The Ambystoma laterale-jeffersonianum Complex

The Ambystoma laterale-jeffersonianum Complex in Central Ontario: Ploidy Structure, Sex
Ratio, and Breeding Dynamics in a Bisexual-Unisexual Community
Author(s): Leslie A. Lowcock, Hugh Griffith, Robert W. Murphy
Source: Copeia, Vol. 1991, No. 1 (Feb. 7, 1991), pp. 87-105
Published by: American Society of Ichthyologists and Herpetologists
Stable URL: http://www.jstor.org/stable/1446251 .
Accessed: 04/02/2011 09:20
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=asih. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
American Society of Ichthyologists and Herpetologists is collaborating with JSTOR to digitize, preserve and
extend access to Copeia.
http://www.jstor.org
Copeia,1991(1),pp. 87-105
The Ambystomalaterale-jeffersonianumComplex in Central
Ontario: Ploidy Structure, Sex Ratio, and Breeding
Dynamics in a Bisexual-unisexual Community
LESLIEA. LOWCOCK,HUGH GRIFFITH
AND ROBERT W. MURPHY
Ploidy ratio was investigated in six Ontario populations of the Ambystoma
complex by collecting blood from individual salamanders
laterale-jeffersonianum
during spring migrations, and analyzing the nuclear DNA content of erythrocytes by flow cytometry. All populations contained diploid male and female A.
laterale as well as a potential mixture of diploid, triploid, and tetraploid female
hybrids, which occurred at variable frequencies. Triploid males, pentaploid females and possible polyploid A. laterale (3n and 4n) were found in one extensively sampled population. Initial separation of A. laterale from hybrids based
on morphological criteria was tested for accuracy by comparison with the cytometric results and found to be exceptionally high (99.6%). Sex ratios in all
populations were biased overwhelmingly towards females, primarily because
female hybrids accounted for up to 84%of a breeding aggregate;however, diploid
female A. laterale outnumbered males in several populations and in two instances, near the northern limit of hybrid biotypes, these account for much of the
biased sex ratio. Timing of breeding, always triggered by precipitation (rain or
snow), varied between sites and was correlated with altitude/latitude. Initial
immigrations at all sites contained biotypes indicative of local primary composition with respect to both ploidy and hybridity. The breeding dynamics of one
population were investigated by daily sampling throughout the breeding period.
Migration occurred in distinct waves. In A. laterale, frequency of immigrating
males declined over the breeding period while frequency of females increased,
a pattern conforming to the typical Ambystomabreeding dynamic. Within hybrids, the percentage of tetraploids increased over the breeding period. Within
waves, there was a general increase in hybrids over time.
T
HE evolution and ecology of unisexual vertebrates is an emerging area of concern
in organismal biology. Reasons for this include:
1) occurrences of complexes containing unisexual and bisexual biotypes in fishes, amphibians
and reptiles are more widespread than previously thought (Dawley and Bogart, 1989); 2)
reproductive modes employed by these forms
can be varied and confounding, even within a
population (Bogart et al., 1987, 1989); 3) unisexuals in a given association seldom form a
homogeneous group (Bogart et al., 1987; Bogart, 1989; Lowcock, 1989); 4) interspecific hybrid origin of most unisexual forms appears to
confer a high degree of tolerance for aberrant
meiotic processes and elevated levels of ploidy
(Dawley and Bogart, 1989); and 5) an evolutionary role exists for some of these forms which
may involve direct speciation, or an effect on
incipient speciation in associated bisexuals (Bogart, 1989; Lowcock, 1989).
One of the most interesting and least studied
of these features is the lack of homogeneity
within the unisexual segment of a population,
and the relative proportions of different unisexual forms with respect to each other and
their bisexual cohorts. Variability in genome
constitution and ploidy among unisexuals has
been carried to an extreme in hybrid complexes
of the North American salamander genus Ambystoma,in which male and female diploid bisexuals breed in aggregations which can also
contain diploid (2n), triploid (3n), tetraploid (4n)
and pentaploid (5n) unisexual di- or trihybrids.
In genetic and ecological studies of such populations, it is desirable to be able to obtain
unambiguous baseline data on ploidy composition and sex-ratio without creating sampling
artifact in the population.
Numerous studies have established that unisexuals in Ambystomatypically outnumber bisexuals in these populations, but these were unable
? 1991 by the American Society of Ichthyologists and Herpetologists
88
COPEIA, 1991, NO. 1
to distinguish between ploidy classes in unisexual hybrids (Clanton, 1934; Uzzell, 1964; Wilbur, 1971). Genetic studies carried out on randomly collected samples (up to 149 individuals
from the same locality; Bogart and Licht, 1986)
have demonstrated that variable ploidy and genomic compositions characterize most unisexual populations. At best, however, these studies
have been only suggestive of ploidy and sex ratios (Bogart, 1989; Bogart et al., 1985, 1987).
Populations of Ambystomacan be very large, and
reliable methods for elucidating ploidy composition in samples of several hundred individuals can be prohibitively time consuming and
necessitate sacrificing of the animals. Methods
of ploidy evaluation that do not require sacrificing, such as erythrocyte area measurements
(Uzzell, 1964; Wilbur, 1971; Austin and Bogart,
1982; Bogart et al., 1985) and ectodermal cell
or nuclei area measurements (Morris and Brandon, 1984; Licht and Bogart, 1987) can also be
slow, labor intensive and lack the precision of
other methods at higher levels of ploidy. Additionally, random sampling of breeding populations may yield biased results due to ecological and behavioral partitioning of males,
bisexual females and genomically variable polyploid biotypes (Bogart et al., 1987).
Flow cytometry (FCM), a rapid and highly
sensitive method for measuring the nuclear
DNA content of cells, has been used in the investigation of a unisexual complex of hybrid
fishes. Dawley and Goddard (1988) employed
FCM in their analysis of diploid, triploid and
diploid-triploid mosaic biotypes of Phoxinus eosneogaeus. The ability to use FCM for non-destructive gathering of ecological data based on
ploidy comparisons among large numbers of individuals is of great benefit to the study of bisexual-unisexual associations. The accuracy of
this method precludes the necessity of chromosome confirmation of ploidy when samples
are run against standards of known ploidy (Tank
et al., 1987). Where hybrids involve two species
of differing DNA content, FCM may also indicate genomic composition (Dawley and Goddard, 1988) although the higher the ploidy, the
less informative this becomes. Unequivocal genotypic classification of polyploids in hybrid
complexes remains most reliably accomplished
through allozyme electrophoresis. Where hybrids involve species of similar DNA content
(such as in Ambystoma),genotypic inferences can
be made only with reference to information
gleaned from molecular investigations and a
priori knowledge of biogeographic distributions of particular bisexual and unisexual biotypes (Lowcock, 1989). In the present investigation, the template genomic composition of
polyploid biotypes associated with populations
of the bisexual diploid A. laterale are known
from genetic and/or biogeographic data (Lowcock, 1989), and FCM is used solely to compare
relative amounts of DNA among individuals.
In this paper, we describe the first application
of FCM to large numbers of unisexuals (> 100),
and obtain the first comprehensive data on ploidy and sex-ratios in breeding aggregations of
the A. laterale-jeffersonianumcomplex (nomenclature of Lowcock et al. [1987]). These data
contain a number of unique and unexpected
results with regard to ploidy levels and ratios
within hybrids. Mechanisms are considered to
account for each ploidy level. Sex-ratios within
syntopic A. laterale failed to conform with theoretical expectations; females outnumbered
males, and we discuss possible reasons for this.
For one population, sampled comprehensively over the breeding period, we compared arrival time of individuals of different sex or ploidy, because timing has been reported to vary
along these parameters (Uzzell, 1964; Panek,
1978; Weller, 1980). At this locality, we also
tested to see if ploidy composition remained
relatively constant among peak immigration
events throughout the breeding period; this
provided a measure of the validity of ploidyratio data generated by one-time samples from
other areas. Local influences of latitude, physiography and climate on timing of breeding were
compared among samples and evaluated.
METHODS AND MATERIALS
Sampling.-Adult salamanders were collected
by various methods during breeding migrations
to five ponds and one series of marshes in central Ontario (Fig. 1). The most successful and
reliable method was a drift-fence/pitfall trap
system (Gibbons and Semlitsch, 1982) at Haliburton Beaver Pond (HBP), Oro Pond (OP),
and Algonquin Highlands Pond (AHP). Other
immigrants were collected as they crossed asphalt barriers (analagous to a fence; manned
and patrolled for the entire evening) separating
overwintering habitat from breeding sites: Coboconk Pond (CP) and Algonquin Lowlands (AL).
