1759

Colonization of high-elevation lakes by long-toed salamanders (Ambystoma macrodactylum) after the extinction of introduced trout populations
W. Chris Funk and William W. Dunlap

Abstract: We surveyed high-elevation lakes for long-toed salamander (Ambystoma macrodactylum) larvae and trout in the northern Bitterroot Mountains of Montana, U.S.A., in 1978, 1997, and 1998. Our objectives were to (i) test whether trout exclude salamander populations; (ii) determine whether lakes in which trout have gone extinct have since been colonized by salamanders; and (iii) estimate the rates of population extinction and colonization in lakes never stocked with trout. In agreement with previous work on the interactions between trout and long-toed salamanders, trout effectively excluded salamander populations from lakes. Somewhat surprisingly, however, salamanders managed to colonize lakes after the extinction of trout populations despite evidence of low levels of interpopulation dispersal in these salamander populations. In lakes never stocked with trout there was no evidence of a decline in salamander populations; 2 of these lakes were colonized and no populations went extinct. Rsum : Nous avons inventori des lacs de haute altitude afin dvaluer labondance des larves de la Salamandre longs doigts (Ambystoma macrodactylum) et des truites dans la chane nordique Bitterroot Mountains du Montana, .-U., en 1978, 1997 et 1998. Nous avions pour objectif (i) dtablir si la prsence de truites exclut les populations de salamandres, (ii) de dterminer si les lacs do les truites sont disparues ont t recoloniss par les salamandres, et (iii) destimer les taux dextinction et de colonisation des populations dans les lacs qui nont jamais contenu de truites. Tel que prvu la lumire des travaux antrieurs sur les interactions truites-salamandres, les truites ont de fait limin les populations de salamandres des lacs. tonnamment, cependant, les salamandres ont russi coloniser les lacs aprs lextinction des populations de truites en dpit du faible taux apparent de dispersion inter-populations chez ces populations de salamandres. Dans les lacs qui nont jamais contenu de truites, les populations de salamandres son restes stables, 2 des lacs ont t coloniss et aucune population nest disparue. [Traduit par la Rdaction] 1767 Funk and Dunlap

Introduction

In the last decade, evidence has accumulated suggesting a worldwide decline in amphibian species above and beyond declines being experienced by all taxa (Blaustein and Wake 1990; Wyman 1990; Wake 1991; Wake and Morowitz 1991; Blaustein 1994; Phillips 1994; Sarkar 1996). Some species have completely vanished, many show signs of reduced population sizes or overall numbers, while others appear stable (Pounds 1991; Crump et al. 1992; Pounds and Crump 1994; Weller and Green 1997; Meyer et al. 1998; Williams and Hero 1998). The decline in amphibian populations is of particular concern in the American and Canadian West, where the population sizes and ranges of a relatively high proportion of species have decreased in the latter half of this century (Blaustein and Wake 1990; Weller and Green 1997).

Received November 30, 1998. Accepted July 20, 1999. W.C. Funk.1 Division of Biological Sciences, University of Montana, Missoula, MT 59812, U.S.A. W.W. Dunlap. MathSoft, Inc., Mount Vernon, WA 98274, U.S.A.
1

Author to whom all correspondence should be addressed (e-mail: wcfunk@selway.umt.edu).

A number of factors have been explored as potential causes of amphibian population declines. Among the most obvious, and likely most important, causes are site-specific factors such as habitat destruction, habitat fragmentation, and the introduction of predator and competitor species (Wyman 1990; Green 1997; Kiesecker and Blaustein 1998; Tyler et al. 1998b). Global factors may also be involved. For example, increased UV-B radiation has been shown to decrease survivorship during the early stages of development in some amphibian species (Blaustein et al. 1994; Blaustein et al. 1997; Anzalone et al. 1998; Lizana and Pedraza 1998). This could result in decreased population growth rates and an increased probability of population extinction. However, the degree to which extinctions of individual amphibian populations will result in the overall decline of species depends on a critical parameter: the rate of colonization of empty habitat patches. Metapopulation theory predicts that if population extinctions are balanced by colonizations, then the set of populations (i.e., the metapopulation) in question will persist (Levins 1970; Hanski and Gilpin 1991). However, if the rate of population extinction exceeds the rate of colonization, the metapopulation is predicted to decline. Although previous authors have discussed the importance of considering dispersal and colonization for understanding the species-wide affects of amphibian population extinctions
1999 NRC Canada

