HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

Ci>

MCZ
LfBRARY

O C CASIONAL PAPERS

A.^^^ t 6 1991

of the

MUSEUM OF NATURAL HISTORY

The University of Kansas * »

Lawrence, Kansas

NUMBER 141, PAGES 1-12 18 APRIL 1991

GENETIC VARIATION IN COASTAL

AND MONTANE POPULATIONS OF

AMBYSTOMA GRACILE

(CAUDATAiAMBYSTOMATIDAE)

Tom a. Titus' and Michael S. Gaines-

Gene flow has long been considered a critical factor in maintaining the
genetic cohesiveness of populations (Jordan, 1905; Mayr, 1 963; Dobzhansky,
1970). This view has been challenged by some authors, who maintain that
selection is the primary detemiinant of genetic cohesion and fragmentation
(Ehrlich and Raven, 1969; Endler, 1973). However, when traits are selec-
tively neutral, lack of gene flow should result in population differentiation
generated by the stochastic forces of mutation and genetic drift, whereas high
levels of gene flow should produce genetic homogeneity (Wright, 1943).
Genetic heterogeneity seems to characterize many natural populations and is
evidence of reduced gene flow. Numerous animal species, e.g., house mice
(Petras, 1967: Selander, 1970), gophers (Patton and Feder, 1981), snails
(Selander and Kaufman. 1975; Jones et al., 1977), lizards (Pounds and
Jackson, 1981), and salamanders (Hedgecock, 1978; Yanev and Wake, 1981),
display substantial amounts of genetic structuring among populations over
very small geographic distances. Banners to gene flow may result from

' Dyche Hall, Museum of Natural History, and Department of Systematics and
Ecology, The University of Kansas, Lawrence, Kansas 66045-2454.

-Haworth Hall. Department of Systematics and Ecology, The University of
Kansas, Lawrence, Kansas 66045-2106.

© Museum of Natural History. The University of Kansas. Lawrence.

2 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP., No. 141

extrinsic or intrinsic factors that limit dispersal. Extrinsic factors include
physiographic barriers and patchily distributed habitat, whereas intrinsic
factors are life history attributes such as homing behavior (Hedgecock, 1978)
and social behavior (Wright, 1965).

Ambystomatid salamanders display intra- and interspecific variation in a
life history attribute that may be important in determining dispersal capa-
bilities: the retention of the gill-breathing aquatic larval morphology into
adulthood, or larval reproduction (sensu Shaffer, 1984). Salamanders ex-
hibiting larval reproduction are restricted to an aquatic environment, whereas
individuals that metamorphose are capable of terrestrial dispersal. Elimina-
tion of overland dispersal, resulting from fixation of larval reproduction, has
been suggested as an explanation for greater genetic differentiation among
some species of Mexican ambystomatid salamanders (Shaffer, 1984). In
addition, obligately nonmetamorphosing salamander species display less
heterozygosity than do their metamorphosing counterparts (Shaffer and
Breden, 1990).

The northwestern salamander, Ambystoma gracile, displays intraspecific
variation in metamorphic failure (Nussbaum et al., 1983; and references
therein ). The incidence of neoteny in this species is positively correlated with
increasing altitude, stability of the aquatic habitat, lack of fish, and slower
larval growth rates (Snyder, 1956; Sprules, 1974a, b; Eagleson, 1976). Some
populations seem to be obligately neotenic (Snyder, 1956; J. Taylor, pers.
comm.). In addition, some nonmetamorphosing populations are in close
geographic proximity to one another and are not connected by aquatic
avenues for larval dispersal.

