THE OPPORTUNISTIC SERPENT: MALE GARTER SNAKES
ADJUST COURTSHIP TACTICS TO MATING OPPORTUNITIES
by
R. SHINE1,2) , T. LANGKILDE 1) and R.T. MASON 3,4)
(1 School of Biological Sciences A08, University of Sydney, NSW 2006, Australia;
3 Department of Zoology, Oregon State University, Cordley Hall 3029, Corvallis, Oregon
97331-2914, USA)
(Acc. 28-VIII-2003)
Summary
Reproductive males encounter potential mates under a range of circumstances that in uence
the costs, beneŽ ts or feasibility of alternative courtship tactics. Thus, males may be under
strong selection to  exibly modify their behaviour. Red-sided garter snakes (Thamnophis sirtalis parietalis) in Manitoba overwinter in communal dens, and court and mate in large aggregations in early spring. The number of males within a courting group varies considerably, as
do the body sizes of both males and females. We manipulated these factors to set up replicated
courtship groups in outdoor arenas, and analysed videotapes of 82 courtship trials to quantify
courting behaviours of male snakes. Larger and more heavy-bodied males courted more vigorously than did their smaller, thinner-bodied rivals, and large females attracted more intense
courtship. The major effect, however, involved the number of rival males competing for copulation. Males in large groups not only reduced their overall vigour of courtship, but also modiŽ ed their tactics in such a way as to beneŽ t from the courtship activities of rival males. That
is, they devoted less energy to inducing female receptivity (which requires energy-expensive
caudocephalic waving) and more effort to behaviour (tail-searching)that enhanced their own
2)
Corresponding author’s e-mail address: [email protected]
We thank Al and Gerry Johnson, Ruth Nesbitt, Heather Waye, Ryan O’Donnell and Deb
Lutterschmidt for help and encouragement, and the Manitoba Dept. of Natural Resources for
permits. Financial support was provided by the Australian Research Council (to RS), and
by a National Science Foundation National Young Investigator Award (IBN-9357245), and
the Whitehall Foundation (W95-04) to RTM. Research was conducted under the authority of
Oregon State University Institutional Animal Care and Use Committee Protocol No. LAR1848B. All research was conducted in accord with the US Public Health Service ‘Policy on
Humane Care and Use of Laboratory Animals’ and the National Institutes of Health ‘Guide
to the Care and Use of Laboratory Animals’.
4)
© Koninklijke Brill NV, Leiden, 2003
Behaviour 140, 1509-1526
Also available online -
1510
SHINE, LANGKILDE & MASON
probability of mating if the female gaped her cloaca. This social parasitism reveals an unsuspected plasticity and complexity in the behavioural tactics of reproducing male snakes.
Keywords: operational sex ratio, plasticity, reproductive success, reptile, sexual selection.
Introduction
A high proportion of all behaviours in most kinds of organisms display some
degree of plasticity: aspects of the organism’s physiological state or environmental cues modify the expression of the behaviour. Plasticity of this kind is
likely to enhance Ž tness when (as is often the case) the animal unpredictably
encounters situations that modify the relative Ž tness consequences of alternative behaviours (e.g. Jackson & Pollard, 1996; Eltz, 1997; Schatz et al.,
1997; Henen et al., 1998). Even very ‘simple’ organisms display  exible
behaviours, indicating that behavioural plasticity does not require complex
information-processing capabilities (e.g. Demott & Moxter, 1991; Jurgens &
DeMott, 1995; Bushmann, 1999; Corbara & Dejean, 2000). In many cases,
organismal responses are modiŽ ed by local circumstances in ways that plausibly enhance Ž tness (e.g. Maynard Smith, 1976; Sinervo & Lively, 1996;
Alonzo & Sinervo, 2001; Candolin & Reynolds, 2002).
Some of the best examples of adaptive behavioural plasticity involve mating tactics. For example, the optimal tactics for mate location often depend
upon local conditions (resources, topography, etc.) that determine the spatial distribution and ease of location of receptive females (Emlen & Oring,
1977; Duvall & Schuett, 1997; Brown & Weatherhead, 1999). In the same
way, a male’s optimal tactics after Ž nding a female will depend on factors
(such as the numbers and sizes of rival males) that may vary unpredictably
over relatively short timespans and spatial scales (Travis & Woodward, 1989;
Dodson & Schwaab, 2001; Candolin & Reynolds, 2002). Hence, the tactics
of reproductive males offer a useful system with which to examine plasticity
in behavioural traits, and the ways in which such plasticity might enhance
mating success.
Reptiles have attracted relatively little attention in this respect, perhaps re ecting an assumption that they display limited behavioural plasticity. However, recent studies have revealed remarkable complexity in reptilian social
systems, including polymorphic male strategies. For example, side-blotched
COURTSHIP TACTICS IN SNAKES
1511
lizards (Uta stansburiana) occur in multiple male morphotypes with distinctive mating tactics (Sinervo & Lively, 1996; Sinervo et al., 2000a, b,
2001; Sinervo, 2001). Does signiŽ cant diversity also arise via behavioural
plasticity within single individuals? Studies on snakes have lagged well behind those on lizards in this respect, but recent research on one species (the
garter snake Thamnophis sirtalis) has revealed  exible reproductive tactics.
