Behavioral Ecology
doi:10.1093/beheco/arp148
Advance Access publication 13 November 2009
How and why the winner effect forms:
influences of contest environment
and species differences
Matthew J. Fuxjagera and Catherine A. Marlera,b
Department of Zoology, University of Wisconsin—Madison, 250 N Mills Street, Madison,
WI 53706, USA and bDepartment of Psychology, University of Wisconsin—Madison,
W Johnson Street, Madison, WI 53706, USA
a
n many animal species, past winning experience enhances
the chances of future victories (reviewed in Hsu et al.
2006). Research indicates that this so-called winner effect
can help to shape social structures and relationships, including the establishment of dominance hierarchies (Bergman
et al. 2003; Dugatkin and Druen 2004) and possibly territorial
interactions (Oyegbile and Marler 2006). Yet, despite the winner effect’s apparent functional relevance, little is understood
about the processes that govern its formation within an
individual on either a proximate or ultimate level.
In general, the winner effect is considered an intrinsic phenomenon, meaning that it is thought to form as the result of
internal changes that transpire after a victory, including
changes in perceived fighting ability (Beaugrand et al. 1991;
Whitehouse 1997; Hsu and Wolf 1999) or actual fighting ability (Parker 1974). Although, a recent study found that individuals that accumulate multiple wins in their own territory
showed a full winner effect when they fought future battles in
that same territory, but showed a diminished winner effect
when they fought future battles in an unfamiliar location
(Fuxjager et al. 2009). This study illustrates that the environmental context of a fight can mediate the intrinsic properties
that underlie the ‘‘expression’’ of a fully formed winner effect;
this study does not address, however, whether the environmental context in which winning experience is accrued influences the ‘‘formation’’ of the winner effect in a given
individual.
I
Address correspondence to M.J. Fuxjager. E-mail: [email protected].
Received 7 May 2009; revised 30 September 2009; accepted 3
October 2009.
The Author 2009. Published by Oxford University Press on behalf of
the International Society for Behavioral Ecology. All rights reserved.
For permissions, please e-mail: [email protected]
Physiologically, steroid hormones are thought to regulate
the winner effect (Dugatkin 1997; Marler et al. 2005), particularly androgen hormones endogenously secreted after an aggressive social encounter (Gleason et al. 2009; Oliveira et al.
2009). Postconflict release of testosterone (T) is posited by
the ‘‘Challenge Hypothesis’’ (Wingfield et al. 1990), and numerous studies, including some in humans, have confirmed
that this phenomenon is taxonomically pervasive (see reviews
by Archer 2006; Hirschenhauser and Oliveira 2006). Other
steroid hormones, such as progesterone (P), that similarly mediate aggression (Kapusta 1998; Davis and Marler 2003; Weiss
and Moore 2004) are also hypothesized to influence plasticity
in winning ability (Davis and Marler 2003). Interestingly, research has indicated that the physical location or social context
of a fight can influence an animal’s postconflict androgen
responsiveness (Oliveira et al. 2005; Hirschenhauser et al.
2008; Fuxjager et al. 2009). This flexibility suggests a potential
mechanism through which extrinsic factors might mediate the
winner effect’s formation in a context-dependent manner
(Oliveira 2004; Gleason et al. 2009), although few studies have
begun to consider and study this possibility.
Another element of the winner effect that remains elusive is
the degree to which it and its underlying mechanisms vary
among species, particularly among those that are closely related. Understanding this variation is important because it will
likely provide a conceptual framework from which hypotheses
can be generated to address the selective pressures responsible for the evolution of the winner effect. For example, the
degree to which individuals of a given species are territorial
might predict the presence of the winner effect because this
enhancement in winning ability likely improves an individual’s competitiveness during aggressive encounters (Walls and
Semlitsch 1991; Kemp 2000; Perry et al. 2004). The few
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
Winning aggressive social encounters can enhance the probability of future victories. This so-called winner effect occurs in
diverse species and is thought to be an intrinsic phenomenon mediated by postencounter hormone release. Yet, recent evidence
suggests the possibility that certain extrinsic factors also influence the winner effect’s formation within an individual, possibly by
affecting the expression of hormone titers that follow a fight. We first investigated in the monogamous and territorial male
California mouse (Peromyscus californicus) whether the effect of residency, an extrinsic factor, influences the winner effect’s
formation and the endogenous secretion of testosterone and/or progesterone after a dispute. We found that California mice
that acquire winning experience in unfamiliar physical locations do not form a full winner effect. Furthermore, this species does
not experience a testosterone pulse after a fight in an unfamiliar cage. These findings indicate that environmental context can
mediate the winner effect’s formation, possibly by affecting the expression of postencounter testosterone pulses. Second, we
compared the winner effect of the California mouse to that of the white-footed mouse (Peromyscus leucopus), a close relative that is
promiscuous and less-territorial. We found that, compared with California mice, white-footed mice exhibit neither a full winner
effect, despite similar past winning experiences, nor a postencounter surge in testosterone. This finding suggests that these
behavioral and physiological phenomena vary among even closely related species and are possibly linked to aspects of social
biology, including the degree to which individuals of each species are territorial. Key words: aggression, Peromyscus, residency,
social behavior, testosterone, winner effect. [Behav Ecol 21:37–45 (2010)]
Behavioral Ecology
38
MATERIALS AND METHODS
Animals
California and white-footed mice were reared in laboratory
colonies at the University of Wisconsin—Madison. Animals
were maintained in accordance with the National Institutes of
Health Guide for the Care and Use of Laboratory Animals and
the appropriate institutional authorities at the University of
Wisconsin—Madison approved of the research described
herein. Each mouse colony was kept under a 14:10 h light:
dark cycle (lights on: 03:00 CST). After weaning, individuals
of the same species were housed in same-sex groups of 3 to 4
mice per cage (size: California mice- 48 3 27 3 16 cm; whitefooted mice- 28 3 18 3 12 cm). California and white-footed
mice were provided with Purina 5001 and 5015 mouse chow,
respectively and water ad libitum.
One week prior to the study’s onset, mice were transported
into testing rooms. Both species were never held in the same
testing room. To make observational work feasible for the experimenter, the 14:10 h light:dark cycle in each room was
shifted so that lights went on at 22:00 CST (see Trainor
et al. 2004; Oyegbile and Marler 2005, 2006; Fuxjager et al.