Despite the interesting life history variation among A. gracile populations
and the potential effect on the genetic structure of populations, microgeographic
genetic variation has not been studied in this species. In addition, few studies
of geographic variation in salamanders involve populations that are geo-
graphically close enough for movement of individuals among them or employ
the large sample sizes necessary for detecting small quantities of genetic
differentiation. We compared microgeographic genetic structure in a group of
coastal metamorphosing populations and a group of montane
nonmetamorphosing populations of A. gracile. Because of the greater poten-
tial for gene flow and concomitant reduction in genetic drift, we predicted that
the metamorphosing populations would show less among-population genetic
variation than populations fixed for larval reproduction.

METHODS AND MATERIALS

Larvae and nontransformed adults were sampled from populations in two
groups, one coastal and one montane. The coastal group was composed of
four populations located ca. 40 km northwest of Corvallis, Oregon, near Falls

TITUS & GAINES. 1991

GENETIC VARIATION IN AMBYSTOMA GRACILE

City: Little Sink Natural Area Upper Pond (UP), Little Sink Natural Area
Lower Pond (LP). Camp Kilowan Pond (CK). and Fire Pond (FP) (Fig. 1).
These coastal populations were in permanent ponds ranging from 0.4 to LO
ha in size. The maximum distance between any two coastal populations was
3 km. All ponds are located in the Teal Creek drainage, but are not directly
connected by waterways that might allow dispersal by larvae. The montane
group was composed of three populations located ca. 20 km west of Sisters,
Oregon, near Suttle Lake: Santiam Pass Pond (SP), Dark Lake (DL), and
Scout Lake (SL) (Fig. 2). The montane populations were in two lakes and a
pond, ranging in size from ca. 0.24 to 1 .0 ha. All bodies of water were per-
manent. The maximum distance between any two montane populations was ca.
7 km. No apparent avenues for aquatic dispersal existed between any of these
populations. The montane and coastal population groups are ca. 150 km apart.

Fig. 1 . Geographic distribution of coastal populations of A. gracile. CK = Camp
Kilowan, FP = Fire Pond, LP = Lower Pond, Little Sink Natural Area, UP = Upper
Pond, Little Sink Natural Area.

UNIV. KANSAS MUS. NAT. HIST. OCC. PAP., No. 141

Fig. 2. Geographic distribution of montane populations of A. gracile. DL= Dark
Lake, SL = Scout Lake, SP = Santiam Pass.

The montane and coastal populations examined in this study differ in the
relative incidence of metamorphosis. Ideally, the frequency of inetamorphosis
should be quantified for each population, but an accurate estimate of the
frequency of metamorphosis would require following a single cohort from
each population at least through the 14-16 ino larval period and preferably for
the entire lifespan of each individual because metamorphosis may occur after
sexual maturity has been reached. Previous attempts at mark-recapture
studies of A. gracile in Dark and Scout Lakes resulted in low recapture rates,
with no recaptures more than 2 wk after remarking (J. Taylor, pers. comm.).
However, a qualitative assessment of the two groups of populations indicates
differences in the frequency of metamoiphosis. Sampling of the Falls City
coastal populations produced only four nontransformed adults (easily distin-
guished on the basis of size) in ca. 500 larvae examined. No metamoiphosing
larvae or transfoirned adults were found in the montane populations of Dark
Lake or Scout Lake after searching both underwater and in the surrounding
terrestrial habitat in 1977, 1978, 1981, 1983, and 1985 (J. Taylor, pers.
comm.). Only one metamoiphosed individual was captured in the Santiam
Pass pond during this study, despite sampling during the breeding season (May-
June) of 1986 and in August 1985, a time when metamoiphosing individuals
were observed in other lakes in the Oregon Cascades (Titus, pers. obs.).