For example, males shift between alternative tactics based on their body size
(size-assortative courtship: Shine et al., 2001b), their ability to compete with
other males (female mimicry: Shine et al., 2000a, 2001c), and the numbers
and body sizes of rival males (participation in courtship: Shine et al., 2000b).
Male mate choice is in uenced also by attributes of the female, with large
and heavy-bodied females attracting more courtship than small, thin conspeciŽ cs (Aleksiuk & Gregory, 1974; Hawley & Aleksiuk, 1976; Gartska et
al., 1982; Shine et al., 2001b). Operational sex ratios within courting groups
also may modify the intensity of courtship by each male (Joy & Crews,
1985).
Previous studies on reproductive behaviour in garter snakes have been
based on visual observation, precluding detailed quantiŽ cation of the activities of individual snakes within ‘mating balls’ (which often contain several animals in simultaneous vigorous activity). Based on previous research
(above), we predicted that male snakes might modify either the intensity or
form of their courtship behaviours in response to three factors: (1) aspects
of their own phenotype (body size and condition) that could in uence their
ability to compete with other males; (2) the potential Ž tness beneŽ ts from a
mating (body size, and thus probable fecundity, of the female being courted);
and (3) the intensity of male-male competition (numbers of rival males).
Thus, we videotaped courting snakes in experimental arenas (allowing us to
manipulate all three of the above variables) to test these predictions.
Methods
Study species and area
Red-sided garter snakes (Thamnophis sirtalis parietalis) are small (<1 m total length) nonvenomous snakes with a broad geographic distribution in North America. We studied these
snakes in the Interlake region north of Winnipeg, Manitoba, in central Canada. Snakes in this
severely cold region spend about eight months per year inactive in communal winter dens,
emerging in early May to court and mate before dispersing to feed and give birth in the surrounding swamps and grasslands (Gregory, 1974, 1977; Gregory & Stewart, 1975). Dens may
1512
SHINE, LANGKILDE & MASON
contain >10,000 snakes and courtship occurs close to the den; thus, densities of snakes are
very high at this time (sometimes, >100/m2 : Shine et al., 2001a). Newly-emerging females
are rapidly located and vigorously courted by dozens or hundreds of males, but the mean size
of courting groups (numbers of males) decreases rapidly as one moves away from the den
(Shine et al., 2001a). In the grasslands surrounding the den, some females are courted by
single males (Shine et al., 2001a).
Male garter snakes locate females using the vomeronasal system, with skin lipids on
females serving as sex pheromones (Mason, 1993). Males align their bodies with that of
the female, pressing their chins against her dorsal and lateral surfaces and positioning their
cloacae next to hers. Males maintain this position with frequent attempts to intertwine their
tailbase beneath that of the female; any movement by the female provokes an immediate
 urry of activity by male tails (‘tail-searching’: Noble, 1937; Blanchard & Blanchard, 1941;
Crews & Gartska, 1982; Whittier et al., 1985). Large males may push the tails of their small
rivals out of the way at such times (Luiselli, 1996; Shine et al., 2000f). The most distinctive
courtship tactic by males involves caudocephalic waving, a series of muscular contractions
that travel from the male’s cloacal area to his head (Noble, 1937; Gillingham, 1987). Recent
studies suggest that these pressure waves push anoxic air from the female’s non-respiratory
posterior airsac anteriorly to her respiratory lung, thus inducing the stress response of cloacal
gaping (Shine et al., 2003a). This gaping facilitates intromission of the male’s hemipenis.
During May 2002 we worked at a communal den containing approximately 10,000 snakes
1.5 km N of the town of Inwood, 250 m E of Highway 17 in central southern Manitoba
(50± 31:580 N 97± 29:710 W). This den has been the focus of considerable previous research
(e.g. Shine et al., 2000b, c, d, e, 2001b, 2003a, b).
Procedures
We set up four small outdoor arenas (‘Space Pop’, Smash Enterprises, Melbourne; circular
with 48 cm diameter, 56 cm deep) in a  at sunny area <10 m from the den, to video-record
courting groups. Unmated females (i.e. lacking a mating plug: Devine, 1977; Shine et al.,
2000e) were collected as they emerged, and one such female was placed in each arena. For
later analysis we grouped these animals into four size categories such that each category
contained approximately equal numbers of snakes (23.7-53.6, 53.7-57.2, 57.3-60.0, 60.170.0 cm SVL). We then added either one, four or 24 males (collected from courting groups
immediately prior to the video trials) to each arena. We conducted 24 trials involving one
male plus a female, 25 trials with 24 males plus a female, and 33 trials with four males
plus a female. In the one-male and four-male trials, all snakes were painted with coloured
mid-dorsal stripes for individual recognition. In the 24-male trials, we painted four ‘focal’
males (and the female) in the same way, but the other males were unpainted. The four painted
males in each four-male and 24-male trial comprised one from each of four arbitrarily-deŽned
size classes, encompassing the normal size variation for males at the Inwood den (as above,
divided such that equal numbers of snakes fell into each of the four size classes). We Ž lmed
the animals with Sony Handycam CCD-TRV46E digital videocameras on tripods. From the
resulting videotapes, we randomly selected two 30-sec samples three minutes apart, and at
least two minutes prior to actual mating. At each of these two time periods we scored four
nominal variables for each focal male: whether or not (a) his body was aligned with that
of the female; (b) he attempted a tail-search; (c) he performed one or more caudocephalic
waves; and/or (d) he chin-pressed the female, within the 30 sec period. We also scored three
COURTSHIP TACTICS IN SNAKES
1513
continuous variables: (a) the distance between his cloaca and the female’s cloaca; (b) the
proportion of his body aligned with that of the female; and (c) the number of caudocephalic
waves that he performed during 30 sec.