Can. J. Zool. 77: 17591767 (1999)

1760

Can. J. Zool. Vol. 77, 1999 whether or not salamander larvae were actually absent from all of these lakes. In contrast, in 1997 and 1998 we surveyed the entire perimeters of all but the 2 largest (Bass Lake and Big Creek Lakes) of the 42 lakes from shore for both salamander larvae and trout as described above. Approximately one-quarter of the perimeters of Bass Lake and Big Creek Lakes were surveyed. In addition, if no salamander larvae and (or) trout were detected in a given lake by searching from shore, snorkel surveys were employed to confirm their purported absence. Snorkel surveys for salamanders consisted of slowly swimming three different 25-m transects running parallel to and approximately 2 m from the shoreline. In these surveys, both the water column and substrate within approximately 2 m either side of the transect line were carefully checked for salamander larvae (Tyler et al. 1998a). Snorkel surveys for fish consisted of slowly swimming three different randomly chosen 50-m transects running parallel to and approximately 3 m from the shoreline. In these surveys, the water column within approximately 3 m either side of the transect line was carefully searched for trout. In no case did snorkel surveys give a different result from surveys conducted from shore, suggesting that shore surveys were sufficient for determining the presence or absence of salamanders and trout.

(Gill 1978; Sjgren 1991; Sjgren Gulve 1994; Hecnar and MCloskey 1996; Driscoll 1997; Vos and Chardon 1998), there is little empirical evidence concerning the capacity of amphibians to colonize empty habitat patches (Sjgren Gulve 1994; Hecnar and MCloskey 1996). Genetic differentiation among populations tends to be high in amphibians relative to other taxa, suggesting that gene flow, and hence dispersal, is limited in amphibians (Larson et al. 1984; Slatkin 1985; Gascon et al. 1996; Gascon et al. 1998). The results of studies using direct methods to measure amphibian dispersal generally support the same conclusion: amphibians are highly philopatric (Daugherty and Sheldon 1982; Crump 1986; Driscoll 1997). These results suggest that amphibian colonization rates may be low as well. In view of these low dispersal rates, the critical question, therefore, is whether dispersal is great enough to allow the colonization of empty habitat patches at a rate equal to or greater than the rate of population extinction. We address this question for a set of long-toed salamander (Ambystoma macrodactylum) populations in high-elevation lakes in the northern Bitterroot Mountains of Montana, U.S.A. Many of these salamander populations were extirpated by the introduction of trout to these historically fishless lakes earlier this century. However, with the cessation of trout stocking in 1984, some of these trout populations have gone extinct, allowing for possible colonization by salamanders. We surveyed these lakes in 1978, 1997, and 1998 to (i) test whether trout exclude salamanders; (ii) determine whether lakes in which trout have gone extinct have since been colonized by salamanders; and (iii) estimate the rates of colonization and population extinction in lakes never stocked with trout.