Salamanders from Fort Irwin Pond (FIP) were
collected by seining and dipnetting. Ponds were
checked frequently before the breeding season
LOWCOCK ET AL.-PLOIDY
IN HYBRID AMBYSTOMA
89
AlgonquinHighland
*
Algonquin Lowland
Fort
L.Huron
Irwin
Haliburton
Oro
Coboconk
L.Ontario
100 km
LErie
Fig. 1. Southern and central Ontario, showing location of the six sites from which migratingsalamanders
were sampled.The dashed line at the top indicatesthe approximatenorthern limit of hybridbiotypes,beyond
laterale.Inset shows area of Ontario enlarged.
which all populationsare pure Ambystoma
to ensure that immigration had not been initiated prior to sampling. Samples were collected
during the first waves of breeding migrations
at all localities except HBP, where samples were
collected daily throughout the immigration period.
pond. Handling time was minimized; the average time from capture to release was less than
24 h. To test the influence of processing on
breeding behavior, several recently revived salamanders were paired and placed in 5 litre
buckets filled with pond water and detritus.
Processing.--Blood preparations for FCM were
obtained in the following manner: animals were
anesthetized in a weak solution of tricaine methanosulfate (MS-222), examined for sex and/or
hybridity, subjected to three measurements with
vernier calipers (snout-vent length [SVL], head
width [HW] and internarial distance [IND]; see
Zeyl and Lowcock [1989] for measurement parameters), and 1-2 drops of blood extracted
from an interphalangeal toe-clip. For each sample, whole blood was mixed in a pyrex depression-plate well, with enough freezing solution
(2-3 drops, Dawley and Goddard, 1988) to effect a pink color. This suspension was transferred by 0.1 ml pipettor or hematocrit tube to
1.8 ml screw-top cryotubes and frozen immediately in liquid nitrogen (-196 C). Salamanders were revived by washing in pond water,
and recuperated briefly in a pan lined with wet
paper towelling prior to transport back to the
Ploidy analysis.-In order to determine the ratios of: 1) A. laterale: hybrids; 2) male: female A.
laterale; 3) males: total females (hybrids + A.
laterale) and; 4) 2n:3n:4n hybrids, we initially
separated A. laterale from putative hybrids based
on visually evaluated morphological criteria, so
FCM would also test the reliability of subjective
field distinctions. When absolute size could not
be used (A. laterale <80.0 mm SVL; Appendix
2, Lowcock, 1986), separation was primarily
based on: 1) dorsal and ventral background color (lighter in hybrids; Clanton, 1934); 2) head
shape and width at the angle of the jaw (wider
in hybrids relative to SVL; Lowcock, 1986); and
3) IND (A. laterale, 2.5-3.9 mm, : = 3.25; syntopic hybrids, 3.3-4.7 mm, R = 4.10; Appendix
2, Lowcock, 1986). Ploidy classes within the unisexual segment of other hybrid complexes of
Ambystoma cannot be reliably determined
through morphological inspection (Zeyl and
90
COPEIA, 1991, NO. 1
1.
TABLE
RESULTS
OF TEST
CROSSES
BETWEEN
cipitation type) were recorded during sample
collection. These were coupled with regional
data obtained from Atmospheric Environment
Services (Environment Canada) weather monitoring stations at closest proximity to each site.
FEMALES AND MALE Ambystomalaterale FROM
HBP WHICH WERE PROCESSED FOR BLOOD.
VARIOUS
Female
Eggs laid
Fertilized
Viable to
hatching
% Successa
A. laterale
A. laterale
153
212
134
189
116
154
87.5/86.6
89.2/81.4
LLJb
167
130
102c
LLJ
LLLJ
143
84
140
47
98c
17c
77.8/78.5
97.9/70.0
56.0/36.2
SFertilization/hatching; survival unknown.
b i.e., A. (2)laterale-jeffersonianum;
nomenclature of Lowcock et al. (1987).
c Many abnormal larvae, survival rates likely low.
Lowcock, 1989), so this was not attempted. Males
were readily identified by a laterally compressed
tail, squarish snout and conspicuously swollen
cloacal area.
To determine nuclear DNA content of erythrocytes, we followed the FCM staining procedures of Dawley and Goddard (1988), substituting propidium iodide (PI; fluoresces over a
broad spectrum, but not UV) for diamidino-2phenylindole (DAPI; fluoresces in the UV) as
the nuclear stain. We used an argon-laser EPICS V Flow Cytometer (Coulter Electronics, Hialeah, FL) to excite PI at 488 nm with 500 mw
of power, and collected red fluorescent wavelengths above the limit established by a 595 X
interference filter. For each sample, 100020,000 nuclei were examined and histograms
were accumulated by the MDADS computer
system (Coulter Electronics). Nuclear DNA
content of samples in each FCM run were determined based on comparisons to a known diploid sample of A. laterale.
Environmentalparameters.--Air and water temperatures, as well as other general and local
conditions (ice cover, elevation, latitude, preTABLE
2.
COMPARISONS
OF DATE
Site
Coboconk (CP)
Oro Pond (OP)
Haliburton(HBP)
Fort Irwin (FIP)
Algonquin Lowlands(AL)
Algonquin Highlands (AHP)
Statistics.-To test for independence of breeding waves to relative frequencies of sexes and
ploidies, chi-square (x2) and G-tests (Sokal and
Rohlf, 1981) were conducted on the daily data
from HBP. In these tests, we include only data
from the 8 d on which significant migration
occurred
individuals) and for which we
(>-40 ploidy ratios by FCM (4-8, 12had determined
14 April).
RESULTS
Salamanders recovered from anesthesia within one-half h and did not appear to suffer illeffects from toe-clips; post-breeding individuals
with toe-clips were captured leaving HBP and
were frequently encountered in the area in summer and autumn of 1988. Test matings involving male A. laterale and female A. laterale or
hybrids demonstrated that processing was not
traumatic to breeding behavior or ability. All
test matings produced viable (fertilized) eggs
within 24 h. Hatching success of eggs laid by
female A. laterale or hybrids were concordant
with expectations based on previous reports
(Uzzell, 1964; Weller, 1980) and our own field
observations (Table 1).
Breeding timing.-Date of first breeding varied
between sites and was positively correlated with
both latitude and altitude (Table 2). Most
breeding migrations coincided with significant
daytime or overnight rainfall. Although evening migrations continued when rain turned to
snow after nightfall (HBP, AHP, AL, FIP), large
OF FIRST SIGNIFICANT
BREEDING,
TIONS AT SIX CENTRAL
ONTARIO
Altitude (m)
Lat(N)/Long(W)
Date
274.3
283.5
335.3
365.8
396.2
472.4
44.39/78.48
44.44/78.56
45.00/78.30
45.15/78.20
45.45/78.20
45.35/78.40
3 April
3 April
3 April
14 April
23 April
28 April
In 24 h period prior to evening migration.
PHYSICAL AND ENVIRONMENTAL
CONDI-
SITES.
Snow
cover
90%
75%
80%
Ice
cover
10%
95%
5%
95%
Precipitation"
(mm)
Temperaturea
(min/max C)
8.4
8.4
11.0
12.4
24.8
9.6
6/12
6/12
-2/12
4/8
-6/7
3/9
LOWCOCK ET AL.-PLOIDY
migrations were not initiated when daytime
precipitation was snow; diurnal snow is typically
associated with unfavorable temperatures.
At HBP, immigration began on the evening
of 3 April and continued until the morning of
28 April. Peak immigrations occurred in distinct waves coinciding with optimal conditions
of temperature and moisture (4, 7, 13-14, 18,
24 April), and were typically followed by "stragglers" (frequently abundant) that arrived during non-optimal conditions (5-6, 8-9, 15, 21,
25-28 April) (Fig. 2). On 11-12 April, immigration initiated under apparently non-optimal
parameters was correlated with other factors.
Immigration was halted by inclement weather
on four occasions, and did not resume until
weather patterns had meliorated. Both sexes
and the locally commonest hybrid ploidy levels
were represented during the first wave of immigration at all sites. Thus, first immigrations
may be reliable indicators of local primary composition of biotypes. This was true at HBP where
ploidy ratios on day 1 were concordant with
overall primary composition, but not exact frequencies; there are generally more male A. laterale in early migrations, and proportions of unisexuals can increase through time within waves.
Because ploidy ratio within hybrids on day 1
(and wave 1 in general) varied significantly from
that of subsequent waves (Table 3), hybrid ploidy ratios could only be roughly estimated for
the latest immigration dates at HBP, in which
hybrids were counted but not differentiated with
respect to ploidy.
Ploidy structure.--A total of 1119 salamanders
were collected and examined. Of these, 1025
individuals were screened for ploidy; the remaining 94 were late immigrants at HBP (1527 April) and could be classified unambiguously
only with respect to sex and hybridity. Hybrid
ploidy ratios were roughly estimated in HBP
samples collected after 14 April (Fig. 2) based
on results from the 710 individuals collected
prior to this date; we observed that the percentage of tetraploids remained identical
through waves 2-3, which accounted for 65%
of the population (migrations after 14 April accounted for only 8.6%). These estimates are for
comparative purposes only and are not necessarily reliable extrapolations of trajectories in
late-immigrant ploidy ratios, which may have
increased or decreased.
Results of FCM were unambiguous, demonstrating that all populations contained diploid
IN HYBRID AMBYSTOMA
TABLE 3.
91
PLOIDY RATIOS OF HYBRIDS IN FIRST THREE
IMMIGRATION
Migration day (n > 20)
WAVES AT HBP.