Analysis
Each animal (either male or female) was used in only a single trial, to avoid pseudoreplication. Data were analysed on an Apple Macintosh G4 computer using the statistical software
package Statview 5 (SAS Institute, 1998). Relevant assumptions of tests (normality of distributions, equality of variances) were checked prior to analysis. The text presents mean values
§ one SE.
We examined male body sizes using two-factor ANOVA with the factors being trial type
(number of males per female) and male size class. Body sizes of males were similar among
the trial types (F2;244 D 0:41, p D 0:67/, with no signiŽ cant interaction between trial type
and male size class (F6;67 D 1:43, p D 0:20/. However, the four size classes of males differed
consistently (F3;244 D 107:39, p < 0:001/. The ‘extra-large’males averaged 50:08§0:37 cm
snout-vent length (SVL) and 37:76 § 0:96 g; ‘large’ males were 46:70 § 0:30 cm and
31:21 § 0:81 g; ‘medium’ males were 42:98 § 0:30 cm and 24:24 § 0:63 g; and ‘small’
males were 40:00 § 0:35 cm and 19:78 § 0:67 g. Analysed in the same way, data on female
sizes showed the same pattern: mean body sizes of females did not differ between these three
types of trials (two-factor ANOVA: F2;70 D 0:89, p D 0:42/, but differed between our four
size classes (F3;70 D 64:80, p < 0:0001I mean SVLs D ‘extra-large’, 63:39 § 0:28 cm;
‘large’, 58:47 § 0:12 cm; ‘medium’, 55:22 § 0:16 cm; ‘small’, 49:19 § 0:54 cm), with no
signiŽ cant interaction between trial type and female size class (F6;70 D 0:83, p D 0:55/.
Previous work has shown that courtship vigour and success in red-sided garter snakes is
in uenced by female body size (large females attract more courtship: Aleksiuk & Gregory,
1974; Hawley & Aleksiuk, 1976) as well as by male body size (large heavy-bodied males
court more vigorously and obtain more matings: Shine et al., 2000f). Thus, we included
female size (SVL) as well as male size (SVL) and shape (body condition, calculated as
residual scores from the linear regression of ln mass vs ln svl), together with trial type
(D number of males per female) as independent variables in multiple logistic regressions.The
behaviours we scored as presence/absence for each focal male were used as the dependent
variable, with the behaviour scored as present if the male showed it during either of the two
observation periods. Because data from the two periods were thus combined into a single
response variable for the analysis, each male contributed only a single data point.
Results
Log-likelihood ratio tests from these logistic regressions yielded very similar
results for the four behaviours we examined in this way (Table 1). A male’s
body size and condition did not in uence his courtship except that heavierbodied males were more likely to perform caudocephalic waves. Figures
1 and 2 show courtship behaviours of males as a function of female body
size (divided into four equal-sized groups) and trial type (number of males).
1514
SHINE, LANGKILDE & MASON
TABLE 1. Results of log-likelihood ratio tests from multiple logistic regressions to evaluate whether phenotypic traits of garter snakes (Thamnophis
sirtalis parietalis), and numbers of males per courting group, in uenced behaviours exhibited by courting males
Variable
Degrees of freedom
Align to female?
Tail-searching?
Caudocephalic waving?
Chin-pressing?
Female
SVL
Male
SVL
Male
condition
Group
size
1
14.44***
5.63*
25.74***
7.76**
1
0.0001
0.32
0.19
0.08
1
0.47
0.09
7.07**
0.28
2
40.53***
24.73***
36.08***
40.09***
SVL D snout-vent length; ‘condition’ D residual score from linear regression of ln mass vs ln
SVL; ‘group size’ D number of males courting the female (1, 4 or 24). Text shows  2 values,
with signiŽ cance levels indicated by asterisks (*p < 0:05; **p < 0:01; ***p < 0:001/.
Values with p < 0:05 shown in boldface.
Large females attracted more intense courtship (alignments, tail-searching,
caudocephalic waving, chin-pressing: see Fig. 1). The major effect on all
these variables, however, was the number of competing males. Courtship
was more vigorous when a male was alone with a female than when he was
part of a group with three other males, and courtship intensity decreased still
further in a large group (Table 1, Fig. 1).