Data analysis
Whether or not trout exclude salamander larvae from lakes was tested by comparing the proportion of lakes without trout that currently contain salamander larvae with the proportion of lakes with trout that currently contain salamander larvae, using Fishers exact test. This test generates an exact probability that two variables are statistically independent. All Fishers exact tests were performed using Zaykin and Pudovkins (1993) program CHIRXC. Since the 18 lakes containing trout in 1978 (Table 1) were not thoroughly searched for salamander larvae during the 1978 surveys, salamander larvae found in these lakes during the 1997 and 1998 surveys cannot unequivocally be said to be the descendants of colonists. They could also be the descendants of individuals that existed sympatrically with the introduced trout populations. Thus, to test whether or not lakes in which trout populations have gone extinct have been colonized by salamanders, we used a statistical approach. We first assumed that the proportion of the 12 lakes with trout in the 1997 and 1998 surveys (Table 1) that also contain salamander larvae is representative of the proportion of the 18 lakes with trout in the 1978 surveys that also contained salamander larvae. We then tested whether the proportion of the 6 lakes in which trout went extinct (Table 1) that currently contain salamander larvae is greater than that which we would expect if the salamander larvae found in these lakes were simply the descendants of individuals that had managed to coexist with trout. This was done using Fishers exact test to test whether the proportion of the 6 lakes in which trout have gone extinct that currently contain salamanders is significantly greater than the proportion of the 12 lakes in which trout have persisted that contain salamanders. A second assumption of this approach is that the lakes in which trout went extinct are a random subsample of the lakes that contained trout in 1978 with respect to the capacity of each lake to support the coexistence of salamander larvae and trout. Two factors that affect the suitability of high-elevation lakes for both salamander larvae and trout, and, therefore, which likely affect a given lakes capacity to support the coexistence of larvae and trout, are water depth and elevation (Bahls 1992; Tyler et al. 1998a; Vos and Chardon 1998). To check this assumption, we looked for systematic differences in surface area (as a surrogate for water depth) and elevation between lakes in which trout went extinct and lakes in which they persisted, by plotting log (surface area) and elevation for each lake (Fig. 5). We then used Fishers exact test to test whether the proportion of lakes in which trout went extinct differed between the 9 lakes with the smallest log (surface area) and the 9
1999 NRC Canada

Materials and methods

Study area
All lakes surveyed are located in the northern Bitterroot Mountains (Fig. 1). Surveyed lakes range in size from approximately 0.0020 to 1.0000 km2 and in elevation from approximately 1790 to 2470 m (estimated from 1 : 24 000 scale U.S. Geological Survey maps) (Table 1). The townshiprange format for locations is as used in all U.S. geological survey and U.S. Forest Service maps. Because of impassable waterfalls, these lakes were historically fishless but served as habitat for breeding populations of long-toed salamanders and spotted frogs (Rana luteiventris). However, many of the lakes were stocked with various species of trout (Oncorhynchus clarki, Oncorhynchus mykiss, Salvelinus fontinalis) earlier this century. Stocking of these lakes by the Montana Department of Fish, Wildlife, and Parks continued until 1984 (C. Clancy, personal communication).

Surveys
We surveyed 42 lakes in 1978, 1997, and 1998 for long-toed salamander larvae and trout. Survey procedures differed between the earlier surveys (1978) and later surveys (1997 and 1998). In 1978, surveys involved carefully scanning the water column, water surface, and bottom of each lake from shore for salamander larvae and trout as described by Thoms et al. (1997). At each lake, surveys continued until either the entire perimeter of the lake had been searched or a trout had been seen. Therefore, in 1978 all lakes were thoroughly surveyed for trout, but only those lakes without trout were thoroughly surveyed for salamander larvae. No salamander larvae were ever seen in lakes containing trout, but because the entire perimeters of these lakes were not searched, it is not known

Funk and Dunlap

1761

Fig. 1. Map of the 42 lakes surveyed for long-toed salamander (Ambystoma macrodactylum) larvae and trout in the northern Bitterroot Mountains of Montana, U.S.A., in 1978, 1997, and 1998. Symbols indicate whether or not lakes were stocked and the presence or absence of salamander larvae and trout in the 1978 and 19971998 survey sessions as specified.

lakes with the largest log (surface area) or between the 9 lakes at the lowest elevation and the 9 lakes at the highest elevation. Change in the proportion of lakes that have never been stocked with trout that contained salamander larvae between the 1978 surveys and the 1997 and 1998 surveys was tested using Fishers exact test.