2009). Behavioral testing occurred during the dark phase under dim red light between 13:00 and 18:00. This study included 150 sexually naive male California mice, of which 46
were focal individuals, 60 were training intruders, and 44 were
testing intruders. This study also included 57 sexually naive
male white-footed mice, of which 24 were focal animals, 22
were training intruders, and 11 were testing intruders. All
animals were between 6 and 13 months old.
Behavioral testing
To test if residency status influences the winner effect’s formation in California mice, we randomly assigned sexually naive
male individuals to 1 of 4 different experimental conditions
(Table 1, n ¼ 11 per group): 1) three wins in the home cage
followed by a test encounter in the home cage, 2) three wins
in an unfamiliar cage followed by a test encounter in the
home cage, 3) three wins in an unfamiliar cage and a test
encounter in an unfamiliar cage, or 4) three handling experiences (i.e., no fights) in the home cage followed by a handling experience instead of a test encounter. Animals in the
last group were used as hormonal controls, in that we collected blood from these individuals in order to provide baseline measurements of circulating hormones. Details regarding
cage types and procedures are described below.
To test for species differences in the winner effect and
postencounter hormone release, we randomly assigned
white-footed mice to 1 of 2 experimental conditions (Table 1,
n ¼ 11 per group): 1) three wins in the home cage followed by
a test in the home cage (identical to group one above) or 2)
three handling experiences (i.e., no fights) followed by a handling experience in lieu of a test encounter. Again, mice that
experienced only handling were used as hormonal controls.
Once individuals were assigned to their respective treatment
groups, they were all treated in the same manner, described
below. On day 1, male mice were weighed and each was paired
with a female that had a small strip of fur shaved from her back
to ease identification (Trainor et al. 2004). On day 11, each
pair was moved from a small standard cage (see dimensions
above) to a larger, transparent Plexiglas cage (30 3 50 3 30
cm), which is referred to as the ‘‘home cage.’’ In each home
cage, a transparent wall divided the inside space into 2 separate compartments (30 3 29 3 30 cm and 22 3 29 3 30 cm).
Further, the dividing wall had 2 small openings at its base
(5 cm diameter) that allowed mice to freely pass between
compartments. For enrichment, a small wood nest-box was
placed in the same location of each home cage. Mouse chow
and water were also provided ad libitum.
Days 13, 15, and 17 consisted of the ‘‘training phase,’’
whereby focal individuals received either winning or handling
experiences. Winning experience occurred in either the focal
animal’s home cage or an unfamiliar cage, whereas handling
experience always occurred in the focal animal’s home cage.
Unfamiliar cages were constructed identically to home cages
(described above) and, prior to their use, were washed thoroughly and lined with fresh aspen chip bedding so that they
were completely unsoiled. Also, unfamiliar cages contained
Table 1
Six treatment groups to which male California and white-footed
mice were randomly assigned
Species
California mouse
White-footed mouse
Training phase
(days 13, 15, and 17)
Testing phase
(day 19)
3
3
3
3
3
3
Encounter in HC
Encounter in HC
Encounter in UC
Handle in HC
Encounter in HC
Handle in HC
wins in HC
wins in UC
wins in UC
handles in HC
wins in HC
handles in HC
Handled individuals were used as controls for hormonal analysis.
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
studies which have examined such species-level differences
typically compare results from separate studies (Oyegbile
and Marler 2006); yet, methodological differences among
assays of the winner effect potentially introduce a confounding factor that limits interpretation of the results (Hsu et al.
2006). To eliminate this confounding factor and determine
whether interspecific variation in the winner effect is associated with aspects of social biology, such as territoriality or
mating system, species comparisons should be made in the
same study and using identical methodology.
Here, we addressed 2 main questions surrounding the winner effect: 1) whether residency status influences both the winner effect’s formation and the hormonal mechanisms thought
to mediate the winner effect (i.e., postencounter T pulses); and
2) whether the winner effect and its underlying endocrine
mechanisms differ between 2 closely related species with different social systems. To address the first issue, we used the
California mouse (Peromyscus californicus), which is a monogamous and highly territorial species (Ribble and Salvioni 1990;
Bester-Meredith et al. 1999; Gubernick and Teferi 2000;
Trainor and Marler 2001; Oyegbile and Marler 2005). To address the second issue, we compared data obtained from
California mice with those obtained from white-footed mice
(Peromyscus leucopus), a close relative that is promiscuous and
less territorial (Wolff et al. 1983; Wolff 1985, 1986; Wolff and
Cicirello 1990; Xia and Millar 1991). We predicted that the
winner effect’s formation in male California mice is influenced by residency status because this environmental factor
can mediate the robustness of the winner effect’s expression
(Fuxjager et al. 2009). Similarly, we predicted that in California mice the hormonal changes known to mediate the winner
effect (i.e., postencounter T pulses) would occur only after
fights in the home cage, as evidence indicates that residency
influences postconflict physiological changes (Carre 2009;
Fuxjager et al. 2009). In regard to the species comparison,
we predicted that white-footed mice would show neither a robust winner effect nor a postencounter hormone titer, similar
to nonresident California mice, because white-footed mice are
significantly less territorial and aggressive (Wolff et al. 1983;
Wolff 1986).
Fuxjager and Marler
•
Environmental mediation of the winner effect
Figure 1
Percentage of focal mice that won the test encounter in each
treatment group. Notation below the horizontal axis denotes:
physical location of 3 prior wins; physical location of test encounter.
‘‘HC’’ ¼ home cage and ‘‘UC’’ ¼ unfamiliar cage. Treatment groups
that are significantly different have different letters above their bars.
intruders were randomly assigned to focal mouse opponents
and used only once.
Behavioral analysis
An observer blind to treatment groups analyzed the videotaped
test encounters, scoring for each individual in a given test encounter the frequency of bites, chases, wrestling bouts (i.e.,
a stint of wrestling lasting at least 3 s), jumps away from the
opponent, retreats from the opponent, and freezes (i.e.,
remaining completely still after initiating or receiving an attack). The observer also recorded each mouse’s attack latency
(i.e., time between the encounter’s onset and an individual’s
first attack toward its opponent). Finally, the observer calculated for all individuals the total attacks directed toward the
opponent (i.e., sum of bites, chases, and wrestling bouts)
and the total losing behavior elicited by the opponent (i.e.,
sum of jumps away, retreats, and freezes). Thus, the ‘‘total
attacks’’ is a relative measure of an individual’s aggressiveness
during the test encounter, whereas the ‘‘total losing behavior’’
is a relative measure of subordinate or losing behavior.