The higher intensity of courtship to large females was evident for each of
the nominal variables that we scored, and this effect was generally stronger
in four-male trials than in the other courting groups (Fig. 1). Analysis of
the continuous variables (Fig. 2) revealed broadly similar patterns. Males
maintained their tailbases closer to the female cloaca in trials with fewer
rivals (F2;228 D 22:75, p < 0:0001I posthoc comparisons all differ at p <
0:05/, with no in uence of female body size (F3;228 D 0:72, p D 0:54I
interaction: F6;228 D 1:20, p D 0:31I see Fig. 2). Similarly, the proportion
of the male’s body aligned with the female was higher in small courting
groups (F2;228 D 36:24, p < 0:0001I posthoc comparisons all differ at
p < 0:05/ but was not affected by female body size (F3;228 D 1:30, p D
0:28I interaction: F6;228 D 1:35, p D 0:24I see Fig. 2). The frequency
of caudocephalic waves was higher in small groups (F2;228 D 28:16, p <
0:0001I posthoc comparisons all differ at p < 0:05/ and was higher for
large females (F3;228 D 4:11, p < 0:008I posthoc tests show that S < L and
COURTSHIP TACTICS IN SNAKES
1515
Fig. 1. Effects of body size of female and number of courting males per female on courtship
behaviours exhibited by garter snakes (Thamnophis sirtalis parietalis). Groups contained
either 1, 4 or 24 males. We scored whether or not each focal male aligned with the female’s
body (a), tail-searched (b), performed caudocephalic waves (c) and/or chin-pressed against
the female (d), in two 30-sec observation periods. Size categories of females were extra-large
(XL), large (L), medium (M) or small (S). Sample sizes (left to right) D 4, 3, 9, 8, 28, 21, 23,
28, 32, 40, 32 and 28. See text for further explanation and statistical tests.
M), with no signiŽ cant interaction between these two factors (F6;228 D 0:98,
p D 0:44I see Fig. 2).
Figures 3 and 4 show male responses as a function of male rather than
female body size. For the continuous variables that we scored for each male,
we again calculated a single score for each animal (mean of the two values
for successive time periods) and used these as dependent variables in twofactor ANOVAs with trial type (number of competing males) and male size
class as the factors. Analyses including female SVL as a covariate (ANCOVA) generated identical conclusions, and so this factor is omitted for
simplicity. As for the nominal variables analysed above, male courtship behaviours were affected by the number of rival males. Males were able to
maintain their cloacae closer to those of the females if no rivals were present
1516
SHINE, LANGKILDE & MASON
Fig. 2. Effects of body size of female and number of courting males per female on courtship
behaviours exhibited by garter snakes (Thamnophis sirtalis parietalis). Groups contained
either 1, 4 or 24 males. We scored mean distances between the cloacae of focal males and
females (a), the proportion of the males’ bodies aligned with those of the females (b), and
the mean number of caudocephalic waves performed by each male within two 30-sec periods
(c). Histograms show mean values and one SE. Size categories of females were extra-large
(XL), large (L), medium (M) or small (S). Sample sizes as for Fig. 1. See text for further
explanation and statistical tests.
(F2;228 D 24:83, p < 0:001I see Fig. 4a). There were no signiŽ cant overall differences among size classes of males in this respect (F3;228 D 0:57,
p D 0:63/, but there was a signiŽ cant interaction between male size class and
number of competing males (F6;228 D 2:79, p < 0:02/. Larger males maintained their tailbases closer to the female’s cloaca than did small males if the
group consisted of only four courting animals, but the pattern was reversed
in the 24-male aggregations (Fig. 4a). The proportion of the focal male’s
body aligned with that of the female was higher in one-on-one courtship
than in large groups (Fig. 4b; F2;228 D 36:98, p < 0:0001I posthoc tests
show that all trial types differ at p < 0:05/, with no in uence of male body
size (F3;228 D 0:38, p D 0:77I interaction: F6;228 D 0:92, p D 0:48/. The
COURTSHIP TACTICS IN SNAKES
1517
Fig. 3. Effects of body size of male and number of courting males per female on courtship
behaviours exhibited by garter snakes (Thamnophis sirtalis parietalis). Groups contained
either 1, 4 or 24 males. We scored whether or not each focal male aligned with the female’s
body (a), tail-searched (b), performed caudocephalic waves (c) and/or chin-pressed against
the female (d), in two 30-sec observation periods. Size categories of males were extra-large
(XL), large (L), medium (M) or small (S). Sample sizes (left to right) D 6, 6, 6, 6, 33, 33, 33,
33, 25, 25, 25 and 25. See text for further explanation and statistical tests.
same pattern is evident in Fig. 3a, which shows that the proportion of males
aligning with the female was lower in groups with more intense male-male
rivalry. Caudocephalic waving was displayed by males regardless of group
size (Fig. 3c) but the rate of waving was higher in small groups (Fig. 4c;
F2;244 D 27:78, p < 0:001I posthoc comparisons all differ at p < 0:05/ and
more frequent by larger than by smaller males (F3;244 D 4:55, p < 0:005I
posthocs show S differ from XL, L and M at p < 0:05/. The interaction between male size and number of rivals was close to statistical signiŽ cance
(F6;244 D 2:10, p D 0:054/, re ecting a trend for very large males to
court more actively in four-male groups than in either of the other trial types
(Fig. 4c).
These analyses show that male courtship behaviours shifted in complex
ways as a function of the numbers of competing males, as well as male and
1518
SHINE, LANGKILDE & MASON
Fig. 4. Effects of body size of male and number of courting males per female on courtship
behaviours exhibited by garter snakes (Thamnophis sirtalis parietalis). Groups contained
either 1, 4 or 24 males. We scored mean distances between the cloacae of focal males and
females (a), the proportion of the males’ bodies aligned with those of the females (b), and the
mean number of caudocephalic waves performed by each male within two 30-sec periods (c).