TITUS & GAINES. 1991

GENETIC VARIATION IN AMBYSTOMA GRACILE

Liver and skeletal muscle were removed in the laboratory from carcasses
frozen whole in liquid nitrogen, or tissues were taken in the field from animals
killed in 2% chloretone and immediately frozen in liquid nitrogen. Carcasses
were fixed in 10% fomialin. Voucher specimens are deposited at The
University of Kansas, Museum of Natural History. Tissue preparation and
electrophoresis generally followed the methods of Hillis (1985). Eighteen
presumptive gene loci (hereafter referred to as loci) were screened, and six
polymoiphic loci (loci with the rare allele present at a frequency <0.005 were
considered polymoiphic) were chosen based on degree of polymorphism and
consistent resolution: Est- 1 . Gpi-A, S-Aat-A, Pgdh-A, Pgm-B, and S-Sod-A.
Of these loci. S-Aat-A. Pgm-B, and S-Sod-A were not polymoiphic in the
montane populations and only Est-1. Gpi-A. and Pgdh-A were used in the
analysis of genetic structure in the montane populations. Enzyme nomencla-
ture follows the recommendations of the International Union of Biochemistry
Nomenclature Committee (lUBNC) (1984). Buffer systems, lUBNC num-
bers, and source tissues are listed in Table 1. Multiple loci were numbered
from anode to cathode. Electromoiphs for each locus (hereafter referred to as
alleles) were lettered a, b, and c in order of decreasing anodal mobility. Alleles
were labeled relative to this study only and are not meant to correspond with
allelic designations from other studies.

Chi-square contingency tables were used to test for allelic heterogeneity
between years within populations, for heterogeneity among populations, and
for comparison of genotypic frequencies to Hardy-Weinberg equilibrium.
Genetic variation was partitioned into within- and among-population varia-
tion using the G-statistics of Nei (1973). Gg^ is equal to Fj^ when calculated
for multiple alleles following Wright (1978) and is hereafter referred to as F^j.

Table 1. Loci examined, abbreviations. International Union of Biochemistry
Nomenclature Committee numbers, and buffer systems for protein electrophoresis in
A. gracile.

Locus

Abbreviation

lUBNC no.

Buffer system

Aspartate

S-Aat-A

2.6.1.1

Poulik

aminotransferase

Esterase (Acetate)

Est-1

3.1.1.1

Poulik

Glucose-6-phosphate

Gpi-A

5.3.1.9

Poulik

isomerase

Phosphogluconate

Pgdh-A

1 . 1 . 1 .44

T.C.-NADP-6.7

dehydrogenase

Superoxide dismutase

Sod-A

1.15.1.1

T.C.-NADP-6.7

6 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP., No. 141

Chi-square contingency tables and mean ¥^j were computed using BIOS YS-
1 release 1.7 (Swofford and Selander. 1981). G-statistics were computed
using LYNS (Marilyn Loveless, University ofGeorgia, Athens). Significance
values for chi-square contingency tables were adjusted for multiple com-
parisons by dividing 0.05 by the number of loci tested in each group of
populations (P < 0.008 for coastal populations, P < 0.016 for montane
populations). Similarly, significance values for chi-square tests for confor-
mation of genotypic frequencies to Hardy-Weinberg equilibrium were adjusted
by dividing 0.05 by the number of loci tested in each population.

RESULTS

Chi-square tests revealed no significant heterogeneity in allelic frequen-
cies between years in any population. For additional analyses, the 1984 and
1985 samples were pooled for populations sampled in both years. Allelic
frequencies and sample sizes for the pooled data are presented in Table 2.

For the coastal populations, no significant deviations from Hardy-Weinberg
equilibrium were found, and chi-square tests revealed no genetic heterogene-
ity among the coastal populations for any of the six polymorphic loci
examined. Statistics of genetic variation for these populations are presented
in Table 3. Total genetic variation, Hj, ranges from 0.114 (Gpi-A) to 0.428
(Pgm-B). Values for Ds^. the among-population variation, are small, ranging
from zero (Pgdh-A) to 0.0024 (Est-1), whereas values for Hj, the within-
population variation, range from 0.114 (Gpi-A) to 0.427 (Pgdh-B) and
account for 98.8-100% of the total variation at these loci. Values for Fj^, the
ratio Dsj/Hj, are therefore low, ranging from zero (Pgdh-A) to 0.01 1 (Est-1 )
(X = 0.006). These populations are genetically very similar to one another
qualitatively as well. Although four loci (S-Aat-A, Est-1, Gpi-A, and Pgdh-
A) have alleles present at frequencies of 0.200 or less (see Table 2), these
relatively rare alleles were present in all four populations examined except for
allele b of Pgdh-A in CK.