Results
Of the 42 lakes surveyed in 1997 and 1998, 12 (28.6%) contained trout and 29 (69.0%) contained salamander larvae (Table 1, Figs. 1 and 2). Only 2 (16.7%) of the 12 lakes with

trout also contained larvae, whereas 27 (90.0%) of the 30 lakes without trout contained larvae (Table 1, Figs. 1 and 2). The proportion of lakes with trout that also contained larvae is significantly different from the proportion of lakes without trout that contained larvae (p < 0.0001; Fig. 2). This suggests that trout excluded larvae from lakes. By 1998, trout populations had gone extinct in 6 (33.3%) of the 18 lakes that contained trout in 1978 (Table 1, Figs. 1 and 3). Salamander larvae were present in 5 (83.3%) of the 6 lakes in which trout went extinct, whereas only 2 (16.7%) of the 12 lakes in which trout persisted contained larvae (Ta 1999 NRC Canada

1762

Can. J. Zool. Vol. 77, 1999 Table 1. Incidence of long-toed salamander (Ambystoma macrodactylum) larvae and trout populations in lakes in the northern Bitterroot Mountains of Montana, U.S.A., in 1978 and 1997 1998.

Drainage Mill Creek Carlton Creek One Horse Creek Sweeney Creek

Lake designation and name WOO CAR, Carlton L. LCA, Little Carlton L. MOH, Reed L. SON, South One Horse L. MIL, Mills L. HOL, Holloway L. DLK, Duffy L. DUF NDF PET, Peterson L. SWE, Sweeney L. ESW ASC BAS, Bass L. WKO, West Middle Fork L. EKO, East Middle Fork L. NKO, North Kootenai L. SKO, South Kootenai L. UDS LDS SH2 SH3 BOB RUT CLA SMC DMC BCL, Big Creek Lakes SFL, South Fork L. WSF ASF PRL, Pearl L. UPR STG HID, Hidden L. UHD BHD TCR GLE, Glen L. BEA, Bear L. BRY, Bryan L.

Location T11N R21W S22 SW1/4 T11N R21W S27 NE1/4 T11N R21W S27 SE1/4 T11N R21W S33 NE1/4 T11N R21W S33 SE1/4 T10N R21W S8 SE1/4 T10N R21W S8 SE1/4 T10N R21W S8/9 T10N R21W S9 NW1/4 T10N R21W S4 SW1/4 T10N R21W S10 SW1/4 T10N R21W S20 NW1/4 T10N R21W S20 NW1/4 T10N R21W S30 NW1/4 T10N R21W S30 SW1/4 T9N R22W S11 SE1/4 T9N R22W S11 SE1/4 T9N R22W S11 NE1/4 T9N R22W S13/14 T9N R21W S27 NW1/4 T9N R21W S27 NW1/4 T9N R21W S22 SE1/4 T9N R21W S27 NE1/4 T9N R21W S22 SE1/4 T9N R21W S27 NW1/4 T9N R21W S27 SW1/4 T9N R21W S28 SE1/4 T9N R21W S33 NW1/4 T9N R22W S33/34/ T8N R22W S5 T8N R22W S17 SE1/4 T8N R22W S16 SW1/4 T8N R22W S16 SW1/4 T8N R22W S8/17 T8N R22W S17 NW1/4 T8N R22W S16 NW1/4 T8N R22W S11 NW1/4 T8N R22W S11 SW1/4 T8N R22W S11 NE1/4 T8N R22W S13 NW1/4 T8N R22W S13 NW1/4 T8N R22W S21 SE1/4 T8N R22W S29 SW1/4

South Fork, Lolo Creek Bass Creek Kootenai Creek

Sharrott Creek

McCalla Creek Big Creek

Sweathouse Creek Bear Creek

Note: A question mark indicates a lake and year in which the incidence of A. macrodactylum was not determined with confidence (see Materials and methods).

ble 1, Figs. 1 and 3). The proportion of lakes containing larvae in which trout went extinct was significantly different from the proportion of lakes containing larvae in which trout

persisted (p = 0.0128; Fig. 3). This suggests at least some of the lakes in which trout went extinct were colonized by salamanders.
1999 NRC Canada

Funk and Dunlap Table 1 (concluded).