% 4n
Approximate
ratio 3n:4n
1 (4 April)
2 (5 April)
3 (6 April)
Wave 1 total
8.8
8.0
2.1
6.7
10:1
12:1
47:1
14:1
4 (7 April)
5 (8 April)
Wave 2 total
12.7
15.7
13.4
7:1
5:1
6:1
6 (12 April)
7 (13 April)
8 (14 April)
9 (15 April)a
Wave 3 total
14.7
13.4
12.7
13.4
13.2
6:1
6:1
7:1
8:1
7:1
Overall total, waves 1-3
12.3
8:1
* Estimated based on within-wave data previous to this date.
A. laterale as well as a variable potential mixture
of; 1) 2n, 3n, 4n, and 5n female hybrids; and
2) rare male hybrids and/or polyploid A. laterale
(Fig. 3). Successive ploidy levels above 2n showed
characteristic increases of DNA corresponding
to the addition of a full haploid set of chromosomes: erythrocyte nuclei contained exactly
1.5 x the DNA content of a diploid, 4n nuclei
contained 2 x the diploid amount and 5n nuclei
2.5 x (Fig. 4). The distribution of ploidy classes
in our samples are compared in Figure 3 and
detailed in Table 4 (AHP, AL, CP, OP, FIP)
and Table 5 (HBP). Aneuploidy was not indicated and ploidy mosaics were not encountered.
A priori assignment of individuals to either
A. laterale or the A. laterale-jeffersonianumhybrid
category was compared with FCM results (Table 6) and found to be exceptionally accurate.
Unless specifically noted, hybrids were initially
assumed to be of 3n or higher ploidy. Based on
this, overall prediction accuracy with respect to
hybridity was 99.98% (range 99.6-100%). An
individual from OP classified a priori as hybrid,
but which had a ploidy determination of 2n
(therefore "misclassified" under our assumption), was reevaluated on morphological
grounds. This animal was clearly a hybrid based
on size alone (>85.0 mm SVL). A second individual from OP was originally classified as a
diploid hybrid, the sole exception to our ploidy
assumption for hybrids, and was confirmed by
FCM. Three polyploid females originally classed
COPEIA, 1991, NO. 1
92
20-
A
HM
precipitation
15-
RS
ERS
E10-H
RS
S
5
L
MMIL
L S
B -
RS
R
IRS
R
S
S
S
R R
R
R
1
O
0.1
water
--
temp..,--•./.-
04
0
20C
air temp.
O15-
max
010
a
min(HBP)
5
,--"min
0-
250
D-8
S5n
200
a
ploidy
a
composition
other
-34n
S150
E
-2n
S2n
100
male
S3n
E
_
509
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
April days
Fig. 2. A comparisonof environmental parameters,numbers of migrants and daily ploidy composition
during the breeding regime at HBP. All X-axes are April days. From the top of the figure: A) precipitation
amount and type (AES data, Environment Canada plus our own local observations),R = rain, H = heavy
rain, L = light rain, M = mist, S = snow. B) Comparisonof water temperature (left Y-axis)to percent ice
cover (right Y-axis).C) Maximumand minimum daily air temperaturesderived by averaging temperatures
from three AES monitoring stations within 10 km of the site. Minimumlocal temperature, as measured at
the site (1 m above ground) is comparedto the area minimumin order to show meliorations.D) Comparison
of daily ploidy composition and numbersof immigrantsduring the breeding regime.
as A. laterale may be 3n or 4n forms of this
species. Measurements of these specimens do
not unambiguously support this conclusion, as
they fall within the range of overlap between
this species and co-occurring unisexuals (Lowcock, 1986). A single 3n male hybrid from HBP
was correctly classified on initial examination in
the field.
LOWCOCK ET AL.-PLOIDY
COB
and
altitude
Increasing
HBP
ORO
93
latitude
FI
5n
2n Hybrid
IN HYBRID AMBYSTOMA
AL
AH
9
3ncid
/
/
Other
4n
3n
2nY
LLi
Day
14
Cumulative
Fig. 3. Relative sex and ploidy composition of Day 1 breeding cohorts examined in central Ontario. Site
abbreviationsas in text. At HBP, the Day 1 (4 April) total is compared to the cumulativetotal (4-28 April).
Absolute values are listed in Tables 4-5.
BREEDING DYNAMICS AT
HBP
Ambystoma laterale (diploids) vs polyploid hybrids.--Significant differences in the ratios of
diploid to polyploid individuals were observed
among the four waves (x2, 3 df = 16.65, P =
0.001; G, 3 df = 7.81). This is attributable to a
very high percentage of polyploid individuals
in the second wave (91.3%). The remaining
waves had percentages close to 80 (Table 7).
Male A. laterale vs female A. laterale.-No dependence was observed between waves and sex
ratios for diploid individuals (x2, 3 df = 2.12, P
= 0.549; G, 3 df = 1.60).
Male A. laterale vsfemales (all ploidies).-No dependence between sex and wave was observed
for the population as a whole (x2, 3 df = 3.96,
P = 0.266; G, 3 df = 5.8).
Triploidvs tetraploidfemales.-Only the first three
waves were included in this test because hybrid
ploidies were not determined for the final wave.
Slight significance was observed (x2, 2 df = 5.684,
P = 0.058; G, 2 df = 6.20). This is attributable
to the higher proportion of triploids in the first
wave (93.3%). Second and third wave polyploids
were 86.5% and 86.1% triploid, respectively.
Comparisons among individual days within
waves (waves 1-3) indicated some intra-wave
heterogeneity of sex and ploidy composition.
Percentages of sexes and ploidies, and relevent
G-scores are provided in Table 7.
Ambystoma laterale (diploids) vs polyploid hybrids.-Although significant variation in the ratios of diploids to polyploid hybrids was observed among waves, within waves the ratios did
not not vary significantly.
94
COPEIA, 1991, NO. 1
20k
2n
()
o
"-
4n
10k
5n
E
D
0
76
149
114
188
250
of fluorescence
Intensity
lateraleand 3n,
of
nuclear
DNA content for 2n Ambystoma
4.
measurements
Separateflow-cytometric
Fig.
4n and 5n hybrid biotypes of A. laterale-jeffersonianum
depicted by an overlay of single parameterhistograms
representingthe integral of red fluorescence. Channel numbers on the X-axis correspondto the position of
the peak for each histogram.
Male A. laterale vsfemale A. laterale.-Significant
differences in diploid male:female ratios did
not occur among days of the first wave, but
within the second and third waves the percentage of diploid females increased significantly
with passing days, reaching 100% in the final
day of each wave.
Male A. laterale vs females (all ploidies).-No significant variation was found in the overall male :
female ratio among days of the first wave. However, within the second and third waves a significant increase in the numbers of females
across days was observed.
there
Triploid vs tetraploidfemales.-Although
was significant variation in the relative numbers
of triploid and tetraploid females across waves,
TABLE
4.
DISTRIBUTION
within waves significant changes in the composition of polyploids did not occur.
DISCUSSION
Ploidy levels and ratios.-The ratio of A. laterale
to hybrids varied between samples but was consistent with the range of variation previously
reported for the A. laterale-jeffersonianumcomplex (Clanton, 1934; Uzzell, 1964; Lowcock,
1989). Unisexual hybrids, dominated by triploids (LLJ; where L = one genome of A. laterale
and J = one genome of A. jeffersonianum), greatly outnumber diploid bisexuals, except in the
periphery of the complex, where the extent of
unisexuality fluctuates widely between populations. Diploid bisexuals eventually predominate
near the distributional limits of hybrid biotypes
OF SEX AND PLOIDY CLASSES IN FIRST IMMIGRATIONS
SITES AS DETERMINED
A. laterale
Site
n
CP
OP
FIP
AL
AHP
95
74
50
54
42
a
m"
11.6/61.1
12.2/45.0
14.0/70.0
18.5/27.7
14.2/25.0
ONTARIO
Hybridsc
f
7.4/38.9/8.3
14.9/55.0/16.9
6.0/30.0/6.9
48.1/72.3/59.1
42.9/75.0/50.0
Percent of sample/% of A. laterale.
b Percent of sample/% of A. laterale/% of females.
c Percent of sample/% of hybrids/% of females.
AT FIVE CENTRAL
BY FLOW CYTOMETRY.
2n
2.7/3.7/3.1
-
-
3n
78.9/97.4/89.2
70.0/96.3/80.0
80.0/100/93.1
29.6/88.9/36.4
42.9/100/50.0
4n
2.1/2.6/2.4
-
3.8/11.1/4.5
LOWCOCKET AL.-PLOIDY IN HYBRID AMBYSTOMA
TABLE 5.
DISTRIBUTION OF SEX AND PLOIDY CLASSESIN DAILY IMMIGRATIONSAT HBP IN 1988 As DETERMINED
BY FLOW CYTOMETRY(4-14 APRIL) AND A PRIORI ASSIGNMENT (15-28 APRIL).