No deviations from Hardy-Weinberg equilibrium were found in the
montane populations. Significant genetic heterogeneity among the montane
populations was found for one locus, Pgdh-A (%- = 1 0.44, 2 df , P < 0.0 1 ), and
was the result of frequency differences in the Santiam Pass (SP) population
relative to Dark Lake (DL) and Scout Lake (SL). Total genetic variation (H.^,)
for three polymoiphic loci ranged from 0.378 (Gpi-A) to 0.496 (Pgdh-A). As
in the coastal populations, D^^^ is small, ranging from 0.002 (Gpi-A) to 0.010
(Pgdh-A), whereas Hs ranges from 0.377 (Gpi-A) to 0.486 (Pgdh-A) and
accounts for 98.3-99.6% of the total genetic variation. Values for F^-y are
correspondingly low, ranging from 0.004 ( Est- 1 ) to 0.020 (Pgdh-A ) ( x = 0.0 1 0).

TITUS & GAINES, 1991

GENETIC VARIATION IN AMBYSTOMA GRACILE

Table 2. Allelic frequencies and sample sizes («) for six polymorphic loci in four
coastal populations (FP, LP, CK, UP) and three montane populations (SL, DL, SP) of
A. gracilc.

Allele,

sample

size

Population

Metamorphosing

Neotenic

Locus

FP

LP

CK

UP

SL

DL

SP

S-Aat-A

a

0.112

0.098

0.167

0.121

0.009

0.005

0.035

b

0.888

0.902

0.833

0.879

0.991

0.995

0.965

n

116

127

30

91

55

104

86

Est-1

a

0.034

0.021

0.048

0.022

—

b

0.903

0.894

0.774

0.872

0.580

0.718

0.787

c

0.021

0.016

0.016

0.022

0.420

0.282

0.213

n

118

96

31

90

114

103

87

Gpi-A

a

0.034

0.021

0.048

0.022

—

—

b

0.941

0.943

0.919

0.944

0.741

0.718

0.787

c

0.025

0.036

0.032

0.034

0.259

0.282

0.213

n

118

96

31

89

114

103

87

Pgdh-A

a

0.934

0.934

0.952

0.933

0.017

b

0.013

0.014

—

0.011

0.557

0.505

0.322

c

0.053

0.052

0.048

0.056

0.446

0.495

0.661

n

114

106

31

90

111

95

87

Pgm-B

a

0.290

0.264

0.258

0.300

—

b

0.702

0.702

0.742

0.689

1.000

1.000

1.000

c

0.008

0.034

0.011

—

—

—

n

62

104

31

90

68

78

85

S-Sod-A

a

0.212

0.207

0.226

0.207

b

0.788

0.793

0.774

0.793

1.000

1.000

1.000

n

111

133

31

82

no

104

80

DISCUSSION

Levels of genetic variation among coastal populations of A. gracile
examined in this study were very low. Although numerous published esti-
mates of average Fsj are available for salamanders (see Larson et al., 1984;
and references therein), few meaningful comparisons can be made because of
the small geographic distances among populations in this study. The average
FsT (0.006) is considerably lower than values for Fj^ reported for populations
of Taricha rivularis within two drainages (0.024 and 0.029) (Hedgecock,
1978). Yanev and Wake (1981) reported a very high average Fgx of 0.470 for

8 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP., No. 141

Table 3. Statistics of genie diversity for polymorphic loci in four coastal and
three montane populations of A. gracile.