Hormonal measurements
Focal individuals were rapidly decapitated 45 min after the test
encounter so that blood from the thoracic cavity (i.e., trunk
blood) and the brain (used for another study) could be collected. Past studies indicated that this time point is when postencounter plasma T peaks in Peromyscus mice (Marler et al.
2005; Oyegbile and Marler 2005). Samples were immediately
centrifuged and stored at 280 C until assayed. Hormone
assays were conducted at the Wisconsin Regional Primate Center. Samples were extracted with ethyl ether, and steroids were
separated using celite chromatography. All hormones were
analyzed separately and in a manner described in detail previously (Bester-Meredith et al. 1999; Trainor and Marler 2001;
Davis and Marler 2003). Briefly, T and P were analyzed separately using enzyme immunoassays (T antibody R156, University of California—Davis diluted to 1:35 000; P antibody
R4861, University of California—Davis, diluted to 1:33 000).
Corticosterone (Cort) samples were analyzed using radioimmunoassay (validated for California mice elsewhere [BesterMeredith and Marler 2001]). The intra- and interassay coefficients of variation for each hormone measured, respectively,
were as follows: T, 2.9% and 4.3% (n ¼ 2 plates); P, 1.3% and
2.3% (n ¼ 2 plates); and Cort, 0% and 3.1% (n ¼ 1 assays).
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
food and a clean nest-box in the larger compartment, although
the latter was placed in a different physical location within
the unfamiliar cage compared with the home cage. Thus, unfamiliar cages differed from home cages in their scent and
arrangement of objects within the cage. Focal mice neither
experienced more than one fight in the same unfamiliar cage
nor encountered the same nest-box during a fight.
Aggressive contests followed the resident–intruder (R-I) paradigm described previously (Trainor et al. 2004; Oyegbile and
Marler 2005, 2006). For encounters in the home cage, the
female was first removed and an opaque divider that blocked
passage and sight between compartments was inserted into
the cage. Focal individuals were kept in the cage’s larger compartment, and intruders were placed in the smaller compartment. The 2 individuals remained separated for 2 min so that
the intruder had time to acclimate to its surroundings. The
divider was then removed and the mice were permitted to
interact for 10 min. For aggressive encounters in the unfamiliar cage, the R-I procedure was slightly modified, such that
focal animals were removed from their home cage and placed
into the unfamiliar cage’s larger compartment with the opaque divider already in place. The opponent was then placed in
the adjacent compartment and allowed to acclimate for 2 min
as described above.
The contest winner was defined as the individual that initiated at least 3 consecutive attacks (described below) that
elicited either avoidance or freezing behavior (described below) from the opponent (Oyegbile and Marler 2005, 2006).
Previous studies of Peromyscus behavior operationally defined
winning behavior as initiating multiple attacks of biting, chasing, and wrestling, whereas losing behavior was defined as
jumping away from an opponent, retreating, and freezing
(Eisenberg 1961). To bias the training encounters so that
focal individuals always won and therefore received the
proper treatment, training intruders were sexually inexperienced, smaller, and losers of prior encounters (Trainor et al.
2004; Oyegbile and Marler 2005). Each training intruder was
randomly assigned to its focal mouse opponent and never
fought the same focal mouse more than once. Training intruders were used only twice for the entire experiment. Focal
mice that did not clearly win all 3 of their training encounters
were removed from the study (California mice: 5.5%; whitefooted mice: 55%).
Focal individuals that experienced handling instead of
an aggressive encounter experienced the same treatment imparted by the experimenter and the R-I paradigm; however, an
intruder was never placed in the home cage. As such, when the
opaque divider was lifted, the focal mouse did not have any
behavioral interactions with a male conspecific.
On day 19, focal mice were subjected to a test encounter, in
either their home cage or another unfamiliar cage (Table 1),
which allowed for the assessment of the strength and magnitude of a winner effect (Oyegbile and Marler 2005, 2006).
Test encounters followed the same R-I paradigms described
above, except in this case the 10 min encounter was videotaped. Also, to reduce the effect of contest asymmetries
among treatment groups during the test encounter, resources
(i.e., nest-box and food) were removed from the home cages
or were absent from unfamiliar cages. The probability of intruders winning in the test encounters was increased compared with training sessions because testing intruders were
sexually experienced, slightly larger (mean 6 standard error;
1.4 6 0.22 g), and had won a single encounter on the previous
day. In effect, this ‘‘intruder advantage’’ decreased the probability that focal mice randomly win their test encounter
(;20%, see Oyegbile and Marler 2005) and therefore
enhanced our ability to statistically detect a winner effect
(Trainor et al. 2004; Oyegbile and Marler 2005, 2006). Test
39
Behavioral Ecology
40
Table 2
Aggressive behavior exhibited by focal individuals or their respective opponents during the test encounter
White-footed
mouse
California mouse
Behavior
Attack latency (s)*
Total attacks*
Total opponent
losing behavior*
3 wins in HC;
test in HC
3 wins in UC;
test in HC
3 wins in UC;
test in UC
3 wins in HC;
test in HC
6.5 6 2.5a
63.0 6 12.0a
39.0 6 10.0a
71.7 6 38.0a
34.7 6 6.1b
6.2 6 3.9b
15.7 6 7.7a
11.3 6 5.0c
7.0 6 4.7b
8.8 6 5.6a
65.5 6 7.4a
23.5 6 4.1a
Statistical analyses
Aggressive behavior
Analyses were based on data obtained during the test encounters and from focal individuals, the one exception being analysis of losing behavior elicited from the opponents of focal
mice. The relative differences in winning ability among treatment groups were analyzed with the Fisher’s Exact test modified for data arranged in a 2 3 4 table (Freeman and Halton
1951). Post hoc comparisons between treatment groups were
conducted with the Fisher’s Exact test, using Holm’s procedure for multiple comparisons to control type I error rate
(Holm 1979). The Fisher’s Exact test was also used to compare the number of individuals of each species that were removed from the experiment during the training phase.
Remaining behavioral and hormonal data were log transformed to meet conditions of normality (Zar 1999) and
analyzed using SPSS 16.0 (Chicago, IL). We used a series of
one-way ANOVAs to test if experimental treatments influence
aggression and postencounter hormone levels, and post hoc
analyses were conducted with Student-Newman-Keuls tests.