1763

Incidence 1978 Elevation (m) 2270 2370 2350 2410 2260 2380 2380 2230 2230 2470 1980 2240 2230 2160 2060 1820 1800 1910 1910 2440 2320 2210 2320 2100 2370 2370 2430 2440 1790 2070 2320 2220 2160 2200 2210 2180 2240 2160 2390 2300 2000 2060 Surface area (km2) 0.0030 0.1250 0.0190 0.0160 0.0290 0.0220 0.0700 0.0630 0.0050 0.0130 0.0730 0.0540 0.0120 0.0230 0.3260 0.0750 0.0200 0.0550 0.0960 0.0050 0.0050 0.0050 0.0030 0.0030 0.0030 0.0050 0.0100 0.0200 1.0000 0.0830 0.0040 0.0020 0.0620 0.0050 0.0080 0.0340 0.0060 0.0020 0.0200 0.0240 0.0560 0.0470 Salamander + ? ? + ? ? ? ? + + ? + + ? ? ? ? ? + + + + + + + + ? ? ? + + ? + + + + ? ? Trout + + + + + + + + + + + + + + + + + + 1997/1998 Salamander + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Trout + + + + + + + + + + + +

Twenty-four (57.1%) of the 42 lakes surveyed were never stocked with trout (Table 1, Figs. 1 and 4). In 1978, 20 (83.3%) of these 24 lakes contained salamander larvae, and

in 1997 and 1998, 22 (91.7%) contained larvae (Table 1, Figs. 1 and 4). This difference in the proportion of lakes with larvae is not significant (p = 0.6662; Fig. 4).
1999 NRC Canada

1764

Can. J. Zool. Vol. 77, 1999

Fig. 2. Frequency histogram of the proportions of the 42 lakes surveyed in 1997 and 1998 with and without long-toed salamander larvae and trout. The proportion of lakes with trout that also contained salamander larvae was significantly different from the proportion of lakes without trout that contained larvae (p < 0.0001).

Fig. 3. Frequency histogram showing the proportion of the 12 lakes in which trout persisted that contained long-toed salamander larvae and the proportion of the 6 lakes in which trout went extinct that contained larvae in the 19971998 survey session. These two proportions were significantly different (p = 0.0128).

There was no systematic difference between lakes in which trout went extinct and those in which trout persisted with respect to lake surface area or elevation (Table 1, Fig. 5). The proportions of lakes in which trout went extinct did not differ between the 9 lakes with the smallest log (surface area) and the 9 lakes with the largest log (surface area) (p = 0.6199), nor between the 9 lakes at the lowest elevation and the 9 lakes at the highest elevation (p = 1.0000).

Discussion
As expected, trout excluded long-toed salamander larvae from lakes in the northern Bitterroot Mountains. This result is in agreement with previous work in the North Cascade

Range in Washington, U.S.A., showing that trout reduced the abundance of long-toed salamander larvae (Tyler et al. 1998a). The likely mechanism by which trout eliminate salamander larvae or reduce their abundance is predation. Recent experimental evidence supports this hypothesis (Tyler et al. 1998b). What is more surprising is that long-toed salamanders colonized lakes relatively quickly (within the 20-year period of this study) despite recent genetic evidence that levels of dispersal between these populations are very low (Tallmon et al. 1999). This is in agreement with two other studies in which colonization events were observed in amphibians (Sjgren Gulve 1994; Hecnar and MCloskey 1996), although the colonists observed by Sjgren Gulve were not re 1999 NRC Canada

Funk and Dunlap

1765

Fig. 4. Frequency histogram of the proportions of lakes never stocked with trout that contained long-toed salamander larvae in the 1978 and 19971998 survey sessions. The two proportions were not significantly different (p = 0.6662).