Histograms show mean values and one SE. Size categories of males were extra-large (XL),
large (L), medium (M) or small (S). Sample sizes as for Fig. 3. See text for further explanation
and statistical tests.
female body sizes. Do these shifts re ect simple changes in intensity, or in
tactics as well? For example, if males are ‘social parasites’, we might expect that a male in a large group would continue to tail-search (to maximise
his chances of intromission if the female gapes her cloaca) but reduce his
rate of energy-expensive behaviours such as caudocephalic waving. Figure 5
shows exactly this situation. A two-factor ANOVA with presence or absence
of tail-searching as one factor and trial type as the other, with rate of caudocephalic waving as the dependent variable, shows a signiŽ cant interaction
term (F2;250 D 5:63, p < 0:005I see Fig. 5). That is, males in large groups
COURTSHIP TACTICS IN SNAKES
1519
Fig. 5. Effects of group size (number of rival males) on the frequency with which courting
male garter snakes performed caudocephalic waves (mean value over two 30-sec periods).
Data (mean values and 1 SE) are shown separately for males that were actively engaged in
tail-searching and those that were not. Sample sizes (left to right) D 8, 16, 55, 75, 70, and 30
males.
continued to tail-search but substantially reduced their rate of caudocephalic
waving.
Discussion
Re ecting their remarkable tolerance of disturbance (Moore & Mason,
2001), the garter snakes in our outdoor arenas courted vigorously, and with
the same behaviours as seen in natural courting groups in the den a few metres away. The arenas allowed us to control group sizes (operational sex ratios) and body sizes of both males and females, and hence to explore the ways
that male tactics are in uenced by these factors. Videotaping permitted unambiguous scoring of behaviours for each focal male. The clear result from
the study is that male garter snakes exhibit signiŽ cant plasticity in courtship
tactics in response to each of the variables that we manipulated. This plasticity involved not only the intensity of effort that a male devoted to courtship,
but also his allocation of effort among different components of courtship
behaviour. In each case, the direction of the shift was consistent with the
adaptationist hypothesis that male courtship plasticity enhances mating success. Below, we consider each of the factors (and interactions between them)
that modiŽ ed the tactics of male snakes in our trials.
1520
SHINE, LANGKILDE & MASON
Female body size
Large females attracted more intense courtship than did small conspeciŽ cs
(Figs. 1 and 2). Similar results have been reported in most previous studies
of the same species, with large females attracting more suitors than small
females (Aleksiuk & Gregory, 1974; Hawley & Aleksiuk, 1976; Gartska et
al., 1982; Shine et al., 2001b). Only one previous study of red-sided garter
snakes has reported that female body size did not in uence courtship intensity (Joy & Crews, 1985). Large females stimulate higher tongue- ick rates
from mate-searching male snakes (Shine et al., 2003c). Large females are
courted or mated by more males in at least two other snake taxa (Laticauda
colubrina — Shetty & Shine, 2002; Nerodia sipedon — Prosser et al., in
press), suggesting that this may be a widespread phenomenon. In red-sided
garter snakes, the preference for large females may be driven by pheromonal
cues rather than (or as well as) visual cues (LeMaster & Mason, 2002; Shine
et al., 2003c). The strength of this preference for large females increases with
male body size (Shine et al., 2001b).
Previous work on this topic has interpreted the greater attractiveness of
large females in terms of potential Ž tness beneŽ ts to the male because female
fecundity increases with maternal body size in this as in most other snake
species (Fitch, 1970; Gregory, 1977). Hence, a male that mates with a large
female may thereby fertilise more ova. The recent discovery that male redsided garter snakes force matings by inducing hypoxic stress to females
(Shine et al., 2003a) suggests another reason why large females are subject
to more vigorous courtship. In natural courting groups, the proportion of a
female’s body that is covered by courting males depends upon her body size;
large females are less encumbered in this respect, especially in the anterior
part of the body containing the vascular lung (Shine et al., 2003a). Thus,
males may court (and especially, caudocephalic-wave) more vigorously to
large females in order to induce sufŽ cient stress to initiate cloacal gaping.
Male body size and body condition
Large males were more vigorous in many aspects of courtship (e.g. Fig. 4c)
and heavier-bodied males performed more frequent caudocephalic waves
(Table 1). An earlier study of the same population reported a similar result,
with more active courtship by large males (Joy & Crews, 1985). Extensive
data from Ž eld and arena matings have shown that larger, heavier-bodied
COURTSHIP TACTICS IN SNAKES
1521
males experience higher mating success than do their small, thinner rivals
(Shine et al., 2000f). Videotape analyses of courting trials (4 males, 1 female)
that resulted in matings showed that the successful males were those that
courted most vigorously (Shine et al., 2003b). Thus, the size advantage for
mating in this system may partly be due to a size-associated shift in courtship
vigour rather than a simple ability of large males to displace small males. The
trend for more vigorous courtship from heavier-bodied males may re ect
the fact that these animals have greater energy reserves: males lose mass
(and thus, body condition) rapidly during their period of residency at the
den during the mating season (Shine et al., 2001a), and leave the den when
their reserves fall below some threshold level (Shine et al., 2003d). Our data
reinforce the complexity of the relationship between body size and male
mating success within red-sided garter snakes. For example, small males not
only court less vigorously (perhaps expending less energy per courtship?)
and obtain fewer matings, but also tend to mate with small females (Shine et
al., 2001b), at least some of which are too small to produce offspring (Shine
et al., 2000d).