We calculated Pearson’s correlation coefficients to test if any
given behavior or hormone was correlated with postencounter
T, P, or Cort (Zar 1999).
The physical location in which both the prior wins and the test
encounter occurred significantly affected the level of aggressive behavior focal mice displayed toward their opponents
and the degree of losing behavior focal mice elicited from their
opponents (Table 2; attack latency: F3,37 ¼ 2.94, P , 0.05; total
number of attacks: F3,37 ¼ 13.94, P , 0.01; total number of
opponent losing behaviors: F3,36 ¼ 8.44, P , 0.01). For attack
latency, there were no significant contrasts among any given
treatment groups. In regard to total attacks, California and
white-footed mice that had winning experience and were
tested in the home cage attacked opponents more than individuals that acquired winning experience in an unfamiliar
cage (Table 2; California mouse HC;HC vs. UC;HC: q ¼
4.16, P , 0.05; California mouse HC;HC vs. UC;UC:
q ¼ 9.22, P , 0.01; white-footed mouse HC;HC vs. UC;UC:
q ¼ 10.61, P , 0.01; Table 2; white-footed mouse HC;HC vs.
UC;HC: q ¼ 5.56, P ¼ 0.01). California mice that acquired
winning experience in an unfamiliar cage but had a test encounter in the home cage attacked opponents more than
those individuals that both acquired wins and had a test encounter in an unfamiliar cage (Table 2; UC;HC vs. UC;UC:
q ¼ 5.05, P , 0.05). Lastly, regarding opponent losing behavior, California and white-footed mice that both acquired wins
and were tested in the home cage elicited more losing behavior
from their opponents than individuals in the other treatment
groups (Table 2; California mouse HC;HC vs. UC;HC: q ¼ 5.23,
P , 0.01; California mouse HC;HC vs. UC;UC: q ¼ 3.99, P ,
0.05; white-footed mouse HC;HC vs. UC;HC: q ¼ 4.94, P , 0.01;
white-footed mouse HC;HC vs. UC;UC: q ¼ 3.75, P , 0.05).
RESULTS
Winner effect
The physical location in which California mice acquired winning experience significantly influenced an individual’s ability
to form a full and robust winner effect (Figure 1; P , 0.005).
Analyses showed that California mice that acquire wins in an
unfamiliar cage, regardless of the test encounter’s location,
showed a diminished winner effect compared with California
mice that acquire prior winning experience in the home cage
(Figure 1; California mouse HC;HC vs. UC;HC: P , 0.004,
California mouse HC;HC vs. UC;UC: P , 0.004). Furthermore, white-footed mice with past winning experience in
the home cage also showed a diminished winner effect compared with California mice that acquired wins and a test encounter in the home cage (Figure 1; California mouse HC;HC
vs. white-footed mouse HC;HC: P , 0.0125). As such, whitefooted mice were behaviorally more similar to California mice
that had winning experiences in unfamiliar cages, as there were
not significant differences between these groups (Figure 1; all
comparisons P ¼ 1.0).
Additional analyses demonstrated that significantly fewer
white-footed mice (11 of 24, ;45%) were able to win the 3 consecutive training encounters compared with California mice
(33 of 35, ;95%) (P , 0.005).
Hormones
Analyses show that the physical location in which a test encounter takes place affects circulating T after the fight (Figure 2A;
F5,60 ¼ 5.00, P , 0.01). Post hoc tests revealed that California
mice fighting a test encounter in the home cage, regardless of
the location in which prior wins were accrued, had significantly higher plasma T 45 min after the dispute compared
with baseline levels (Figure 2A, California mice only; HC;HC
vs. Handled Control: q ¼ 4.10, P , 0.05; UC;HC vs. Handled
Controls: q ¼ 5.49, P , 0.01). Furthermore, California mice
experiencing a test encounter in an unfamiliar cage did not
show a detectable difference in postencounter plasma T compared with baseline (Figure 2A, California mice only; UC;UC
vs. Handled Controls: q ¼ 0.97, P . 0.90). Unlike California
mice, postencounter plasma T in white-footed mice fighting in
the home cage was not different from this species’ basal level
(Figure 2A, white-footed mice only: HC;HC vs. Handled
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
Data are displayed as means 6 standard error, but were natural log transformed for statistical analyses. ‘‘HC’’ means
home cage and ‘‘UC’’ means unfamiliar cage. The asterisks adjacent to each behavior indicate an overall significant
difference among treatment groups. Significant differences between values of a given behavior are indicated by
differences in superscript letters.
Fuxjager and Marler
•
Environmental mediation of the winner effect
41
winner effect (i.e., enhanced ability to win aggressive contests
following past victories, reviewed by Hsu et al. 2006). This
result demonstrates that the physical environment in which
individuals acquire wins can influence the formation of the
winner effect. Furthermore, these findings are consistent with
our physiological data showing that postencounter T pulses
responsible for mediating the winner effect (Gleason et al.
2009; Oliveira et al. 2009) are observed in California mice
only after they fight in their familiar home cage. In contrast
to the California mouse, the closely related white-footed
mouse does not show a robust winner effect nor a postencounter T pulse, which might explain why this species does
not exhibit a full winner effect.
Winning ability and aggressive behavior
Controls: q ¼ 2.11, P . 0.50). Interestingly, T levels in whitefooted mice were highly variable, although this pattern of variability is consistent with that observed in white-footed mice in
previous studies (Marler et al. 2003; Oyegbile and Marler
2006).
Additional analyses of postencounter P and Cort showed that
while neither hormone changed after a fight relative to its respective baseline level, significant species differences in basal values were detected (Figure 2B,C; P: F5,60 ¼ 21.11, P , 0.01; Cort:
F5,60 ¼ 8.17, P , 0.01); thus, California mice had higher P
(Figure 2B; California mouse groups vs. white-footed mouse
groups: q . 7.60, P , 0.01) and Cort (Figure 2C; California
mouse groups vs. white-footed mouse groups: q . 3.26, P ,
0.05) than white-footed mice.
Correlations
In both California and white-footed mice, there was no significant correlation between any given behavior and T, P, or Cort
(Table 3). However, in California mice, there was a significant
positive correlation between postencounter P and Cort levels
(Table 4; r ¼ 0.653, slope ¼ 0.43, P , 0.001). We detected no
other significant hormone–hormone correlations (Table 4).