Population

type

Locus

H,

Hs

DST

FsT

Coastal

S-Aat-A

0.202

0.201

0.001

0.003

Est-1

0.213

0.211

0.002

0.011

Gpi-A

0.114

0.114

0.000

0.001

Pgdh-A

0.122

0.122

0.000

0.000

Pgm-B

0.428

0.427

0.001

0.002

S-Sod-A

0.332

0.332

0.000

0.000

Montane

Est-1

0.472

0.470

0.002

0.004

Gpi-A

0.378

0.377

0.002

0.004

Pgdh-A

0.496

0.486

0.010

0.020

Batrachoseps campi in a study that included many populations separated by
>5 km. Pierce and Mitton (1980) reported significant genetic heterogeneity
among four populations of Ambystoma tigrinum separated by a maximum of
1 km, but this may have represented a contact zone between /\. /. mavortium
and A. t. nehidosum.

In the coastal metamorphosing populations, the larger number of potential
dispersers should reduce genetic differentiation among populations because
of increased gene flow. Low Fst values and the distribution of alleles at several
loci suggest that gene flow does occur among these populations. Four loci (S-
Aat-A, Est-1 , Gpi-A, and Pgdh-A) have one allele present at high frequencies
in all populations, and the rare alternative allele(s) was also detected in all
populations. The only exception to this pattern was the absence of allele b of
Pgdh-A in CK, but the sample size was relatively small (/; = 31) and this allele
may actually have been present but not detected. If no gene flow had occuned
after an initial colonization event, these rare alleles should have been
eliminated from at least some of these populations as a result of founder
effects. Homing behavior has been reported for other ambystomatid species
(Stenhouse, 1985; and references therein) and should enhance among-
population differentiation by reducing dispersal. If philopatry is exhibited by
A. gracile, it has not been strong enough to produce significant among-
population differentiation in the coastal A. gracile populations, possibly be-
cause dispersal by a small number of individuals is sufficient to "swamp out"
the effects of genetic drift (Kimura and Maruyama. 197 1 ).

The low incidence of metamorphosis in the montane populations should
reduce the likelihood of gene flow among them and should enhance among-
population genetic variation. The montane populations did exhibit a higher

TITUS & GAINES. 1 99 1 GENETIC VARIATION IN AMBYSTOMA GR AGILE 9

average F^j (0.010), but this was primarily the result of allelic frequency
differences in the Santiam Pass population at one locus (Pgdh-A); this
population is ca. 7 km from the Scout Lake population and ca. 8 km from the
Dark Lake population, whereas the maximum distance between any two
coastal populations is ca. 3 km. Thus, slightly more differentiation among the
montane populations relative to the coastal populations could also be the
result of isolation by distance rather than, or in addition to, differences in the
frequency of metamoiphosis.

Perhaps the most interesting outcome of this study is the apparently high
level of genetic continuity among the montane populations where the prob-
ability of an individual metamorphosing, dispersing, and successfully mating
in a different population seems low. At least three hypotheses can be posited
to account for this pattern. ( 1 ) Metamoiphosis and dispersal may be more
common in the study populations than previously thought. (2) Metamoipho-
sis may be common in neighboring lakes (e.g.. Cache Lake and Hand Lake,
ca. 3-A km away) and gene flow from these populations could maintain
genetic continuity among the study populations. (3) Metamorphosis and
dispersal may have been more common in these populations in the recent past;
the Dark and Scout lake populations number in the thousands (Taylor, 1983),
and the combined effects of large effective population sizes and high levels
of variation at the three polymorphic loci examined would make genetic drift
an extremely slow process. Future testing of these hypotheses should include
sampling with genetic markers more sensitive to the effects of mutation and
genetic drift. Variation in the D-loop region of the mitochondrial genome may
provide additional information because of a higher realized mutation rate and
matrilineal inheritance (reviewed by Brown, 1985).