Fig. 5. Bivariate plot of log (surface area) and elevation for all of the 18 lakes stocked with trout. The proportion of lakes in which trout went extinct did not differ between the 9 lakes with the smallest log (surface area) and the 9 lakes with the largest log (surface area) (p = 0.6199), nor between the 9 lakes at lowest elevation and 9 lakes at highest elevation (p = 1.0000).

productive and Hecnar and MCloskey did not specify whether colonists were reproductive or not. Thus, it is not known whether these colonizations represent successful colonization events resulting in the founding of new populations or simply temporary residency of nonreproductive individuals. We could not find any studies that explicitly reported the failure of colonization in amphibians, but some studies of amphibian dispersal hypothesize that colonization is unlikely, owing to low levels of dispersal (Driscoll 1997). In those lakes that were never stocked with trout, longtoed salamander populations showed no signs of decline. None of these populations went extinct; in fact, 2 lakes were colonized. Thus, if anything, these populations appear to be increasing in number. According to recent evidence, long-

toed salamander populations may be at particular risk of declining, since embyronic development in this species is negatively affected by UV-B radiation (Blaustein et al. 1997). Since UV-B radiance associated with stratospheric ozone depletion is increasing at temperate latitudes (Blumthaler and Ambach 1990), the negative effect of UV-B could become manifest at the population level. Our evidence suggests that this is not happening in these long-toed salamander populations, although population sizes, which we did not monitor, rather than overall numbers could be declining. Conclusions Our results show that the populations of long-toed salamanders surveyed have the capacity to colonize lakes despite
1999 NRC Canada

1766

Can. J. Zool. Vol. 77, 1999 Blaustein, A.R., Hoffman, P.D., Hokit, D.G., Kiesecker, J.M., Walls, S.C., and Hays, J.B. 1994. UV repair and resistence to solar UV-B in amphibian eggs: a link to population declines? Proc. Natl. Acad. Sci. U.S.A. 91: 17911795. Blaustein, A.R., Kiesecker, J.M., Chivers, D.P., and Anthony, R.G. 1997. Ambient UV-B radiation causes deformities in amphibian embryos. Proc. Natl. Acad. Sci. U.S.A. 94: 13 735 13 737. Blumthaler, M., and Ambach, W. 1990. Indication of increasing solar ultraviolet-B radiation flux in alpine regions. Science (Washington, D.C.), 248: 206208. Crump, M.L. 1986. Homing and site-fidelity in a neotropical frog, Atelopus varius (Bufonidae). Copeia, 1986: 438444. Crump, M.L., Hensley, F.R., and Clark, K.L. 1992. Apparent decline of the golden toad: underground or extinct? Copeia, 1992: 413420. Daugherty, C.H., and Sheldon, A.L. 1982. Age-specific movement patterns of the frog Ascaphus truei. Herpetologica, 38: 468474. Driscoll, D.A. 1997. Mobility and metapopulation structure of Geocrinia alba and Geocrinia vitellina, two endangered frog species from southwestern Australia. Aust. J. Ecol. 22: 185195. Gascon, C., Lougheed, S.C., and Bogart, J.P. 1996. Genetic and morphological variation in Vanzolinius discodactylus: a test of the river hypothesis of speciation. Biotropica, 28: 376387. Gascon, C., Lougheed, S.C., and Bogart, J.P. 1998. Patterns of genetic population differentiation in four species of Amazonian frogs: a test of the riverine barrier hypothesis. Biotropica, 30: 104119. Gill, D.E. 1978. Effective population size and interdemic migration rates in a metapopulation of the red-spotted newt, Notophthalmus viridescens. Evolution, 32: 839849. Gilpin, M.E., and Soul, M.E. 1986. Minimum viable populations: the processes of species extinctions. In Conservation biology: the science of scarcity and diversity. Edited by M. Soul. Sinauer Associates, Sunderland, Mass. pp. 1334. Green, D.M. 1997. Perspectives on amphibian population declines: defining the problem and searching for answers. In Amphibians in decline: Canadian studies of a global problem. Edited by D. Green. Society for the Study of Amphibians and Reptiles, St. Louis, Mo. pp. 309328. Hanski, I., and Gilpin, M. 1991. Metapopulation dynamics: a brief history and conceptual domain. Biol. J. Linn. Soc. 42: 316. Hecnar, S.J., and MCloskey, R.T. 1996. Regional dynamics and the status of amphibians. Ecology, 77: 20912097. Hitchings, S.P., and Beebee, T.J. 1997. Genetic substructuring as a result of barriers to gene flow in urban Rana temporaria (common frog) populations: implications for biodiversity conservation. Heredity, 79: 117127. Kiesecker, J.M., and Blaustein, A.R. 1998. Effects of introduced bullfrogs and smallmouth bass on microhabitat use, growth, and survival of native red-legged frogs (Rana aurora). Conserv. Biol. 12: 776787. Larson, A., Wake D.B., and Yanev, K.P. 1984. Measuring gene flow among populations having high levels of genetic fragmentation. Genetics, 106: 293308. Levins, R. 1970. Extinction. In Some mathematical questions in biology. Vol. 2. Edited by M. Gerstenhaber. American Mathematical Society, Providence, R.I. pp. 77107. Lizana, M., and Pedraza, E.M. 1998. The effects of UV-B radiation on toad mortality in mountainous areas of central Spain. Conserv. Biol. 12: 703707. Meyer, A.H., Schmidt B.R., and Grossenbacher, K. 1998. Analysis of three amphibian populations with quarter-century long timeseries. Proc. R. Soc. Lond. B Biol. Sci. 265: 523538.
1999 NRC Canada