DISCUSSION
In California mice, winning experience accrued in an unfamiliar physical location tempers the formation of a full and robust
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
Figure 2
Amount of circulating testosterone (A), progesterone (B), and
corticosterone (C) in focal individuals 45 min after the test
encounter. Notation below the horizontal axis denotes: physical
location of 3 prior wins; physical location of test encounter. ‘‘HC’’ ¼
home cage and ‘‘UC’’ ¼ unfamiliar cage. ‘‘Handled Controls’’ are
those animals used to measure baseline hormone levels because they
experienced no fights. Treatment groups that are significantly
different have different letters above their bars. Bars represent
a mean value and error bars represent the standard error.
In the territorial California mouse, the winner effect’s full formation occurs only if past wins are acquired in a familiar home
cage. Previous research showed that the winner effect is largely
an intrinsic phenomenon, (Beacham 1988; Beaugrand et al.
1991, 1996; Whitehouse 1997; Hsu and Wolf 1999; Hsu et al.
2006), but our data are provocative because they suggest that
these intrinsic properties can be governed by the environmental context in which winning experience occurs. Another experiment in California mice shows that the expression of an
otherwise fully formed winner effect is similarly influenced by
a fight’s physical location (Fuxjager et al. 2009). Although it is
possible that the effect of contest location interacts with an
effect of social context to influence the winner effect, both
the results from Fuxjager et al. (2009) and those presented
here illustrate that contest location is a potent factor that
influences both how the winner effect forms within an individual and how it is expressed in the future. To this end, our
data illustrate an intriguing property of the winner effect that
has not yet been described: it is likely bound to the environment in which it was formed. Studies in male crickets (Gryllus
bimaculatus) have shown a similar phenomenon regarding
plasticity in fighting behavior; namely, the depressed fighting
ability a male cricket normally experiences after it loses a fight
is muted once it engages in flight and presumably arrives at
a different location (Hofmann and Stevenson 2000). Future
studies should examine how long the winner effect lasts in an
individual, particularly in territorial species since the dynamics
that guide territoriality are flexible and, in some cases, change
between breeding seasons (Vitousek et al. 2008).
The data presented here build further support for a connection between the winner effect and the residency effect, the
latter of which is defined as an enhanced ability for territory
owners to win contests that occur on their home range or
territory (Krebs 1982; Waage 1988; Olsson and Shine 2000).
Although the functional significance of this relationship
requires more investigation, it is possible that the residencydependent nature of the winner effect is related to territoriality (Oyegbile and Marler 2006; Fuxjager et al. 2009). For
example, residency might be a cue that triggers the formation
of the winner effect, such that an individual’s ability to win
aggressive contests is enhanced during instances of territory
settlement or defense. At the same time, this mechanism presumably allows individuals to avoid investing energy into costly
aggressive encounters that are unrelated to territoriality and
therefore most likely offer fewer benefits associated with victory. Indeed, this interpretation is consistent with the comparative data present herein, in that the white-footed mouse,
which is less territorial and wanders in search of potential
mates (Wolff 1986; Wolff and Cicirello 1990), does not exhibit
as robust of a winner effect as the highly territorial and monogamous California mouse (Ribble and Salvioni 1990; Oyegbile and Marler 2005). Furthermore, the observed species
Behavioral Ecology
42
Table 3
Statistical information regarding correlations between any given behavioral measure and postencounter hormone level for each species
Correlated variables
Hormone
Behavior
Correlation
coefficient (r)
P value
California mouse
Testosterone
Attack latency (s)
Total attacks
Total opponent losing
Attack latency (s)
Total attacks
Total opponent losing
Attack latency (s)
Total attacks
Total opponent losing
Attack latency (s)
Total attacks
Total opponent losing
Attack latency (s)
Total attacks
Total opponent losing
Attack latency (s)
Total attacks
Total opponent losing
0.020
0.211
20.080
20.055
0.199
20.055
20.101
0.188
0.094
0.148
0.060
20.17
20.074
20.485
20.497
0.177
0.006
0.142
0.92
0.26
0.68
0.77
0.29
0.78
0.60
0.32
0.63
0.66
0.86
0.62
0.829
0.13
0.12
0.60
0.99
0.68
Progesterone
Corticosterone
White-footed mouse
Testosterone
Progesterone
Corticosterone
behavior
behavior
behavior
behavior
behavior
behavior
No correlations are significant (P , 0.05).
differences in the winner effect are not due to methodological
biases (Hsu et al. 2006) because individuals of each species
were examined in the same experiment and using the same
protocol. Finally, although we hypothesize that territoriality
and mating strategy are potential factors contributing to observed interspecific divergence in the winner effect, other
factors might also influence this process. For example, in
some species, the winner effect functions in part to help shape
emerging social structures, such as dominance hierarchies
(Dugatkin 1997; Bergman et al. 2003).
Our data also show that the level of aggressive behavior individuals exhibit during the test encounter varies among treatment groups and species, but not necessarily in the manner
we might otherwise expect if aggressive behavior is always directly associated with winning ability. This is clearly demonstrated by our finding that California mice trained and
tested in their home cage win more test encounters than
white-footed mice with identical treatment and experience;
however, focal animals in both groups do not show significant
differences in either the total number of attacks directed at an
opponent or the total number of losing behaviors elicited from
an opponent. These data suggest that the winner effect is not
reducible to simple changes in aggressive behavior that occur
as a result of winning experience or fighting in a home cage.
Instead, it is likely that other species-specific and dynamic
properties between 2 individuals engaged in a fight underlie
this behavioral phenomenon (Oliveira et al. 2009), which
lends credence to the argument that the winner effect is an
evolved adaptation rather than a by-product of regulatory
processes guiding aggressive behavior (Rutte et al. 2006).
Postencounter hormone levels
Postencounter T increases only after a fight in the home cage.
Previous research suggested this possibility (Fuxjager et al.