The montane populations occur in an area affected by Pleistocene vol-
canism, whereas the area containing the coastal populations has not expe-
rienced a geological disturbance since the Miocene (Baldwin, 1976). There-
fore, if effective population size and gene flow were equal among populations
in both groups, more differentiation would be expected among the coastal
populations. However, little differentiation has occurred among the coastal
populations, further supporting the hypothesis that gene flow is a common
occurrence. The geologic evidence indicates that colonization of the montane
habitat and the subsequent evolution of larval reproduction may have been
relatively recent, reducing the time available for genetic drift to have
occurred.

Lower heterozygosity has been reported for nonmetamorphosing sala-
mander species compared with metamoiphosing species ( Shaffer and Breden,
1990). If this overall pattern is the result of factors operating at the level of
populations, then nonmetamoiphosing populations within a species should
also display lowerheterozygosity than theirmetamoiphosing counterparts. In
A. gracile, levels of within-population variation might be expected to be

10 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP.. No. 141

lower in the montane populations relative to the coastal populations because
of reduced gene flow accompanying high levels of metamorphic failure and
genetic drift or, as suggested by Shaffer (1984), because of founder effects
resulting from rare dispersal events by metamorphosed individuals. Most
individuals in this study were examined for only a few polymorphic loci and
mean heterozygosity {H) will therefore be inflated, precluding comparisons
between other populations and species. However, H was reported for one
coastal (Upper Pond ) and one montane ( Scout Lake ) population using 1 8 loci
chosen regardless of level of variation (Titus, 1990). Heterozygosity was
higher in Upper Pond (H = 0.170, SD = 0.052) compared with Scout Lake
(// = 0. 1 2 1 , SD = 0.054), but these values were within one standard error of
one another and were the highest heterozygosities reported for six populations
of A. gracile. A lower level of heterozygosity {H = 0.092, SD = 0.036) occurs
in a population near Toad Creek, Oregon, ca. 30 km northwest of the Scout
Lake population, and sampling at Toad Creek revealed no nontransformed
adults (Titus, pers. obs.). Thus, the pattern of lower heterozygosity seen in
nonmetamorphosing species so far is not reflected at the population level in
A. gracile. However, relatively few populations have been studied, and
additional populations should be examined to evaluate this hypothesis further.

SUMMARY

AUozyme variation was compared in coastal metamorphosing popula-
tions and montane nonmetamorphosing populations of the northwestern
salamander, Kmhy stoma gracile. In both groups of populations, 98-100% of
the total genetic variation for each locus can be attributed to within-popula-
tion variation. No significant genetic heterogeneity was found among coastal
populations, whereas significant heterogeneity was found among montane
populations at one locus, Pgdh-A. The slightly greater genetic differentiation
observed among montane populations relative to the coastal populations may
be the result of larger geographic distances among these populations rather
than life history differences. The high level of genetic continuity among
nonmetamoiphosing montane populations could be the result of ( 1 ) undetec-
ted metamorphosis and dispersal within these populations, (2) dispersal of
metamorphosed individuals from surrounding populations into the study
populations, or (3) a combination of recent evolution of larval reproduction,
large effective population sizes, and initially large amounts of genetic
variation at polymorphic loci, impeding genetic drift. No relationship be-
tween heterozygosity and metamoiphic failure has been observed in popu-
lations of A. gracile studied so far.

TITUS & GAINES. 1991 GENETIC VARIATION IN AMBYSTOMA GRACILE 1 1

ACKNOWLEDGMENTS

We are indebted to T. N. Titus. B. J. Titus, and L. Spring for providing food
and housing for field workers. D. M. Hillis provided laboratory space and
technical assistance during the initial stages of the project. We also thank R.
de Sa, D. Danley. B. Dutton. E. Dutton, L. S. Dryden, and K. Wollter for
assistance in the field. P. Wolf and R. Matson provided valuable assistance
with data analysis. Laboratory space was provided by L. Spring at Western
Oregon State College. W. E. Duellman, C. Haufler, M. L. Johnson, S. Johnson,
R. F. Johnston, J. Taylor, and K. Wollter made critical comments on the
manuscript. Financial support was provided by grants from The University of
Kansas General Research Fund (MSG) and the Department of Systematics
and Ecology, the Herpetological Research Fund of the Museum of Natural
History, and Sigma Xi (TAT).