A. laterale
Date
n
4*
96
5*
62
6*
55
7*
225
8*
61
Hybrids"
ma
fa
6.3/37.5
10.4/62.5/11.1b
76.00/91.3/81.1
12.9/66.7/13.8b
74.2/92.0/79.3
(n = 6)
(n = 10)
6.5/33.3
(n = 4)
3n
(n = 73)
4n
(n = 7)
6.5/8.0/6.9
(n = 8)
(n = 46)
(n = 4)
9.1/71.4
3.6/28.6/4.0
85.5/97.9/94.0
1.8/2.1/2.0
(n = 5)
4.0/42.9
(n = 9)
(n = 2)
5.3/57.1/5.6
(n = 12)
(n = 47)
78.7/86.8/81.9
(n = 177)
(n = 1)
11.6/12.7/12.0
(n = 26)
-
6.6/100/6.6
77.0/82.5/77.0
14.8/15.7/14.8
(n = 4)
-
(n = 47)
100
(n = 9)
-
1
-
11
6
50.0/75.0
16.7/25.0/33.3
16.7/50.0/33.3
(n = 3)
16.6/50.0/33.3
(n = 1)
(n = 1)
(n = 1)
65.9/85.3/76.3
11.4/14.7/13.2
44
13.6/60.0
13*
85
14*
75
4-14
710
9.1/40.0/10.5
(n = 4)
(n = 6)
7.1/33.3
14.1/66.7/15.2
(n = 29)
(n = 5)
68.2/86.6/73.4
10.6/13.4/11.4
(n = 6)
(n = 12)
(n = 58)
(n = 9)
1.3/8.3
14.7/96.7/15.1
72.0/85.7/74.0
10.7/12.7/10.9
(n = 1)
(n = 11)
(n = 54)
(n = 8)
9.0/61.5/9.6
(n = 64)
75.1/88.0/79.7
(n = 533)
9.9/11.6/10.5
(n = 70)
5.6/38.5
(n = 40)
Other
7.3/8.8/7.7
9
12*
95
0.44/0.49/0.46c
(n = 1)
1.6/1.8/1.6c
(n = 1)
-
1.3/1.6/NAd
(n = 1)
0.28/0.33/0.30c
(n = 2)
0.14/0.16/NAd
(n = 1)
14.6 (n = 104)
% of population
by category
m
15*
18*
25
45
21
5
24
12
25
26
27
28
2
1
1
3
4-28
f
804
% of population
by category
6.7/30.0
Undifferentiated hybrids'
47.1/100/47.1
52.9/52.9
(n = 8)
(n = 17;2D
15.5/70.0/16.7
77.8/83.3
(n = 3)
(n = 7)
(n = 35;4)
-
20.0/100/20.0
(n = 1)
25.0/100/25.0
80.0/80.0
(n = 4)
75.0/75.0
(n = 3)
100
-
(n = 9;19
100
100
-
-
5.3/33.9
(n = 43)
10.4/66.1/11.1
(n = 84)
15.7 (n = 127)
*
Significant migrations (n > 20).
Percentages arranged as in Table 4.
b Includes one
polyploid female (LLL or LLLL).
c 5n female.
d 3n male.
e Percent of sample/% of females.
I Estimated number of 4n females in
sample.
a
85.4 (n = 606)
100
84.3/100/88.9
(n = 677)
84.3 (n = 677)
96
TABLE
COPEIA, 1991, NO. 1
6.
PERCENT
ACCURACY
PREDICTION
OF A PRIORI ASSIGNMENTS
ON RESULTS
n
Site
Initial
WITH RESPECT TO HYBRIDITY
CP
OP
74
98.6
50
54
42
710
100
100
100
99.6
3 2n laterale/2 x 3n; 1 x 4n
98.6-100
( = 99.7)
1 polyploid hybrid
and
2n hybrid
3 2n laterale
unknown
FIP
AL
AHP
HBP
Overall
(n = 1025)
Final
Reevaluated
Misclassified/determineda
-
95
100
BASED
OF FLOW CYTOMETRY.
100
1 polyploid hybrid/2n
2n hybridb
100
3n/4n laterale?c
100
100
100
99.6
99.6-100
( = 99.98)
a Our
classification/ploidy determination by flow cytometry.
b Hybrid by morphology, misclassified because all
hybrids were assumed to be polyploid.
c Could not be classified
unambiguously as A. laterale.
(Lowcock, 1989). This is exemplified in our
study by the two most northern samples (AL,
AHP), which are close to these limits, and show
a significantly higher proportion of diploid bisexuals in initial breeding waves (Fig. 3). Based
on data from HBP, we would expect these ratios
to hold over the course of the breeding season.
To what extent distributional limits of hybrid
biotypes are governed by geological, ecological,
historical or physiological factors is unknown.
Although cytogenetic tendencies within populations may play a limited role, it has been arTABLE 7.
PERCENT COMPOSITION
gued that several of the former variables, and
their inherent interactions, form the major
components of limitation (Lowcock, 1989).
First immigration results for AL and AHP
could be interpreted as artifactual if diploid bisexuals breed much earlier than unisexuals in
northern areas. However, such a regimen would
serve only to lower the likelihood of a unisexual
being mated, ultimately providing another
mechanism by which increases in the unisexual
component of a population could be intrinsically controlled, resulting in the same relative
OF DIPLOID AND POLYPLOID
FEMALES AT HBP,
BY DAY AND BY MIGRATION
WAVE.Numbers in parentheses indicate sample sizes.
Day
Percent polyploid
of population
Wave
G-score/df
1
1
2
0.96/2
0.52/1
83.3 (96)
90.7 (225)
80.7 (62)
93.4 (61)
87.3 (55)
-
3
2.92/3
77.3 (25)
78.8 (44)
84.0 (85)
4
Percent female of
diploids
Percent female of
population
Percent triploid
of polyploids
-
1
2
3
4
12.82/3b
-
1
2
3
4
1
2
3
2
77.8 (45)
62.5 (16)
57.1 (21)
40.0(10)
66.7 (9)
66.7 (12)
100.0 (4)
66.7 (18)
-
4.40/1b
8.38/3b
-
93.8
96.0
86.4
93.2
(96)
(225)
(25)
(43)
93.6 (62)
100.0 (61)
92.9 (44)
2.84/2
0.40/1
0.10/2
91.3 (80)
87.2 (203)
85.3 (34)
92.0 (50)
83.9 (56)
86.5 (67)
3.04/2
3.98/1b
0.48/2
2
Significant differences in percent among waves.
b Significant differences in percent within waves.
-
3
-28.6 (7)
-
4
-
68.0 (75)
-
Total for wave
83.6 (213)a
91.3 (286)
79.0 (229)
77.8 (45)
91.7(12)
-
100.0 (8)
-
57.1
64.0
72.9
66.7
(35)
(25)
(48)
(9)
90.0 (55)
97.3 (85)
-
100.0 (75)
-
93.0
96.9
93.9
93.2
(213)
(286)
(229)
(43)
97.9 (48)
87.1 (62)
-
93.0 (213)a
86.5 (259)
86.1 (163)
LOWCOCK ET AL.-PLOIDY
counterbalance of diploid bisexuals that were
observed.
Variable ploidy in unisexual A. laterale-jeffersonianum has only been considered recently
(Sessions, 1982; Bogart, 1982, 1989) and its extent has not been quantified. Our results show
that the most common deviation from triploidy
in central Ontario hybrids is an elevation to the
tetraploid state (LLLJ). Tetraploids were found
at three of six sites, and probably occur at lower
frequencies in the others. The typical route to
tetraploidy in hybrid Ambystomais fertilization
of an unreduced triploid egg (Bogart and Licht,
1986; Bogart et al., 1987). Thus, when hybrids
enter a population, the initial preponderance
of tetraploidy likely reflects the interaction of:
1) the frequency at which unreduced eggs are
produced by triploids; and 2) the probability of
fertilization. Once the breeding aggregation includes tetraploid females, the recruitment dynamic of these forms will change depending on
what type of eggs/offspring they produce. Because the known progeny of allotetraploid females of the analagous A. laterale-texanumcomplex are triploid and tetraploid (Bogart and
Licht, 1986; Bogart et al., 1987), this is likely
also true of the A. laterale-jeffersonianumcomplex. Thus, under certain circumstances, tetraploid genotypes may accumulate in a population. Bogart et al. (1989) recently found that
the temperature at which eggs are exposed to
sperm dictates the relative frequency of gynogenetic, hybridogenetic and ploidy-elevated offspring produced by a particular female. This
seriously confounds the issue, e.g., conditions
favorable to the de novo synthesis of tetraploids
from triploid eggs may not be the same as those
conducive to gynogenetic activation of unreduced tetraploid eggs. The situation begs greater scrutiny, because other important considerations, such as reproductive
viability of
females
and
tetraploid
potential ecological role
for tetraploid biotypes, are not understood.
Tetraploids produce at least some viable unreduced ova, as evidenced by the low frequency
of pentaploid female hybrids (LLLLJ) at HBP.