evidence of low levels of interpopulation dispersal. From a management perspective, this suggests that these salamander populations have the potential to recover from the introduction of trout if trout populations go extinct or are removed. Even then, however, there may be long-term consequences for salamander populations. For example, salamander populations founded by colonists may have higher probabilities of extinction than other populations, owing to demographic stochasticity and negative inbreeding effects associated with small population size during colonization events (Allendorf and Leary 1986; Gilpin and Soul 1986; Mills and Smouse 1994; Hitchings and Beebee 1997; Newman and Pilson 1997; Saccheri et al. 1998). Now that it is known that amphibians can colonize empty habitat patches relatively quickly, future research should be directed towards understanding what factors affect rates of colonization. For example, one question of interest is how the rate of colonization is related to the distance from a source population. A related question concerns where the colonists come from: do they come from the closest occupied habitat patches or from source populations farther away? Answers to these questions will be useful for insuring the persistence of sets of amphibian populations.

Acknowledgements
Particular thanks are extended to Andrew Sheldon for providing advice during the early stages of this project and for thoroughly reviewing the manuscript. We thank Chris Clancy from Montana Fish, Wildlife, and Parks and the Bitterroot National Forest for providing information on the stocking history of the Selway Bitterroot Wilderness lakes and Judith Fraser from the Bitterroot National Forest for facilitating fieldwork in the Selway Bitterroot Wilderness. We are also indebted to Brett Walker for assisting with snorkel surveys and to Kathryn Warner, Christina Von der Ohe, and Chester for helping with other aspects of fieldwork. W. Chris Funk was supported by an National Science Foundation (NSF) Graduate Research Fellowship (DGE9616153) and a Declining Amphibian Populations Task Force Seed Grant during the 1997 and 1998 surveys and during the writing of this paper. William W. Dunlap was supported by an NSF Graduate Research Fellowship and a Sigma Xi grant during the 1978 surveys.

References
Allendorf, F.W., and Leary, R.F. 1986. Heterozygosity and fitness in natural populations of animals. In Conservation biology: the science of scarcity and diversity. Edited by M. Soul. Sinauer Associates, Sunderland, Mass. pp. 5776. Anzalone, CR., Kats, L.B., and Gordon, B.S. 1998. Effects of solar UV-B radiation on embryonic development in Hyla cadaverina, Hyla regilla, and Taricha torosa. Conserv. Biol. 12: 646653. Bahls, P. 1992. The status of fish populations and management of high mountain lakes in the western United States. Northwest Sci. 66: 183193. Blaustein, A.R. 1994. Chicken Little or Neros fiddle: a perspective on declining amphibian populations. Herpetologica, 50: 8587. Blaustein, A.R., and Wake D.B. 1990. Declining amphibian populations: A global phenomenon? Trends Ecol. Evol. 5: 203204.