LITERATURE CITED

Baldwin, E. M. 1976. Geology of Oregon. Kendall-Hunt Publishing Co., Dubuque,

Iowa.
Brown, W.M. 1985. The mitochondrial genome ofanimals, pp. 95-1 30. In: Maclntyre,

R. J. (ed.). Molecular Evolutionary Genetics. Plenum Press, New York.
DoBZHANSKY, T. 1 970. Genetics and the Origin of Species. Columbia Univ. Press, New

York.
Eagleson, G. W. 1976. A comparison of the life-histories and growth patterns of

populations of the salamander Am/^v^towa gracile (Baird) from permanent low

altitude and montane lakes. Can. J. Zool. 54:2098-20 11.
Ehrlich, p. R., and p. H. Raven. 1 969. Differentiation of populations. Science 1 65: 1 228-

1232.
Endler, J. A. 1973. Gene flow and population differentiation. Science 179:243-250.
Hedgecock, D. 1978. Population subdivision and genetic divergence in the red-bellied

newt, Taricha rivulahs. Evolution 32:271-286.
Hillis, D. M. 1985. Evolutionary genetics of the Andean lizard genus PhoUdoholus

(Sauria: Gymnophthalmidae): phylogeny. biogeography, and a comparison of tree

construction techniques. Syst. Zool. 34:109-126.
International Union of Biochemistry. Nomenclature Committee. 1984. Enzyme

Nomenclature, 1984. Academic Press, New York.
Jones, S., B. H. Leith, and P. Rawlings. 1977. Polymorphism in Cepaea: a problem

with too many solutions? Annu. Rev. Ecol. Syst. 8:109-143.
Jordan, D. S. 1905. The origin of species through isolation. Science 22:542-562.
Kimura, M., and T. Maruyama. 1971. Pattern of neutral polymorphism in a geo-
graphically structured population. Genet. Res. 11:247-269.
Larson, A., D. B. Wake, and K. P. Yanev. 1984. Measuring gene flow among popu-
lations with high levels of genetic fragmentation. Genetics 106:293-308.

12 UNIV. KANSAS MUS. NAT. HIST. OCC. PAP.. No. 141

Mayr, E. 1963. Animal Species and Evolution. Belknap Press. Cambridge, Massa-
chusetts.
Nei, M. 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad.

Sci. USA 70:3321-3323.
NussBAUM, R. A., E. D. Brodie, Jr., and R. M. Storm. 1983. Amphibians and Reptiles

of the Pacific Northwest. Univ. Press Idaho, Moscow.
Patton, J. L., AND J. H. Feder. 1981. Microspatial genetic heterogeneity in pocket

gophers: nonrandom breeding and drift. Evolution 35:912-920.
Petras, M. L. 1967. Studies of natural populations of Miis. I. Biochemical polymor-
phisms and their bearing on breeding structure. Evolution 21:259-274.
Pierce, B. A., and J. B. Mitton. 1980. Patterns of allozyme variation m Amby stoma

tigrimim mavortium and A. t. nehiilosiim. Copeia 1980:594-605.
Pounds, J. A., and J. F. Jackson. 198 1 . Riverine barriers to gene flow and the differ-
entiation offence lizard populations. Evolution 35: 516-528.
Selander, R. K. 1970. Behavior and genetic variation in natural populations. Am.