Pentaploidy was not anticipated because: 1) it
has not been revealed in previous genetic studies of hybrid complexes of Ambystoma;2) mortality of 3n and 4n eggs and larvae is exceedingly high (Bogart et al., 1987; Licht and Bogart,
1987; Licht, 1989); 3) in autopolyploid Ambystoma, fecundity decreases as ploidy increases
(Humphrey and Fankhauser, 1956); and 4) potential tetraploid progenitors typically repre-
IN HYBRID AMBYSTOMA
97
sent a low percentage of hybrids (Bogart, 1989;
Bogart et al., 1985, 1987).
This first recorded instance of naturally occurring pentaploid vertebrates is important from
two standpoints. First, all experimentally induced and lab-raised pentaploid urodeles have
been autopolyploids of greatly reduced viability, suffering severe developmental, physiological and behavioral debilities that precluded the
attainment of sexual maturity (Fankhauser,
1945; Fankhauser and Humphrey, 1950). Haliburton Beaver Pond pentaploids were clearly
hybrids, visibly swollen with eggs, some of which
were ovulated, had no noticeable deformities,
and possessed body sizes indicating ages of 3 yr
or more. Thus, compared to autopentaploids,
these allopentaploid salamanders demonstrate
an enhanced tolerance of multiple chromosome
sets. Secondly, autopentaploids were commonly
generated by fertilization of a tetraploid egg
produced by a diploid parent (suppression of
both meiotic divisions; Fankhauser, 1945). Our
pentaploids could not have been produced via
this route; HBP pentaploids are hybrids, and
potential progenitors (diploid hybrid genotypes) do not occur here. Pentaploid A. lateralejeffersonianum at HBP are presumably generated by the fertilization of an unreduced
tetraploid ovum produced by a hybrid tetraploid female, in analogy to the production of
tetraploids by triploids. Test matings between
HBP tetraploids and male A. texanum yielded
numerous pentaploids, confirming the scenario
of origin for pentaploid biotypes at HBP (J. P.
Bogart, pers. comm.).
It may be assumed, as with triploids and tetraploids, that there is a greatly reduced viability
of pentaploid embryos and larvae, and that those
surviving to adulthood are a highly selected
fraction of those produced. Thus, the rarity of
pentaploids at HBP is in accordance with the
proportion of available tetraploid progenitors.
Based on this, the apparent scarcity of tetraploid biotypes in other populations lessens the
probability of pentaploids occurring there. Potential production of pentaploids by the union
of polyploid sperm with polyploid eggs cannot
be ruled out. The most likely source of such
sperm, however, is triploid males, which are rare
and of unknown viability.
Diploid hybrids (LJ) are perhaps our most
perplexing finding with regards to hybridity.
These biotypes were recorded only at OP, where
they have been reported previously (Lowcock,
1989). If present at other sites, the frequency
98
COPEIA, 1991, NO. 1
must be extremely low. The large sample from
HBP indicates diploid hybrids do not occur there
at all. Diploid hybrid A. laterale-jeffersonianum
are known from active zones of hybridization
in Ontario and New England (Bogart, 1989;
Lowcock, 1989), where the opportunity for de
novo synthesis exists because of the presence of
both parental bisexuals. The origin and status
of diploid hybrids outside hybrid zones is currently unknown. It is typically assumed that diploid hybrid female Ambystomaproduce unreduced diploid eggs, giving rise to triploids
through backcrosses to males of either parental
species. Under this assumption, diploid hybrids
are not self-perpetuating unless their unreduced eggs are activated gynogenetically, a possibility that is unresolved. If LJ hybrids at OP
produce some meiotically reduced eggs, in addition to unreduced hybrid ova, then they could
give rise to three types of progeny through syngamy with L sperm: LL, LJ and LLJ. It is also
possible to generate a diploid hybrid from a
triploid, either by syngamy of L sperm with a
rare haploid J egg, or the gynogenetic activation of a diploid hybrid meiotic product. Reduction mechanisms for producing diploids from
triploids have been speculated elsewhere (Sessions, 1982; Bogart, 1989; Lowcock, 1989) but
remain unsubstantiated. If J eggs are occasionally produced by triploids, and assuming equal
frequency and viability of the reciprocal meiotic
product (i.e., LL), we would also expect to find
some incidence of triploid A. laterale (LLL) in
the population. These biotypes did not occur
at OP, but may have been present at HBP, along
with autotetraploids (LLLL); three females classified originally as diploid A. laterale on morphological grounds were shown to be polyploid
by FCM (Table 6). This is not unusual, as triploid A. texanum and A. laterale have been encountered in A. laterale-texanum complex populations on Pelee Island (Bogart and Licht, 1986;
Licht, 1989; Sever et al., 1989). Despite this
potential connection, mutual exclusion of LJ
and LLL biotypes at OP and HBP suggests that
these forms may also be generated autonomously. Erroneous a priori classification of the
individuals at HBP is possible, but seems unlikely given prediction accuracies (Table 6).
Triploid males have been reported on several
occasions, and appear to be a cosmopolitan, but
rare, feature of hybrid complexes of Ambystoma
(Clanton, 1934; Servage, 1979; Morris and
Brandon, 1984). Clanton (1934) reported three
out of 1300 individuals examined (0.23%), and
the single individual in our study represented
0.14% of the population and 0.16% of hybrids
at HBP (Table 5). Morris and Brandon (1984)
reported an individual that was sterile, but it
may have been a trihybrid (A. laterale-jeffersonianum-texanum)and/or a tetraploid. It is unlikely
that complete sterility is universal in these forms
(Fankhauser, 1945). Sperm smears from testes
of A. laterale-jeffersonianum complex triploid
males showed active, though occasionally deformed sperm (pers. obs. with J. P. Bogart).
Potential partial viability notwithstanding, the
rarity of triploid males precludes any meaningful contribution to the population, although test
matings should be conducted to evaluate reproductive capacities.
Sex ratio.-Overall sex ratios show males to be
a predictably limited resource (Tables 4-5, 8).
Males in first day immigrations comprised an
average of 12.8% of migrants, and were most
plentiful in populations with high ratios of diploids (AL, AHP). At HBP, where males were
most scarce, the figure fell from an initial 6.3%
to 5.3% over the entire breeding period (Table
5). This is expected given the breeding dynamic
of Ambystoma(see below). Corresponding total
(A. laterale plus polyploid) female: male sex ratios were congruent with previous reports obtained under controlled (i.e., fenced) conditions
(Table 8). Estimates prior to Uzzell (1964) are
difficult to reconcile with these observations,
and are suspect in that they represent variable
samples combined over periods of up to 5 yr,
some of which: 1) contain numerous newly
transformed individuals (in which sex determination is tenuous); 2) were generated from
unfenced or inadequately fenced ponds; or 3)
were not inventoried daily from the first day of
immigration.
In general, our figures, and those reported
by Uzzell (1964), Wilbur (1971) and Weller
(1980), show the male component of A. lateralejeffersonianum complex populations in Ontario
and southern Michigan to be greater than that
in the A. laterale-texanum complex. Bogart and
Licht (1986) and Licht (1989) found 5.6% males
in their randomly collected cumulative samples
from Pelee Island. Aside from potential collection artifact, the situation is confounded on Pelee (an active hybrid zone) by the presence of
A. laterale, A. texanum and diploid male A. laterale-texanum, in addition to a large component
of diploid hybrid females. Diploid male A. laterale-jeffersonianumare known from in and around
LOWCOCK ET AL.-PLOIDY
TABLE 8.
IN HYBRID AMBYSTOMA
99
COMPARISON BETWEENSEX AND PLOIDY RATIOS IN THE Ambystomalaterale-jeffersonianumCOMPLEX
AS REPORTED
Ploidy ratioa
3.75
1.56
3.21/6.07
18.00
11.00
4.90
13.33
0.69
1.00
8.06
4.69
IN THE LITERATURE
AND AS DETERMINED
Sex ratiob
Mean ratio (Weller)
Actual ratioe
Reference
Population is LL/LLJ based
3.28
6.71/14.16
8.40
7.63
7.22
6.14
4.40
6.00
17.69
6.23
7.00
3.59
13.00
2.14
51.25
52.00
Population is JJ/JJL based
8.00
73.00
76.80/100
16.33
17.00
11.37
9.06*
7.66*
23.00
15.03
13.62
13.67
12.66
BY THIS STUDY.
Uzzell (1964)
Wilbur (1971)
Wilbur (1971)c
Wilbur (1971)
this study (CP)
this study (OP)
this study (FIP)
this study (AL)
this study (AHP)
this study (HBP)
Wilbur (1971)
Uzzell (1964)
Uzzell (1964)
Uzzell (1964)
Clanton (1934)
Clanton (1934)
Clanton (1934)
Bishop (1941)
Peckhamand Dineen (1955)
Wacasey(1961)c
Weller (1980)
Weller (1980)
Weller (1980)d
1973; Weller (1980)
1974; Weller (1980)
1975; Weller (1980)
1976; Weller (1980)
1977; Weller (1980)
1973-77
females: non-hybrid females.
bSHybrid
Total females: males.
c One population in two separate years.
d Cumulative data in a 5 yr study; the ploidy ratio is from a sample collected randomly over this period. A breakdown of Weller's data
(below
solid line) yields slightly different figures than his own total. Weller suggests that the data from 1973-74 (*)
may not be reliable due to sampling
problems (e.g., dipnetting, partial fencing, mistiming).