Number of rival males
Our most striking results involve the effect of operational sex ratio (i.e.
number of competing males in a courting group) on the courtship tactics
of males. For almost every variable that we analysed, the focal males were
most active in courtship when they were alone with a female, less active
when they were competing with three other males, and less active again when
competing with 23 rivals. To our knowledge, no previous study has reported
this kind of decrease in courtship intensity in large groups. Indeed, previous
research suggested either an increased courtship intensity in large groups
(Joy & Crews, 1985), or (anecdotally) that a male’s zeal for courtship was
unaffected by group size (Blanchard & Blanchard, 1941). Nonetheless, the
pattern in our data is extremely strong and consistent, with some courtship
behaviours an order of magnitude less frequent in large groups than small
ones (e.g. Fig. 2c). Because our data are based on values per male, the total
number of caudocephalic waves, chin-presses, body alignments and tailsearches to a female would nonetheless increase with group size (i.e. the
differences between courtship intensities per male to females in 1-male vs
24-male groups is <24-fold; see Figs.).
1522
SHINE, LANGKILDE & MASON
Why was courtship vigour of males highest in small groups, and especially if the male was alone with the female? On a proximate level, a male
alone with a female may have maximal access to pheromonal cues from
her; in large groups the female is obscured by the bodies of rival males. In
terms of Ž tness beneŽ ts, two aspects could be important. First, a solitary
male would have to court very vigorously in order to induce sufŽ cient hypoxic stress in the female and hence, facilitate mating. Second, opportunities
for one-on-one courtship are rare and likely to be short-lived in the crowded
den environment (Shine et al., 2001a) and hence, a male in this situation may
beneŽ t from inducing female receptivity very rapidly, before any rivals arrive
to potentially ‘steal’ a mating that would otherwise be his.
The decreased intensity of courtship in large groups was manifested in
a shift in tactics as well as an overall decline in effort. Thus, not only did
the proportion of actively tail-searching males decline in large groups (from
around 70% in 1-male courtship to 30% in 24-male courtship: Fig. 3b), but
even the males that continued to tail-search showed a precipitous decline
(about 7-fold: Fig. 5) in their frequency of caudocephalic waving. This latter
behaviour is the most vigorous (and presumably, energy-expensive) component of male courtship, and plays a critical role in inducing the female cloacal gaping that is required for successful intromission (Shine et al., 2003a).
The continued tail-searching by these males shows that they were aware of
the female and were attempting to obtain a copulation, but nonetheless substantially reduced their energy expenditure by reducing their waving rate.
Because females in large groups are subject to high rates of caudocephalic
waving simply because of the large numbers of (infrequently-waving) males,
a male in such a group can ‘afford’ to reduce caudocephalic waving (and
thus, save energy) and instead, focus on ensuring that his cloaca is immediately adjacent to the female’s. This positioning is a major predictor of male
mating success (Shine et al., 2003b).
Thus, male garter snakes modify both their overall effort and their speciŽ c
tactics as a function of the operational sex ratio within courting groups. The
net effect is to save energy, and perhaps facilitate accurate cloacal positioning
because of the reduced need for simultaneous performance of con icting
tasks. In essence, the response constitutes ‘social parasitism’, whereby a
male can enhance his own mating opportunities by exploiting the activities of
his rivals. Female mimicry in the same population provides another example
of this phenomenon: a male that elicits courtship from other males thereby
COURTSHIP TACTICS IN SNAKES
1523
obtains beneŽ ts (protection from predators, warming) at some cost in risk
and energy to his rivals (Shine et al., 2001c). Manipulation of the behaviour
of other males (and females, through induction of hypoxic stress: Shine et
al., 2003a) thus plays a signiŽ cant role in male mating tactics within this
population.
Interactions between body size and number of rivals
For two of the variables we measured, a male’s body size in uenced his
behaviour differently in different-sized groups. In relatively small (4-male)
groups, large male snakes performed more frequent caudocephalic waves,
and maintained their tailbases closer to that of the female; these patterns
were reversed in the large (24-male) aggregations (Fig. 4a, c). If mating
success is more deterministic (and thus, more consistently related to male
body size) in small groups than in large groups (Crews, 2000), large males
may be favoured to specialise on small groups and to devote high levels of
effort in such circumstances. In contrast, small males may be at less of a
disadvantage in the more stochastic situation of large groups; indeed, their
small size may facilitate cloacal apposition with the female amidst the tangle
of competing tails (Fig. 4a). Hence, males may adjust their courtship tactics
depending on the degree to which greater strength and vigour is likely to
enhance their probability of mating.
The end result of these patterns is that male garter snakes  exibly modify
the intensity and nature of their courtship depending upon their own body
size, their body condition (energy reserves), the size of the female, and the
numbers of rival males. These factors also interact with each other in complex ways. In each case, the patterns of response support predictions from the
adaptationist hypothesis that males should behave in ways likely to maximise
their mating success while minimising their energy costs. Such plasticity is
likely to be widespread in snakes as well as in other kinds of organisms.
References
Aleksiuk, M. & Gregory, P.T. (1974). Regulation of seasonal mating behavior in Thamnophis
sirtalis parietalis. — Copeia 1974, p. 682-689.
Alonzo, S.H. & Sinervo, B. (2001). Mate choice games, context-dependent good genes, and
genetic cycles in the side-blotched lizard, Uta stansburiana. — Behav. Ecol. Sociobiol.
49, p. 176-186.