Zool. 10:53-66.
Selander. R. K., and D. W. Kaufman. 1975. Genetic structure of the brown snail

(Helix aspera). I. Microgeographic variation. Evolution 29:385-401.
Shaffer, H. B. 1984. Evolution in a paedomoiphic lineage. I. An electrophoretic

analysis of the Mexican ambystomatid salamanders. Evolution 38:1 194-1206.
Shaffer, H. B., and F. Breden. 1990. The relationship between allozyme variation and

life history: nontransforming salamanders are less variable. Copeia 1990:1016-

1023.
Snyder, R. C. 1956. Comparative features of the life-histories oiAmbystoma gracile

(Baird) from populations at low and high altitudes. Copeia 1956:41-50.
Sprules, W. G. 1974a. The adaptive significance of paedogenesis in North American

species oiAmbystoma (Amphibia: Caudata): an hypothesis. Can. J. Zool. 52:393-

400.
. 1974b. Environmental factors and the incidence of neoteny \n Ambystoma

gracile (Baird) (Amphibia: Caudata). Can. J. Zool. 52:1545-1552.
Stenhouse, S. L. 1985. Migratory orientation and homing in Ambystoma macidatum

and Ambystoma opacum. Copeia 1985:631-637.
Swofford, D. L., and R. B. Selander. 1 98 1 . BIOS YS- 1 : a FORTRAN program for the

comprehensive analysis of electrophoretic data in population genetics and

systematics. J. Hered. 72:282-283.
Taylor. J. 1983. Orientation and flight behavior of aneotenic salamander (Am/jy^roma

gracile) in Oregon. Am. Midi. Nat. 109:40^9.
Titus. T. A. 1 990. Genetic variation in two subspecies ofAmlnstoma gracile (Caudata:

Ambystomatidae). J. Heipetol. 24:107-111.
Wright. S. 1943. Isolation by distance. Genetics 28:1 14-138.
. 1965. The inteipretation of population structure by F-statistics. with special

regard to systems of mating. Evolution 19:395—420.

1978. Evolution and the Genetics of Populations, vol. 4. Variability within

and among Natural Populations. Univ. Chicago Press. Chicago.
Yanev, K. P., AND D. B . Wake. 1 98 1 . Genie differentiation in a relict desert salamander.
Batrachoseps campi. Heipetologica 37: 16-28.

UNIVERSITY OF KANSAS

MUSEUM OF NATURAL HISTORY

PUBLICATIONS

The University of Kansas Publications, Museum of Natural History,
beginning with volume 1 in 1946, was discontinued with volume 20 in
1971. Shorter research papers formerly published in the above series are
now published as The University of Kansas Museum of Natural History
Occasional Papers. The University of Kansas Museum of Natural
History Miscellaneous Publications began with number 1 in 1946.
Longer research papers are published in that series. Monographs of the
Museum of Natural History were initiated in 1970. Authors should
contact the managing editor regarding style and submission procedures
before manuscript submission. All manuscripts are subjected to critical
review by intra- and extramural specialists; final acceptance is at the
discretion of the Director.

This publication is printed on acid-free paper. Occasional Papers and
Miscellaneous Publications are typeset using Microsoft- Word and Aldus
PageMaker' on a Macintosh computer. - Museum of Natural History,
The University of Kansas, Lawrence.

Institutional libraries interested in exchanging publications may obtain
the Occasional Papers and Miscellaneous Publications by addressing the
Exchange Librarian, The University of Kansas Library, Lawrence,
Kansas 66045-2800, USA. Individuals may purchase separate numbers
from the Office of Publications, Museum of Natural History, The
University of Kansas, Lawrence, Kansas 66045-2454, USA.

Editor. Robert D. Holt

Managing Editor. Joseph T. Collins

Design and Typesetting: Kate Shaw and Joseph T. Collins

Printkd b**

University of Kansas Printing Service

Lawrence, Kansas

DATE DUE

-MAR n 1996

DEMCO. INC 38-2931

2044 093 361 707