2009), although this work showed neither that postencounter T ‘‘increased’’ relative to basal levels, nor that it
was a response to the aggressive interaction itself. We experimentally addressed these limitations and showed that
T does in fact increase after an aggressive social interaction
in a context-specific manner. Furthermore, our results suggest that the expression of a postencounter T pulse is not
strongly influenced by the location and context of past victories, implying that such experiences have little or no bearing on postencounter T responsiveness. However, this
implication should be interpreted cautiously, not only because previous work has suggested that past winning experiences do in fact influence postconflict T responsiveness
(Oyegbile and Marler 2005), but also because we cannot
entirely rule out the effect of intrinsic factors on T responsiveness (i.e., males might adjust postencounter hormone
responses with prospect to future encounters or there might
be an effect of a male’s ability to mount a postencounter
T response). We also note that there are potential differences in the hormonal milieu and responsiveness between
animals living in the laboratory and the field (Calisi and
Table 4
Statistical information regarding hormone–hormone correlations for each species
Species
Correlated variables
Correlation coefficient (r)
P value
California mouse
Testosterone versus progesterone
Testosterone versus corticosterone
Progesterone versus corticosterone
Testosterone versus progesterone
Testosterone versus corticosterone
Progesterone versus corticosterone
0.314
0.150
0.653
0.014
20.074
20.245
0.076
0.41
0.001*
0.98
0.83
0.47
White-footed mouse
Asterisks (*) denote significant correlations (P , 0.05).
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
Species
Fuxjager and Marler
43
Bentley 2009), although a number of field and laboratory
studies alike have demonstrated postcontest T pulses
(Hirschenhauser et al. 2003; Hirschenhauser and Oliveira 2006).
In regard to California mice, our data suggest an interesting
caveat to the Challenge Hypothesis (Wingfield et al. 1990), in
that contextual cues, such as a fight’s location, can help mediate within-individual variation in the expression of postencounter T pulses. Because this physiological phenomenon in
part underlies the winner effect’s formation (Gleason et al.
2009; Oliveira et al. 2009), our data provide strong support for
the hypothesis that environmental context can regulate the
winner effect’s formation in a given individual by affecting
steroid hormone action. Our findings join a larger body of
work that shows that other factors, like the social context of
a fight, similarly modulate expression of postencounter T
pulses (Oliveira et al. 2001, 2005; Hirschenhauser et al.
2008) and therefore possibly also the winner effect (Oliveira
2004; Gleason et al. 2009). Other hypotheses propose that
postencounter T helps regulate the persistence of territorial
behavior (Wingfield 1994) and/or territorial site preferences
in the wild (Marler et al. 2005; Gleason et al. 2009). Again, our
findings are consistent with these different but nonmutually
exclusive hypotheses, as the territorial California mouse,
which demonstrates a clear winner effect, releases T after
fights that are linked to territorial defense; whereas the less
territorial white-footed mouse, which shows a weak winner
effect, does not release T after disputes in the home cage.
Other reasons might explain why white-footed mice do not
show changes in postconflict T changes, such as the influence
of mating system. For instance, phylogenetic analyses in birds
have indicated that polygynous species show lower androgen
responsiveness to aggressive social encounters compared with
monogamous species (Wingfield et al. 1990, 2000; Hirschenhauser et al. 2003). Our results also conform to this pattern
because white-footed mice mate with multiple partners (Xia
and Millar 1991) and California mice demonstrate strict
monogamy (Ribble and Salvioni 1990; Gubernick and Teferi
2000).
For both species, P levels did not change after aggressive contests in any of the treatment groups, limiting potential interpretation of P’s effect on aggression and winning ability.
Recent work in diverse species, including California mice,
has shown that P changes during or sometimes after aggressive
encounters (Kapusta 1998; Davis and Marler 2003; Weiss and
Moore 2004; Rubenstein and Wikelski 2005; Fuxjager et al.
2009). Yet, the role P plays in mediating aggression or winning
ability remains elusive, in part because its effects vary substantially among species and between sex. For instance, in female
rodents P can enhance aggression (Svare et al. 1986; Meisel
and Sterner 1990), whereas in male rodents it can suppress
aggression (Erpino and Chappelle 1971; Gravance et al.
1996). The issue becomes more complex considering that
separate studies often use different techniques to stage aggressive interactions and therefore alter the contextual stimuli
associated with a given contest. This is probably an important
point when trying to understand how P regulates aggression
because this hormone is involved in mediating other physiological phenomena often associated with aggressive contests,
such as anxiety (Schino 1998; Reddy et al. 2005; Schino et al.
2007). Furthermore, in California mice, there is a positive
association between postencounter P and Cort. This indicates
that adrenal glands might contribute to the postencounter P
levels (Resko 1969) because this is the primary source of Cort
secretions (Singer and Stack-Dunne 1955). To our knowledge,
it is unknown whether the site of P release predicts its physiological function, although this adds to the interesting questions about P’s role in regulating aggressive behavior and
winning ability.
Our results suggest that Cort does not change after an aggressive interaction in either species, even when that interaction occurs in a novel physical setting. Because Cort is often
associated with increased stress (Nelson 2000), our data are
consistent with the hypothesis that, at least in male California
mice, aggressive encounters in a variety of contexts do not
dramatically elevate physiological measures of stress.
Finally, the data presented here indicate species differences
in basal steroid hormone levels. Compared with California
mice, white-footed mice have lower basal levels of P and Cort,
and probably higher basal T. The latter finding is a nonsignificant trend that has been statistically confirmed in a prior study
(Marler et al. 2003). Hormonal mechanisms that serve as
a physiological basis for life-history traits are likely to be targets of natural selection (Hau 2007; Bokony et al. 2009); thus,
it is not surprising that basal hormones differ between California and white-footed mice, because these species show different degrees of territoriality and strategies of acquiring
mates. In regard to androgens, a phylogenetic analysis in Peromyscus mice of baseline T suggests that species differences are
more likely to be related to ecological constraints imposed by
the surrounding habitat rather than a given mating system
(Marler et al. 2003). Certainly, this line of thinking agrees with
additional analyses in birds that suggest that circulating hormones like Cort might be elevated in individuals when the risk
of being stressed is both high and foreseeable (Romero 2002;
Bokony et al. 2009). Yet, our results should be viewed in the
context of broader phylogenetic analyses with caution because
few experiments have been able to assess baseline hormones
among a wide range of free-living mammals (Romero 2002),
and differences in endocrine patterns might exist among
vertebrate classes.
CONCLUSION
We show that extrinsic factors, namely the effect of residency
status, influence the formation of the full winner effect in California mice. This process might occur as a result of extrinsic
regulation of hormone release after an aggressive interaction,
providing evidence for an interesting mechanism through
which environmental stimuli can regulate plasticity in winning
behavior in a context-specific manner. Furthermore, we show
interspecific variation in the winner effect and its corresponding mechanisms, which might be due to species differences in
territoriality and strategies used to acquire mates. This fact supports the notion that the winner effect is an evolved adaptation.
