. Deimling, R.L. Hoffman, and T. Tyler. 1997. Using mark-recapture methods to estimate fish abundance in small mountain lakes. Northwest Science 71:39-44.

Hecnar, S.J. and T. M'Closkey. 1997. Changes in the composition of a ranid frog community following bullfrog extinction. American Midland Naturalist 137:145-150. Hintze, J.L. 1998. NCSS 2000. Kaysville, Utah. Jackson, M.A., and R.D. Semlitsch. 1993. Paedomorphosis in the salamander Ambystoma talpoideum: effects of a fish predator. Ecology 74:342-350.

Kats, L.B., J.W. Petranka, and A. Sih. 1988. Antipredator defenses and the persistence of amphibian larvae with fishes. Ecology 69:1865-1870.

Kiesecker, J.M., and A.R. Blaustein. 1998. Effects of introduced bullfrogs and smallmouth bass on microhabitat use, growth, and survival of native red-legged frogs (Rana aurora). Conservation Biology 12: 776-787.

Liss, W.J., G.L. Larson, E.K. Deimling, R. Gresswell, R. Hoffman, M. Kiss, G. Lomnicky, C.D. McIntire, R. Truitt, and T. Tyler. 1995. Ecological effects of stocked trout in naturally fishless high mountain lakes: North Cascades National Park Service Complex, WA, USA. Technical report NPS/PNROSU/NRTR-95-03. National Park Service, Pacific Northwest Region, 909 First Avenue, Seattle, WA 98104.

MacArthur, R.H., J.M. Diamond, and J.R. Karr. 1972. Density compensation in island faunas. Ecology 53:330-342.

McClanahan, T.R. and N.A. Muthiga. 1988. Changes in Kenyan coral reef community structure and function due to exploitation. Hydrobiologia 166:269-276.

Neish, I.C. 1971. Comparisons of size, structure, and distributional patterns of two salamander populations in Marion Lake, British Columbia. Journal of the Fisheries Research Board of Canada 28:49-58.

Nussbaum, R.A., E.D. Brodie, and R.M. Storm. 1983. Amphibians and Reptiles of the Pacific Northwest. The University Press of Idaho, Moscow, Idaho. 332 p.

Resetarits, Jr., W.J. 1995. Competitive asymmetry and coexistence in size-structured populations of brook trout and spring salamanders. Oikos 73:188-198. Ricklefs, R.E. 1979. Ecology, second edition. Chiron Press, Inc., New York. 966 p. Semlitsch, R.D. 1987. Interactions between fish and salamander larvae: costs of predator avoidance or competition? Oecologia 72:481-486.

Sih, A. 1987. Predators and prey lifestyles: an evolutionary and ecological overview. In Kerfoot, W.C. and A. Sih (eds.), Predation: Direct and Indirect Impacts on Aquatic Communities, p. 203-224, University Press of New England, Hanover, New Hampshire, U.S.A. 386 p.

Sih, A., J.W. Petranka, and L.B. Kats. 1988. The dynamics of prey refuge use: a model and tests with sunfish and salamander larvae. American Naturalist 132:463-483.

Sih, A., L.B. Kats, and R.D. Moore. 1992. Effects of predatory sunfish on the density, drift, and refuge use of stream salamander larvae. Ecology 73:1418-1430.

Sprules, W.G. 1974. The adaptive significance of paedogenesis in North American species of Ambystoma (Amphibia: Caudata): an hypothesis. Canadian Journal of Zoology 52:393-400.

Stangel, P.W., and R.D. Semlitsch. 1987. Experimental analysis of predation on the diel vertical migrations of a larval salamander. Canadian Journal of Zoology 65:1554-1558.

Taylor, J. 1983. Orientation and flight behavior of a neotenic salamander (Ambystoma gracile) in Oregon. American Midland Naturalist 109:40-49.

Tyler, T., W.J. Liss, L.M. Ganio, G.L. Larson, R. Hoffman, E. Deimling, and G. Lomnicky. 1998a. Interaction between introduced trout and larval salamanders (Ambystoma macrodactylum) in high-elevation lakes. Conservation Biology 12:94-105.

Tyler, T.J., W.J. Liss, R. L. Hoffman, and L.M. Ganio. 1998b. Experimental analysis of trout effects on survival, growth, and habitat use of two species of Ambystomatid salamanders. Journal of Herpetology 32:345-349.

Figure 1. Percent of Ambystoma gracile observed during day (versus night) 2 m and 5 m surveys in Lower Panther and Upper Panther Lakes. No fish were observed in Lower Panther lake after June 1997. FR indicates when Lower Panther Lake was considered fishless.

Figure 2. Percent of Ambystoma gracile larvae observed in the immediate shoreline areas of Lower Panther and Upper Panther Lakes during day search versus day 2 m + 5 m surveys. No fish were observed in Lower Panther Lake after June 1997. FR indicates when Lower Panther Lake was considered fishless.

Chapter 1 | 2 | 3 | Table of Contents