LOur calculation
using total number of individuals in Weller (1980).
active hybrid zones of that complex in southern
Ontario and New England (Bogart, 1989), but
reproductive capacities and contributions of
these biotypes to their respective local populations have not been studied. Whether a lower
percentage of males on Pelee Island is an artifact of random collection in and around breeding ponds, or is directly related to the inherent
genotypic structure and reproductive dynamics
of the A. laterale-texanumcomplex, is unknown.
Cryptic behavior by breeding A. laterale-texanum complex males, and difficult collecting con-
ditions in breeding ponds, have been cited as
factors in biased samples (Bogart and Licht,
1986; Bogart et al., 1987). Uncontrolled sampling (in or outside of the breeding season) in
these complexes can result in samples which are
biased in favor of unisexuals (e.g., Uzzell, 1964;
Downs, 1978; Kraus, 1985a, 1985b; and pers.
obs. of collections at the University of Guelph
[Guelph, Ontario], Royal Ontario Museum [Toronto, Ontario], and National Museum of Natural Sciences [Ottawa, Ontario]). Considering
bisexual-unisexual population structure, and if
100
COPEIA, 1991, NO. 1
all parameters of ecological distribution are
equal, the probability of encountering unisexuals in the field can be an order of magnitude
higher than that for diploid bisexuals. However, ecological partitioning of biotypes is apparently operative in some hybrid complexes (Bogart et al., 1987), and this could greatly affect
encounter probabilities.
The geographically variable male component
will influence local evolutionary trajectories
within hybrid complexes via reproductive output, especially if fluctuating ecological factors
affect male recruitment temporally. In this context of intrinsic determinism, no perspective can
be gained unless sex ratios within A. laterale are
examined as a subset of the total female: male
sex ratio in populations. Our results unexpectedly show that this relationship may also vary
geographically (Table 4; Fig. 3). In three populations (AL, AHP, HBP), ratios differ significantly from the expected 1:1 (P < 0.001; X2
goodness of fit tests). In each case, this is due
to a much higher proportion of female A. laterale. This is confounding from several standpoints. First, although they may fluctuate from
year to year, sex ratios in other pond-breeding
species of Ambystomatypically show an overwhelming bias in favor of males (Shoop, 1960;
Semlitsch, 1983; Sexton et al., 1986). This is
also true for populations of A. laterale (Cook,
1967; Gilhen, 1984; Lowcock, 1986) and A. jeffersonianum (Uzzell, 1964; Douglas, 1979; Weller, 1980) that occur outside of the range of
hybrid biotypes. Skewed breeding sex ratios in
favor of male Ambystomahave been attributed
to: 1) differential mortality between sexes (higher male survival; Husting, 1965; Whitford and
Vinegar, 1966); 2) failure of females to breed
in consecutive years (Husting, 1965); or 3) delayed maturation of females (Wacasey, 1961;
Weller, 1980; Philips and Sexton, 1989). It seems
reasonable that the latter point may also be considered in the context of selection for precocious maturation of males.
Skewed diploid sex ratios favoring females
within hybrid complex populations could be related to forces other than those outlined. Given
their early breeding habits, higher male mortality is possible, but seems unlikely and has not
been observed in the field. A higher proportion
of diploid females breeding annually is possible,
especially if many go unmated because of the
large number of unisexuals; unmated diploid
females would presumably resorb their eggs,
not suffer the energetic costs of egg-laying, and
be more likely to breed in consecutive years. In
this case, we would expect populations containing low percentages of hybrids to show diploid
sex ratios in breeding aggregations closer to
equity or even biased towards males in early
immigrations. We did not observe this, although it was reported by Clanton (1934) and
fits with ratios reported by Cook and Gorham
(1979), Gilhen (1984) and Lowcock (1986) for
east-coast populations with few unisexuals.
Lower proportions of males breeding in every
year seems unlikely given the energy budget
difference between sexes, i.e., there would certainly be no advantage to this. Indeed, despite
the complex and irregular breeding schedule
of individuals, Weller (1980) found that a greater proportion of male A. jeffersonianumbred annually over a 5 yr period in a hybrid complex
population involving that species. Accelerated
maturation of females and/or delayed maturation of males is the final possibility based on
previous conjectures. There are no data available to support such a conclusion, and in fact,
body size data of breeding adults at HBP argue
for the opposite tenet, i.e., precocoius maturation of males.
Cytogenetic factors contributing to higher
proportions of diploid females may be more
convincing. The most obvious is a higher overall production of female offspring by female A.
laterale. Clanton (1934) states that although both
sexes in adult A. laterale were "equally represented" in populations containing hybrids, 60%
of progeny of diploid females were female. Uzzell (1964) presents similar data for juvenile A.
jeffersonianum (53.6-66.7% female; 9 = 61.6%).
If this varies among populations, it could explain our observation of higher numbers of females. A mechanism of meiotic drive involving
sex chromosomes need be invoked to consider
this possibility, and substantiation would require investigation at the cellular level. One can
envision selection favoring enhanced production of diploid females in populations with hybrid biotypes. Because they are the only route
to the production of males, these females serve
to maintain the "precious" diploid component,
which is the lynch-pin of bisexual-unisexual
communities. The possibility that some diploid
females are produced by unisexual hybrids has
been alluded to (Bogart, 1980; Lowcock, 1989;
Lowcock and Bogart, 1989), but appears to be
an event of extremely low frequency. Such a
conclusion, however, may be an artifact of those
methods used to determine its occurrence; allo-
LOWCOCKET AL.-PLOIDY IN HYBRID AMBYSTOMA
zyme electrophoresis cannot demonstrate the
production of diploids from unisexuals unless
there has been introgression at marker loci from
other species involved in the complex under
consideration (Lowcock, 1989). It is apparent
from other data presented here that such a
mechanism would be variable in occurrence,
frequency and success. The cytogenetic tendencies which have contributed to variable ploidy composition between populations may also
be operative in this case. Because sex ratios in
first day immigrants vary among populations,
and are biased overwhelmingly towards females
in only a few, such a mechanism seems likely.
It is of interest to note that Bogart et al. (1987)
reported a perfect 1:1 sex ratio for non-hybrids
in the A. laterale-texanum-tigrinumcomplex on
Kelleys Id., although the number of A. texanum
examined was small. Finally, Beneski et al. (1986)
counted up to seven times as many male emigrants as immigrants in their study of A. macrodactylum, concluding that they had missed a
component of initial migration, despite fencing
the pond before ice-off. It is unlikely that this
occurred at HBP, where the male component
was lowest; our fence was erected while the pond
was frozen to the margins, with an average ice
depth of 20 cm, snow cover of equal depth in
the bush at 100%, and 80% on south-facing
slopes. Males would have had to overwinter in
the pond in order not to be inventoried.
final considerBreeding dynamics at HBP.-A
ation, which also bears on proportions of females in first day immigrations, is breeding dynamics. Onset and duration of migrations were
directly related to environmental factors. Migratory movement was nocturnal and typically
limited to periods preceded by (rain) or during
(rain or snow) precipitation (Fig. 2). Wave 3
began on the evening of 11 April and built
through the 12th, despite no precipitation on
either day. These movements were triggered
by a high humidity gradient near the ground
surface at night, the result of localized evaporation/condensation caused by an ice-free pond
and strong sunlight during the day. Migrations
were often initiated despite partial snow cover
in an area, indicating some animals were in a
position to respond to environmental cues despite ground frost. It has been speculated (Weller, 1980) that the ground must be thawed before initiation of migration. Observations of
syntopic A. maculatum and Notophthalmus viridescens showed they followed this pattern,
101
reaching peak immigration several days later
than A. laterale-jeffersonianumcomplex individuals at northern sites, where ground frost and
snow cover were extensive.
Individuals of the A. laterale-jeffersonianum
complex are extremely cold tolerant. They frequently migrated across large expanses of snow
and ice and/or moved in great numbers at nearfreezing air temperatures into 1-3 C water. Individuals frozen in drop buckets with excess
moisture could thaw and walk away. Those frozen in dry conditions dessicated rapidly and died,
or suffered debilitating nerve damage, becoming partially incapacitated upon thawing. Ambystomalaterale, especially males, were more tolerant than hybrids. That immigration is readily
initiated under such conditions, coupled with
the subarctic range limits of this species, argues
convincingly for an evolutionary history tied to
a boreal or sub-boreal climatic regimen, possibly in association with the proglacial environments of an ice age. Pleistocene events have
often been evoked for the splitting of A. laterale
from its sister species A. jeffersonianum (Uzzell,
1964); however, Lowcock (1989) has presented
genetic data that argues for a more distant split,
perhaps associated with pre-Pleistocene ice ages.