FUNDING
National Science Foundation Graduate Research Fellowship
(to M.J.F.); National Science Foundation Grant (IOS-0620042
to C.A.M.).
We thank Jin Park and Andrea Charles for their assistance with the execution of the study. We thank Kyla Davidoff, Jan Davidoff, Katherine
Cronin, Erin Gleason, Elizabeth Becker, and Josh Pultorak for helpful
comments regarding the manuscript. We also thank Dan Wittwer and
the Wisconsin Primate Research Center for conducting the hormone
assays.
REFERENCES
Archer J. 2006. Testosterone and human aggression: an evaluation of
the challenge hypothesis. Neurosci Behav Rev. 30:319–345.
Beacham JL. 1988. The relative importance of body size and aggressive experience as determinants of dominance in pumpkinseed
sunfish, Lepomis gibbosus. Anim Behav. 36:621–623.
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
Environmental mediation of the winner effect
•
44
Kemp DJ. 2000. Contest behavior in territorial male butterflies: does
size matter? Behav Ecol. 11:591–596.
Krebs JR. 1982. Territorial defense in the great tit (Parus major): do
residents always win? Behav Ecol Sociobiol. 11:185–194.
Marler CA, Bester-Meredith JK, Trainor BC. 2003. Paternal behavior
and aggression: endocrine mechanisms and nongenomic transmission of behavior. Adv Study Behav. 32:263–323.
Marler CA, Oyegbile TO, Plavicki J, Trainor BC. 2005. Response to
Wingfield’s commentary on ‘‘A continuing saga: the role of testosterone in aggression’’. Horm Behav. 48:256–258.
Meisel RL, Sterner MS. 1990. Progesterone inhibition of sexual behavior is accompanied by an activation of aggression in female Syrian
hamsters. Physiol Behav. 47:415–417.
Nelson RJ. 2000. An introduction to behavioral endocrinology. Sunderland (UK): Sinauer Associates Inc.
Oliveira RF. 2004. Hormones, social context and animal communication. In: McGregor PK, editor. Animal communication networks.
Cambridge: Cambridge University Press. p. 481–520.
Oliveira RF, Carneiro LA, Canario AVM. 2005. No hormonal response
in tied fights. Nature. 437:207–208.
Oliveira RF, Lopos M, Carneiro LA, Canario AVM. 2001. Watching
fights raises fish hormone levels. Nature. 409:784–784.
Oliveira RF, Silva A, Canario AVM. 2009. Why do winners keep winning? Androgen mediation of winner but not loser effects. Proc R
Soc Lond Ser B Biol Sci. 276:2249–2256.
Olsson M, Shine R. 2000. Ownership influences the outcome of male–
male contests in the scincid lizard, Niveoscincus microlepidotus. Behav
Ecol. 11:587–590.
Oyegbile TO, Marler CA. 2005. Winning fights elevates testosterone
levels in California mice and enhances future ability to win fights.
Horm Behav. 48:259–267.
Oyegbile TO, Marler CA. 2006. Weak winner effect in a less aggressive
mammal: correlations with corticosterone but not testosterone.
Physiol Behav. 89:171–179.
Parker GA. 1974. Assessment strategy and evolution of fighting behavior. J Theor Biol. 47:223–243.
Perry G, Levering K, Girard I, Garland T. 2004. Locomotor performance
and social dominance in male Anolis cristatellus. Anim Behav. 67:37–47.
Reddy DS, O’Malley BW, Rogawski MA. 2005. Anxiolytic activity of
progesterone in progesterone receptor knockout mice. Neuropharmacology. 48:14–24.
Resko JA. 1969. Endocrine control of adrenal progesterone secretion
in the ovariectomized rat. Science. 164:70–71.
Ribble DO, Salvioni M. 1990. Social organization and nest-co-occupancy
in Peromyscus californicus, a monogamous rodent. Behav Ecol Sociobiol. 26:9–15.
Romero LM. 2002. Seasonal changes in plasma glucocorticoid concentrations in free-living vertebrates. Gen Comp Endocrinol. 128:1–24.
Rubenstein DR, Wikelski M. 2005. Steroid hormones and aggression
in female Galapagos marine iguanas. Horm Behav. 48:329–341.
Rutte C, Taborsky M, Brinkhof MWG. 2006. What sets the odds of
winning and losing? Trends Ecol Evol. 21:16–21.
Schino G. 1998. Reconciliation in goats. Behaviour. 135:343–356.
Schino G, Rosati L, Geminiani S, Aureli F. 2007. Post-conflict anxiety
in Japanese macaques (Macaca fuscata) aggressor’s and victim’s perspectives. Ethology. 113:1081–1088.
Singer B, Stack-Dunne MP. 1955. The secretion of Aldosterone and
corticosterone by the rat adrenal. J Endocrinol. 12:130–145.
Svare B, Miele J, Kinsley C. 1986. Mice: progesterone stimulates aggression in pregnancy-terminated females. Horm Behav. 20:194–200.
Trainor BC, Bird IM, Marler CA. 2004. Opposing hormonal mechanisms of aggression revealed through short-lived testosterone manipulations and multiple winning experiences. Horm Behav. 45:
115–121.
Trainor BC, Marler CA. 2001. Testosterone, paternal behavior, and
aggression in the monogamous California mouse (Peromyscus californicus). Horm Behav. 40:32–42.
Vitousek MN, Rubenstein DR, Nelson KN, Wikelski M. 2008. Are hotshots always hot? A longitudinal study of hormones, behavior, and
reproductive success in male marine iguanas. Gen Comp Endocrinol. 157:227–232.
Waage JK. 1988. Confusion over residency and the escalation of damselfly territorial disputes. Anim Behav. 36:586–595.
Walls SC, Semlitsch RD. 1991. Visual and movement displays function
as agonistic behavior in larval salamanders. Copeia. 1991:936–942.
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014
Beaugrand J, Goulet C, Payette D. 1991. Outcome of dyadic conflict in
male green swordtail fish, Xiphophorus helleri—effects of body size
and prior dominance. Anim Behav. 41:417–424.
Beaugrand JP, Payette D, Goulet C. 1996. Conflict outcome in male
green swordtail fish dyads (Xiphophorus helleri): interaction of body
size, prior dominance/subordination experience, and prior residency. Behaviour. 133:303–319.
Bergman DA, Kozlowski CP, McIntyre JC, Huber R, Daws AG, Moore
PA. 2003. Temporal dynamics and communication of winner-effects
in the crayfish, Orconectes rusticus. Behaviour. 140:805–825.