Funk and Dunlap Mills, L.S., and Smouse, P.E. 1994. Demographic consequences of inbreeding in remnant populations. Am. Nat. 144: 412431. Newman, D., and Pilson, D. 1997. Increased probability of extinction due to decreased genetic effective population size: experimental populations of Clarki pulchella. Evolution, 51: 354362. Phillips, K. 1994. Tracking the vanishing frogs. St. Martins Press, New York. Pounds, J.A. 1991. Amphibian watch: new clues in the case of the disappearing amphibians. Wildl. Conserv. 94: 1618. Pounds, J.A., and Crump, M.L. 1994. Amphibian declines and climate disturbance: the case of the golden toad and the harlequin frog. Conserv. Biol. 8: 7285. Saccheri, I., Mikko, K., Kankare, M., Vikman, P., Fortelius, W., and Hanski, I. 1998. Inbreeding and extinction in a butterfly metapopulation. Nature (Lond.), 392: 491494. Sarkar, S. 1996. Ecological theory and anuran declines. BioScience, 46: 199207. Sjgren, P. 1991. Extinction and isolation gradients in metapopulations: the case of the pool frog (Rana lessonae). Biol. J. Linn. Soc., 42: 135147. Sjgren Gulve, P. 1994. Distribution and extinction patterns within a northern metapopulation of the pool frog (Rana lessonae). Ecology, 75: 13571367. Slatkin, M. 1985. Gene flow in natural populations. Annu. Rev. Ecol. Syst., 16: 393430. Tallmon, D.T., Funk, W.C., and Allendorf F.W. 1999. Rapid genetic differentiation of long-toed salamander populations. Copeia. In press. Thoms, C., Corkran, C.C., and Olson, D.H. 1997. Basic amphibian survey for inventory and monitoring in lentic habitats. In Sampling amphibians in lentic habitats. Edited by D. Olson, W.

1767 Leonard, and R. Bury. Society for Northwestern Vertebrate Biology, Olympia, Wash. pp. 3546. Tyler, T.J., Liss, W.J., Ganio, L.M., Larson, G.L., Hoffman, R., Deimling, E., and Lomnicky, G. 1998a. Interaction between introduced trout and larval salamanders (Ambystoma macrodactylum) in high-elevation lakes. Conserv. Biol. 12: 94105. Tyler, T.J., Liss, W.J., Hoffman, R.L., and Ganio, L.M. 1998b. Experimental analysis of trout effects on survival, growth, and habitat use of two species of ambystomatid salamanders. J. Herpetol. 32: 345349. Vos, C.C., and Chardon, J.P. 1998. Effects of habitat fragmentation and road density on the distribution pattern of the moor frog Rana arvalis. J. Appl. Ecol. 35: 4456. Wake, D.B. 1991. Declining amphibian populations. Science (Washington, D.C.), 253: 860. Wake, D.B., and Morowitz, H.J. 1991. Declining amphibian populations-a global phenomenon. Alytes, 9: 3342. Weller, W.F., and Green, D.M. 1997. Checklist and current status of Canadian amphibians. In Amphibians in decline: Canadian studies of a global problem. Edited by D. Green. Society for the Study of Amphibians and Reptiles, St. Louis, Mo. pp. 309328. Williams, S.E., and Hero, J.M. 1998. Rainforest frogs of the Australian Wet Tropics: guild classification and the ecological similarity of declining species. Proc. R. Soc. Lond. B Biol. Sci. 265: 597602. Wyman, R.L. 1990. Whats happening to the amphibians? Conserv. Biol. 4: 350352. Zaykin, D.V., and Pudovkin, A.I. 1993. Two programs to estimate significance of chi-square values using pseudo-probability tests. J. Hered. 84: 152.

1999 NRC Canada