Greater tolerance of harsh conditions by male
Ambystomaappears to be a general attribute of
the genus. Males typically enter ponds earlier
and leave later than females, remaining in the
area up to twice as long as their female counterparts (Shoop, 1960; Weller, 1980; Sexton et
al., 1986). This has been variously attributed
to: 1) differential rate and/or distance of travel;
2) differential response to environmental cues;
and 3) differential orientation ability. Philips
and Sexton (1989) discounted the latter in their
study of A. maculatum, but were unable to substantiate whether the first two were operative.
Douglas (1979) maintained that the selective
advantage of females lagging behind males is
reduction of risk of environmental danger, particularly low temperatures, a selective force
which would be exacerbated in the case at hand.
Based on this, one can construct an idealized
schemata for immigration/emigration in pondbreeding Ambystoma,altering it to reflect the
observations made in those studies, i.e., that
females are represented in early samples but
that males, typically more abundant, reach peak
proportion in the breeding population earlier
than females (Fig. 5).
Assuming this dynamic to be operative among
diploid bisexuals within hybrid complex popu-
COPEIA, 1991, NO. 1
102
100
A
A
Immigration
3n
..
Emigration
B
75
"C,
100
C
t-
50-0
D
o
female
5n
3n male
50
25-
9
50-~
4n
,
-
1 2
3 4
Diploid
5 6
7 8
immigration
9101112131415
days
Fig. 5. Breeding dynamicsin pond-breedingAmbystoma.In each case, the X-axis is a time function.
From the top of the figure: A) idealized breeding
regimen for males and females based on the assumptions that 1) malesoutnumberfemales, 2) malesenter
the pond earlier, stay longer and depart later than
females,and 3) there are no extenuatingcircumstances that would meliorate the selective forces driving
this system. B) Likely shape of immigrationcurves in
reality. Curvesare normalizedto show relative numbers of each sex for ease of direct comparison.Males
and females begin migrating at the same time, but
more males immigrate early so that this sex reaches
its peak first. Stars indicate expected onset of emigration for females (black)and males (white).C) Trajectories for cumulativetotals over time based on the
curves in (B). Stars as in (B). D) Trajectories for cumulative totals of immigrants as measured at HBP
showing conformity to model in (C). In this case the
starsshow actual dates of commencementof emigration for each sex.
1 2345
00
6 7 8 9 10
a
Days
Fig. 6. Relative percentages of polyploid hybrids
during significant migration-days(n > 20) at HBP
(daysas in Table 3; Day 10 = 18 April).
lations (Uzzell, 1964; Weller, 1980), we compared the data from HBP to such a regimen
(Fig. 5). No clear linear relationships existed
between proportions of sexes and breeding
times; however, differential achievement of peak
proportions by male and female A. laterale occurred; approx. 25% of females immigrated after the peak proportion of males (April 14).
Major waves (2-3) appeared to comprise separate episodic breeding events; within a wave
over time, generally showing a: 1) slow increase
and rapid decrease of males; 2) rapid increase
then rapid decrease in females; and 3) corresponding increase in proportions of polyploids
(Fig. 2; Table 5). Hybrids may be slower to respond to environmental cues, though the increase in relative proportions of these forms is
LOWCOCK ET AL.-PLOIDY
most readily explained by the decrease in diploids. Assuming no mate discrimination by
males, polyploids will theoretically do well in
the competition for sperm due to sheer numbers, despite the fact that they reach peak proportions after all diploids within waves. For selection to favor hybrid biotypes, the majority
would have to arrive within the period at which
the male proportion of the population peaked.
The two initial waves were characterized by
higher proportions of hybrids, representing up
to 96% of the female component on a given day
(Table 5). Overall, ploidy ratios of hybrids did
not vary significantly during the immigration
period, showing only a modest increase in the
tetraploid component over day 1 (Table 3; Fig.
3). During wave 2, however, a significant increase in the proportion of tetraploids occurred. The capture of pentaploid females was
coincident with the increase in hybrid biotypes
of elevated ploidy. This observation is problematic: are higher polyploids harder to stimulate
into migration than triploids, or are they behaving like diploid females, to which they are
more similar by virtue of extra genomes from
A. laterale? From the onset of wave 2, polyploid
ratios remained relatively constant (Fig. 6).
The issue of whether hybrid biotypes arrive
at ponds ahead of diploid females has not been
studied.The possibility has been raised (Panek,
1978; Weller, 1980; Lowcock, 1986), and the
available data are suggestive. The time necessary for the proportion of males and females
arriving at the pond to approach equity is greater in pure populations (Douglas, 1979) than in
mixed diploid/triploid aggregations (Weller,
1980). Thus, the abrogation in lag time may be
caused by an influx of hybrid females. Unfortunately, previous data suffer from a lack of
differentiation between ploidy levels (and hybridity) of females, and as such, are inconclusive. Results of this study, however, in which all
females were distinguished with respect to hybridity, lends support to this interpretation.
Conclusions.-Understanding
of the evolution
and ecology of bisexual-unisexual communities
of vertebrates is greatly aided by non-destructive large-scale studies of composition and interaction. Such investigations can ultimately inform and be informed by the vast repository of
genetic information available from these populations. Our several unexpected findings with
regard to ploidy structure, sex ratio and breeding dynamics, suggest that in addition to pop-
IN HYBRID AMBYSTOMA
103
ulation differences, these parameters may all
vary temporally, and that long-term studies of
breeding and transforming cohorts would be
productive means of evaluation. What emerges
from our study is clear evidence that unisexuals
affect population structure and breeding dynamics in syntopic bisexuals. We would expect
that this would obtain for other ecological considerations in these associations.
ACKNOWLEDGMENTS
We are grateful for the extensive field and
laboratory assistance of C. Garland and M. Mosquito. A great debt of gratitude is extended to
C. Smith, who operated the University of Toronto Department of Immunology flow cytometer for incalculable periods and at odd hours.
We are indebted to R. M. Dawley for his help
with FCM technique. R. Elinson and J. P. Bogart carried out the tetraploid-A. texanum cross.
C. Rutland compiled the bibliography. The
study was supported by Natural Sciences and
Engineering Research Council of Canada grant
A3148 to RWM, and some field work was supported by a Friends of Algonquin Park Research grant to LAL.
LITERATURE CITED
AUSTIN, N., AND J. P. BOGART. 1982. Erythrocyte
area and ploidy determination in the salamanders
of the Ambystoma
jeffersonianumcomplex. Copeia
1982:485-488.
BENESKI,
J. T., JR., E.J. ZALISKOANDJ. H. LARSEN,
JR. 1986. Demographyand migratorypatterns of
the eastern long-toed salamander, Ambystomamacro-
Ibid. 1986:398-408.
dactylumcolumbianum.
BOGART, J. P. 1980. Evolutionary implications of
polyploidyin amphibiansand reptiles, p. 341-378.
In: Polyploidy:biological relevance. W. H. Lewis
(ed.). Plenum Publ. Corp., New York, New York.
. 1982. Ploidyand genetic diversityin Ontario
salamandersof the Ambystoma
comjeffersonianum
plex revealed through an electrophoretic examination of larvae. Can.J. Zool. 60:848-855.
. 1989. A mechanism for interspecific gene
-exchangevia all-femalesalamanderhybrids,p. 170179. In: The evolution and ecology of unisexual
vertebrates.R. DawleyandJ. P. Bogart (eds.). New
York State Museum Bulletin, 466 Albany, New
York.
--,
R. P. ELINSON AND L. E. LICHT. 1989. Tem-
peratureand sperm incorporationin polyploidsalamanders.Science 246:1032-1034.
-, AND L. E. LICHT. 1986. Reproduction and
the origin of polyploids in hybrid salamandersof
104
COPEIA, 1991, NO. 1
the genus Ambystoma.Can. J. Genet. Cytol. 28:605northwestern Ohio and southeastern Michigan: a
617.
study of the consequences of hybridization. Copeia
S
, M. J. OLDHAMANDS.J. DARBYSHIRE. 1985:309-324.
1985. Electrophoretic identification of Ambystoma LICHT, L. E. 1989. Reproductive parameters of unilaterale and Ambystomatexanum as well as their dipsexual Ambystomaon Pelee Island, Ontario, p. 209loid and triploid interspecific hybrids (Amphibia:
217. In: The evolution and ecology of unisexual
vertebrates. R. Dawley and J. P. Bogart (eds.). New
Caudata) on Pelee Island, Ontario. Can.J. Zool. 63:
340-347.
York State Museum Bulletin, 466, Albany, New
L. A. LoWCOCK,
C. W. ZEYLANDB. K. MABLE.
York.
-~,
1987. Genome constitution and reproductive bi, AND J. P. BOGART. 1987. Ploidy and develology of hybrid salamanders, genus Ambystoma,on
opmental rate in a salamander hybrid complex (genus Ambystoma).Evolution 41:918-920.
Kelleys Island in Lake Erie. Ibid. 65:2188-2201.
W. 1934. An unusual situation in the sal- LOwcocK, L. A. 1986. Genetics, zoogeography, and
CLANTON,
amander Ambystomajeffersonianum
(Green). Occ. Pap.
morphological aspects of peripheral populations of
Mus. Zoo. Univ. Michigan 290:1-14.
the Ambystoma
jeffersonianumcomplex. Unpubl. M.S.