Bester-Meredith JK, Marler CA. 2001. Vasopressin and aggression in
cross-fostered California mice (Peromyscus californicus) and whitefooted mice (Peromyscus leucopus). Horm Behav. 40:51–64.
Bester-Meredith JK, Young LJ, Marler CA. 1999. Species differences in
paternal behavior and aggression in Peromyscus and their associations with vasopressin immunoreactivity and receptors. Horm
Behav. 36:25–38.
Bokony V, Lendvai AZ, Liker A, Angelier F, Wingfield JC, Chastel O.
2009. Stress response and the value of reproduction: are birds prudent parents? Am Nat. 173:589–598.
Calisi RM, Bentley GE. 2009. Lab and field experiments: are they the
same animal? Horm Behav. 56:1–10.
Carre JM. 2009. No place like home: testosterone responses to victory
depend on game location. Am J Hum Biol. 21:392–394.
Davis ES, Marler CA. 2003. The progesterone challenge: steroid hormone changes following a simulated territorial intrusion in female
Peromyscus californicus. Horm Behav. 44:185–198.
Dugatkin LA. 1997. Winner and loser effects and the structure of
dominance hierarchies. Behav Ecol. 8:583–587.
Dugatkin LA, Druen M. 2004. The soical implications of winner and
loser effects. Biol Lett. 271:S488–S489.
Eisenberg JF. 1961. Studies of the behavior of Peromyscus maniculatus
gambelii and Peromyscus californicus parasiticus. Behaviour. 19:177–207.
Erpino MJ, Chappelle TC. 1971. Interactions between androgens and
progesterone in mediation of aggression in the mouse. Horm
Behav. 2:265–272.
Freeman GH, Halton JH. 1951. Note on the exact treatment of contingency, goodness of fit and other problems of significance.
Biometrika. 38:141–149.
Fuxjager MJ, Mast G, Becker EA, Marler CA. 2009. The ‘home advantage’ is necessary for a full winner effect and changes in postencounter testosterone. Horm Behav. 56:214–219.
Gleason ED, Fuxjager MJ, Oyegbile TO, Marler CA. 2009. Tesosterone
release and social context: when it occurs and why. Front Neuroendocrinol. 30:460–469.
Gravance CG, Casey PJ, Erpino MJ. 1996. Progesterone does not inhibit aggression induced by testosterone metabolites in castrated
male mice. Horm Behav. 30:22–25.
Gubernick DJ, Teferi T. 2000. Adaptive significance of male parental
care in a monogamous mammal. Proc R Soc Lond Ser B Biol Sci.
267:147–150.
Hau M. 2007. Regulation of male traits by testosterone: implications
for the evolution of vertebrate life histories. Bioessays. 29:133–144.
Hirschenhauser K, Oliveira RF. 2006. Social modulation of androgens
in male vertebrates: meta-analyses of the challenge hypothesis.
Anim Behav. 71:265–277.
Hirschenhauser K, Winkler H, Oliveira RF. 2003. Comparative analysis
of male androgen responsiveness to social environment in birds:
the effects of mating system and paternal incubation. Horm Behav.
43:508–519.
Hirschenhauser K, Wittek M, Johnston P, Mostl E. 2008. Social context
rather than behavioral output or winning modulates post-conflict
testosterone responses in Japanese quail (Coturnix japonica). Physiol
Behav. 95:457–463.
Hofmann HA, Stevenson PA. 2000. Flight restores fight in crickets.
Nature. 403:613.
Holm S. 1979. A simple sequentially rejective multiple test procedure.
Scand J Stat. 6:65–70.
Hsu YY, Earley RL, Wolf LL. 2006. Modulation of aggressive behaviour
by fighting experience: mechanisms and contest outcomes. Biol
Rev. 81:33–74.
Hsu YY, Wolf LL. 1999. The winner and loser effect: integrating multiple experiences. Anim Behav. 57:903–910.
Kapusta J. 1998. Gonadal hormones and intrasexual aggressive behavior in
female bank voles (Clethrionomys glareolus). Aggressive Behav. 24:63–70.
Behavioral Ecology
Environmental mediation of the winner effect
45
Weiss SL, Moore MC. 2004. Activation of aggressive behavior by progesterone and testosterone in male tree lizards, Urosaurus ornatus.
Gen Comp Endocrinol. 136:282–288.
Whitehouse MA. 1997. Experience influences male–male contests in
the spider Argyrodes antipodiana (Theridiidae: Araneae). Anim Behav. 53:913–923.
Wingfield JC. 1994. Regulation of territorial behavior in the sedentary song sparrow, Melospiza melodia morphna. Horm Behav. 28:
1–15.
Wingfield JC, Hegner RE, Dufty AM, Ball GF. 1990. The challenge
hypothesis—theoretical implications for patterns of testosterone
secretion, mating systems, and breeding Strategies. Am Nat.
136:829–846.
Wingfield JC, Jacobs JD, Tramontin AD, Perfito N, Meddle S, Maney
DL, Soma KK. 2000. Toward an ecological basis of hormone-behavior interactions in reproduction of birds. In: Wallen K,
Schneider JE, editors. Reproduction in context, social and envi-
ronmental influences on reproduction. Cambridge (MA): MIT
Press. p. 85–128.
Wolff JO. 1985. The effects of density, food, and interspecific interference on home range size in Peromyscus leucopus and Peromyscus maniculatus. Can J Zool. 63:2657–2662.
Wolff JO. 1986. Life history strategies of white-footed mice (Peromyscus
leucopus). Va J Sci. 37:208–220.
Wolff JO, Cicirello DM. 1990. Mobility versus territoriality—alternative
reproductive strategies in white-footed mice. Anim Behav. 39:
1222–1224.
Wolff JO, Freeberg MH, Dueser RD. 1983. Interspecific territoriality in
two sympatric species of Peromyscus (Rodentia, Cricetidae). Behav
Ecol Sociobiol. 12:237–242.
Xia X, Millar JS. 1991. Genetic evidence of promiscuity in Peromyscus
leucopus. Behav Ecol Sociobiol. 28:171–178.
Zar JH. 1999. Biostatistical analysis. Upper Saddle River (NJ): Prentice
Hall.
Fuxjager and Marler
•
Downloaded from http://beheco.oxfordjournals.org/ at Wake Forest University on August 7, 2014