1524
SHINE, LANGKILDE & MASON
Blanchard, F.N. & Blanchard, F.C. (1941). Mating of the garter snake Thamnophis sirtalis
sirtalis (Linnaeus). — Pap. Mich. Acad. Sci., Arts Lett. 27, p. 215-234.
Brown, G.P. & Weatherhead, P.J. (1999). Female distribution affects mate searching and
sexual selection in male northern water snakes (Nerodia sipedon). — Behav. Ecol.
Sociobiol. 47, p. 9-16.
Bushmann, P.J. (1999). Concurrent signals and behavioral plasticity in blue crab (Callinectes
sapidus Rathbun) courtship. — Biol. Bull. 197, p. 63-71.
Candolin, U. & Reynolds, J.D. (2002). Adjustments of ejaculation rates in response to risk of
sperm competition in a Ž sh, the bitterling (Rhodeus sericeus). — Proc. Roy. Soc. Lond.
Ser. B 269, p. 1549-1553.
Corbara, B. & Dejean, A. (2000). Adaptive behavioral  exibility of the ant Pachycondyla
analis (D Megaponera foetens) (Formicidae: Ponerinae) during prey capture. — Sociobiol. 36, p. 465-483.
Crews, D. (2000). Reply to Shine et al. (2000). — Anim. Behav. 59, p. F12.
— — & Gartska, W.R. (1982). The ecological physiology of a garter snake. — Sci. Am. 247,
p. 136-144.
Demott, W.R. & Moxter, F. (1991). Foraging on cyanobacteria by copepods — Responses to
chemical defenses and resource abundance. — Ecology 72, p. 1820-1834.
Devine, M.C. (1977). Copulatory plugs, restricted mating opportunities and reproductive
competition among male garter snakes. — Nature 267, p. 345-346.
Dodson, G.N. & Schwaab, A.T. (2001). Body size, leg autotomy, and prior experience as factors in the Ž ghting success of male crab spiders, Misumenoides formosipes. — J. Insect
Behav. 14, p. 841-855.
Duvall, D. & Schuett, G.W. (1997). Straight-line movement and competitive mate-searching
in prairie rattlesnakes, Crotalus viridis viridis. — Anim. Behav. 54, p. 329-334.
Eltz, T. (1997). Foraging in the ant-lion Myrmeleon mobilis Hagen 1888 (Neuroptera:
Myrmeleontidae): Behavioral  exibility of a sit-and-wait predator. — J. Insect Behav.
10, p. 1-11.
Emlen, S.T. & Oring, L.W. (1977). Ecology, sexual selection, and the evolution of mating
systems. — Science 197, p. 215-223.
Fitch, H.S. (1970). Reproductive cycles in lizards and snakes. — Univ. Kansas Mus. Nat.
Hist., Misc. Publ. 52, p. 1-247.
Gartska, W.R., Camazine, B. & Crews, D. (1982). Interactions of behaviour and physiology
during the annual reproductive cycle of the red-sided garter snake (Thamnophis sirtalis
parietalis). — Herpetol. 38, p. 104-123.
Gillingham, J.C. (1987). Social behavior. — In: Snakes: ecology and evolutionary biology
(R.A. Seigel, J.T. Collins & S.S. Novak, eds). McGraw-Hill, New York, p. 184-209.
Gregory, P.T. (1974). Patterns of spring emergence of the red-sided garter snake (Thamnophis
sirtalis parietalis) in the Interlake region of Manitoba. — Can. J. Zool. 52, p. 10631069.
— — (1977). Life-history parameters of the red-sided garter snake (Thamnophis sirtalis
parietalis) in an extreme environment, the Interlake region of Manitoba. — Nat. Mus.
Can., Publ. Zool. 13, p. 1-44.
— — & Stewart, K.W. (1975). Long-distance dispersal and feeding strategy of the red-sided
garter snake (Thamnophis sirtalis parietalis) in the Interlake of Manitoba. — Can. J.
Zool. 53, p. 238-245.
COURTSHIP TACTICS IN SNAKES
1525
Hawley, A.W.L. & Aleksiuk, M. (1976). Sexual receptivity in the female red-sided garter
snake (Thamnophis sirtalis parietalis). — Copeia 1976, p. 401-404.
Henen, B.T., Peterson, C.C., Wallis, I.R., Berry, K.H. & Nagy, K.A. (1998). Effects of climatic variation on Ž eld metabolism and water relations of desert tortoises. — Oecol.
117, p. 365-373.
Jackson, R.R. & Pollard, S.D. (1996). Predatory behavior of jumping spiders. — Ann. Rev.
Entomol. 41, p. 287-308.
Joy, J.E. & Crews, D. (1985). Social dynamics of group courtship behavior in male red-sided
garter snakes (Thamnophis sirtalis parietalis). — J. Comp. Psychol. 99, p. 145-149.
Jurgens, K. & DeMott, W.R. (1995). Behavioral  exibility in prey selection by bacterivorous
nano agellates. — Limnol. Oceanog. 40, p. 1503-1507.
LeMaster, M.P. & Mason, R.T. (2002). Evidence for a sex pheromone mediating mate choice
in the red-sided garter snake, Thamnophis sirtalis parietalis. — J. Chem. Ecol., in press.
Luiselli, L. (1996). Individual success in mating balls of the grass snake, Natrix natrix: size
is important. — J. Zool. 239, p. 731-740.
Mason, R.T. (1993). Chemical ecology of the red-sided garter snake, Thamnophis sirtalis
parietalis. — Brain Behav. Evol. 41, p. 261-268.