COOK,F. R. 1967. An analysis of the herpetofauna
thesis, University of Guelph, Guelph, Ontario.
of Prince Edward Island. Nat. Mus. Can. Bull. 212:
. 1989. Recent and palaeobiogeography in hy1-60.
brid complexes of Ambystoma:interpreting unisex, AND S. W. GORHAM.1979. The occurrence
ual-bisexual data through time and space, p. 180of the triploid form in populations of the Blue209. In: The evolution and ecology of unisexual
vertebrates. R. Dawley and J. P. Bogart (eds.). New
spotted Salamander, Ambystoma laterale, in New
Brunswick. J. New Brunswick Mus. 1979:154-161.
York State Museum Bulletin, 466, Albany, New
DAWLEY, R. M., ANDJ. P. BOGART (Eds.). 1989. EvoYork.
lution and ecology of unisexual vertebrates. New
, ANDJ. P. BOGART. 1989. Electrophoretic evYork State Museum Bulletin, 466, Albany, New
idence for the multiple origins of triploidy in the
York.
Ambystomajeffersonianumcomplex. Can. J. Zool. 67:
1988. Diploid-triploid
, AND K. GODDARD.
350-356.
mosaics among unisexual hybrids of the minnows
-, L. E. LICHTANDJ. P. BOGART. 1987. NoPhoxinus eos and Phoxinus neogaeus. Evolution 42:
menclature in hybrid complexes of Ambystoma:no
649-659.
case for the erection of hybrid "species." Syst. Zool.
DOUGLAS,M. E. 1979. Migration and sexual selection
36:328-336.
in Ambystomajeffersonianum.Can. J. Zool. 57:2303MORRIS,M. A., AND R. A. BRANDON. 1984. Gyno2310.
genesis and hybridization between Ambystomaplatineum and Ambystomatexanum in Illinois. Copeia
DowNs, F. L. 1978. Unisexual Ambystomafrom the
Bass Islands of Lake Erie. Occ. Pap. Mus. Zoo. Univ.
1984:324-337.
Michigan 685:1-36.
PANEK,F. M. 1978. A developmental study of AmG. 1945. The effects of changes in
FANKHAUSER,
bystomajeffersonianum and A. platineum (Amphibia,
chromosome number on amphibian development.
Urodela, Ambystomatidae). J. Herp. 12:265-266.
Quart. Rev. Biol. 20:20-78.
PHILIPS, C. A., AND O.J. SEXTON. 1989. Orientation
FANKHAUSER,
G., AND R. R. HUMPHREY.1950. Chroand sexual differences during breeding migrations
mosome number and development of progeny of
of the spotted salamander Ambystomamaculatum.
triploid axolotl females mated with diploid males.
Copeia 1989:17-22.
J. Exp. Zool. 115:207-250.
SEMLITSCH, R. D. 1983. Structure and dynamics of
GIBBONS, J. W., AND R. D. SEMLITSCH. 1982. Tertwo breeding populations of the eastern tiger salrestrial drift fences with pitfall traps: an effective
amander, Ambystomatigrinum. Ibid. 1983:608-616.
technique for quantitative sampling of animal popSERVAGE,D. L. 1979. Biosystematics of the Ambysulations. Brimleyana 7:1-16.
toma jeffersonianum complex in Ontario. Unpubl.
GILHEN, J. 1984. Amphibiansand reptiles of Nova
M.S. thesis, Univ. Guelph. Guelph, Ontario, CanScotia. Nova Scotia Museum Publ., Halifax, Nova
ada.
Scotia.
S. K. 1982. Cytogenetics of diploid and
SESSIONS,
HUMPHREY, R. R., AND G. FANKHAUSER.
1956. Structriploid salamanders of the Ambystomajeffersonianum
ture and functional capacity of the ovaries of higher
complex. Chromosoma 84:599-621.
polyploids (4N, 5N) in the Mexican axolotl (Siredon SEVER, D. M., L. E. LICHT AND J. P. BOGART. 1989.
or Ambystomamexicanum).J. Morph. 98:161-197.
Male cloacal anatomy in a hybrid population of AmHUSTING, E. L. 1965. Survival and breeding strucbystoma (Amphibia: Caudata). Herpetologica 45:
ture in a population of Ambystomamaculatum. Cope161-167.
ia 1965:352-362.
SEXTON, O. J., J. BIZER, D. C. GAYOU, P. FREILING
KRAUS, F. 1985a. A new unisexual salamander from
ANDM. MOUTSEOUS.1986. Field studies of breedOhio. Occ. Pap. Mus. Zoo. Univ. Michigan 709:1ing spotted salamanders, Ambystomamaculatum, in
24.
eastern Missouri, USA. Milwaukee Public Museum
S1985b. Unisexual salamander lineages in
Contrib. Biol. Geol. 67:1-19.
LOWCOCK ET AL.-PLOIDY
SHOOP,C. R. 1960. The breeding habits of the mole
salamanders, Ambystomatalpoideum (Holbrook), in
southeastern Louisiana. Tulane Stud. Zool. 8:6582.
SOKAL, R. R., AND F.J. ROHLF. 1981. Biometry.2nd
ed. W. H. Freeman & Company, San Francisco,
California.
TANK, P. W., R. K. CHARLTON AND E. R. BURNS.
1987. Flow cytometry analysis of ploidy in the axolotl, Ambystomamexicanum.J. Exp. Zool. 243:423433.
UZZELL, T. M. 1964. Relations of the diploid and
comtriploid species of the Ambystomajeffersonianum
plex (Amphibia, Caudata). Copeia 1964:257-300.
WACASEY, J. W. 1961. An ecological study of two
sympatric species of salamanders, Ambystomamaculatum and Ambystomajeffersonianum, in Southern
Michigan. Unpubl. Ph.D. dissert., Michigan State
University.
WELLER, W. F. 1980. Migration of the salamanders
Ambystomajeffersonianum (Green) and A. platineum
Cope to and from a spring breeding pond, and the
growth, development and metamorphosis of their
IN HYBRID AMBYSTOMA
105
young. Unpubl. M.S. thesis, University of Toronto,
Toronto, Ontario, Canada.
WHITFORD, W. G., AND A. VINEGAR. 1966. Homing,
survivorship, and overwintering of larvae in spotted
salamanders (Ambystomamaculatum). Copeia 1966:
515-519.
WILBUR,H. M. 1971. The ecological relationship of
the salamander Ambystomalaterale to its all-female,
gynogenetic associate. Evolution 25:168-179.
ZEYL,C. W., AND L. A. LOWCOCK.1989. A morphometric analysis of hybrid salamanders (genus
Ambystoma)and their progenitors on Kelleys Island
in Lake Erie. Can. J. Zool. 67:2376-2383.
DEPARTMENT
OFICHTHYOLOGY
ANDHERPETOLOGY,ROYALONTARIOMUSEUM,100 QUEEN'S
PARK,TORONTO,ONTARIO M5S 2C6, CANADAAND RAMSAYWRIGHTZOOLOGICAL
LAUNIVERSITYOF TORONTO 25
BORATORIES,
HARBORDST., TORONTO,ONTARIO,CANADA. Accepted 22 April 1990.
Copeia, 1991(1), pp. 105-110
The Influence of Photoperiod and Position of a Light
Source on Behavioral Thermoregulation in
Crotaphytuscollaris (Squamata: Iguanidae)
LYNNETTE M. SIEVERTAND VICTOR H. HUTCHISON
Temperature selection of male Crotaphytus collaris acclimated to 25 ? 1 C was
measured over a 24 h period in a thermal gradient with either uniform light
over the entire gradient, a point source of light over the hot end of the gradient,
or a point source of light over the cold end of the gradient and an 16L/8D, 12L/
12D or 8L/16D photoperiod. Both photoperiod and the position of the light
source over the gradient significantly influenced the diel cycle of thermal selection. The position of the light also significantly affected mean selected body
temperatures. Photoperiod and light position were important both as separate
and conjunctional factors influencing behavioral thermoregulation. We concluded that light position, heat, and photoperiod are used as separate cues in
behavioral thermoregulation.
changes throughout a repPHOTOPERIOD
tile's activity season provide reliable seasonal cues. Reptiles often thermoregulate within narrow limits when provided the opportunity,
and some show seasonal differences in temperature preference (Patterson and Davies, 1978;
Sievert and Hutchison, 1989a; Van Damme et
al., 1986). Photoperiod affects many behavioral
and physiological responses including repro-
duction (Licht, 1971a, 1971b), DNA synthesis
in the testis (Noeske and Meier, 1983), temperature selection (Graham and Hutchison, 1979;
Rismiller and Heldmaier, 1982, 1988), diel patterns of temperature selection (Rismiller and
Heldmaier, 1982; Spellerberg, 1974), and thermal tolerance (Hutchison and Kosh, 1964; Licht,
1968).
Photoperiods with a longer photophase (16L/
? 1991 by the American Society of Ichthyologists and Herpetologists