Maynard Smith, J. (1976). Evolution and the theory of games. — Am. Sci. 64, p. 41-45.
Moore, I.T. & Mason, R.T. (2001). Behavioral and hormonal responses to corticosterone in
the male red-sided garter snake, Thamnophis sirtalis parietalis. — Physiol. Behav. 72,
p. 669-674.
Noble, G.K. (1937). The sense organs involved in the courtship of Storeria, Thamnophis, and
other snakes. — Bull. Amer. Mus. Nat. Hist. 73, p. 673-725.
Prosser, M.R., Weatherhead, P.J., Gibbs, H.L. & Brown, G.P. (in press). Genetic analysis
of the mating system and opportunity for sexual selection in northern water snakes
(Nerodia sipedon). — Behav. Ecol.
SAS Institute (1998). Statview 5. — Cary, Indiana.
Schatz, B., Lachaud, J.P. & Beugnon, G. (1997). Graded recruitment and hunting strategies
linked to prey weight and size in the ponerine ant Ectatomma ruidum. — Behav. Ecol.
Sociobiol. 40, p. 337-349.
Shetty, S. & Shine, R. (2002). The mating system of yellow-lipped sea kraits (Laticauda
colubrina: Laticaudidae). — Herpetol. 58, p. 170-180.
Shine, R., Elphick, M.J., Harlow, P.S., Moore, I.T., LeMaster, M.P. & Mason, R.T. (2001a).
Movements, mating and dispersal of red-sided gartersnakes from a communal den in
Manitoba. — Copeia 2001, p. 82-91.
— —, Harlow, P.S., LeMaster, M.P., Moore, I. & Mason, R.T. (2000a). The transvestite
serpent: why do male gartersnakes court (some) other males? — Anim. Behav. 59,
p. 349-359.
— —, Langkilde, T. & Mason, R.T. (2003a). Cryptic forcible insemination: male snakes
exploit female physiology, anatomy and behavior to obtain coercive matings. — Am.
Nat., in press.
— —, — — & — — (2003b). Courtship tactics in garter snakes: how do a male’s morphology
and behavior in uence his mating success? — Anim. Behav., in press.
— —, O’Connor, D., LeMaster, M.P. & Mason, R.T. (2001b). Pick on someone your own
size: ontogenetic shifts in mate choice by male garter snakes result in size-assortative
mating. — Anim. Behav. 61, p. 1-9.
1526
SHINE, LANGKILDE & MASON
— —, — — & Mason, R.T. (2000b). Female mimicry in gartersnakes: behavioural tactics of
‘she-males’ and the males that court them. — Can. J. Zool. 78, p. 1391-1396.
— —, — — & — — (2000c). The problem with courting a cylindrical object: how does an
amorous male snake determine which end is which? — Behav. 137, p. 727-739.
— —, — — & — — (2000d). Sexual con ict in the snake den. — Behav. Ecol. Sociobiol.
48, p. 392-401.
— —, Olsson, M.M. & Mason, R.T. (2000e). Chastity belts in gartersnakes: the functional
signiŽ cance of mating plugs. — Biol. J. Linn. Soc. 70, p. 377-390.
— —, — —, Moore, I., LeMaster, M.P., Greene, M. & Mason, R.T. (2000f). Body size
enhances mating success in male gartersnakes. — Anim. Behav. 59, p. F4-F11.
— —, Phillips, B., Waye, H., LeMaster, M. & Mason, R.T. (2001c). Advantage of female
mimicry to snakes. — Nature 414, p. 267.
— —, — —, — —, — — & — — (2003c). The lexicon of love: what cues cause sizeassortative courtship by male garter snakes? — Behav. Ecol. Sociobiol. 53, p. 234-237.
— —, — —, — — & Mason, R.T. (2003d). Behavioral shifts associated with reproduction
in garter snakes. — Behav. Ecol. 14, p. 251-256.
Sinervo, B. (2001). Runaway social games, genetic cycles driven by alternative male and
female strategies, and the origin of morphs. — Genetica 112, p. 417-434.
— —, Bleay, C. & Adamopoulou, C. (2001). Social causes of correlational selection and
the resolution of a heritable throat color polymorphism in a lizard. — Evolution 55,
p. 2040-2052.
— — & Lively, C.M. (1996). The rock-paper-scissorsgame and the evolution of alternative
male strategies. — Nature 380, p. 240-243.
— —, Miles, D.B., Frankino, W.A., Klukowski, M. & DeNardo, D.F. (2000a). Testosterone,
endurance, and Darwinian Ž tness: Natural and sexual selection on the physiological
bases of alternative male behaviors in side-blotchedlizards. — Horm. Behav. 38, p. 222233.
— —, Svensson, E. & Comendant, T. (2000b). Density cycles and an offspring quantity and
quality game driven by natural selection. — Nature 406, p. 985-988.
Travis, J. & Woodward, B.D. (1989). Social context and courtship  exibility in male sailŽ n
mollies, Poecilia latipinna (Pisces, Poeciliidae). — Anim. Behav. 38, p. 1001-1011.
Whittier, J.M., Mason, R.T. & Crews, D. (1985). Mating in the red-sided gartersnake,
Thamnophis sirtalis parietalis: differential effects on male and female sexual behavior. — Behav. Ecol. Sociobiol. 16, p. 257-261.