Heightened aggression and winning contests increase corticosterone but decrease testosterone in male Australian water dragons

    Heightened Aggression and Winning Contests Increase Corticosterone but Decrease Testosterone in Male Australian Water Dragons Troy A. Baird, Matthew B. Lovern, Richard Shine PII: DOI: Reference:

S0018-506X(14)00115-9 doi: 10.1016/j.yhbeh.2014.05.008 YHBEH 3725

To appear in:

Hormones and Behavior

Received date: Revised date: Accepted date:

15 December 2013 25 May 2014 27 May 2014

Please cite this article as: Baird, Troy A., Lovern, Matthew B., Shine, Richard, Heightened Aggression and Winning Contests Increase Corticosterone but Decrease Testosterone in Male Australian Water Dragons, Hormones and Behavior (2014), doi: 10.1016/j.yhbeh.2014.05.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Heightened Aggression and Winning Contests Increase Corticosterone but Decrease Testosterone in Male Australian Water Dragons

(Corresponding author), Permanent address: Department of Biology, University of

SC

1

RI

PT

Troy A. Baird 1,3, Matthew B. Lovern 2, Richard Shine3

NU

Central Oklahoma, 100 North University Drive, Edmond, OK, U.S.A. 73034. (405) 9745776, [email protected]

Affiliation: School of Biological Sciences, University of Sydney, N.S.W., Australia.

2

Department of Zoology, Oklahoma State University, Stillwater, OK, U.S.A. 74078.

TE

D

MA

3

3

AC CE P

(405) 744-5551, [email protected]

School of Biological Sciences, University of Sydney, N.S.W., Australia. (61 2) 9351-

3772. [email protected]

1

ACCEPTED MANUSCRIPT Abstract Water dragons (Intellegama [Physignathus] lesueurii) are large (to > 1 m) agamid lizards

PT

from eastern Australia. Males are fiercely combative; holding a territory requires

RI

incessant displays and aggression against other males. If a dominant male is absent, injured or fatigued, another male soon takes over his territory. Our sampling of blood

SC

from free-ranging adult males showed that baseline levels of both testosterone and

NU

corticosterone were not related to a male’s social tactic (territorial versus non-territorial), or his frequency of advertisement display, aggression, or courtship behavior. Even when

MA

we elicited intense aggression by non-territorial males (by temporarily removing territory owners), testosterone did not increase with the higher levels of aggression that ensued.

TE

D

Indeed, testosterone levels decreased in males that won contests. In contrast, male corticosterone levels increased with the heightened aggression during unsettled

AC CE P

conditions, and were higher in males that won contests. High chronic male-male competition in this dense population may favor high testosterone levels in all adult males to facilitate advertisement and patrol activities required for territory maintenance (by dominant animals), and to maintain readiness for territory take-overs (in non-territorial animals). Corticosterone levels increased in response to intense aggression during socially unstable conditions, and were higher in contest winners than losers. A positive correlation between the two hormones during socially unstable conditions, suggests that the high stress of contests decreased androgen production. The persistent intense competition in this population appears to exact a high physiological cost, which together with our observation that males sometimes lose their territories to challengers, may indicate cycling between these two tactics to manage long-term energetic costs.

2

ACCEPTED MANUSCRIPT Keywords: Aggression, corticosterone, hormones, lizard, steroids, testosterone, territory defense

PT

Introduction

RI

Vertebrates in which males display two or more alternative reproductive tactics provide robust model systems for the study of interactions between hormones and aggressive

SC

behavior (Baird and Hews, 2007). Steroid hormones activate male aggression associated

NU

with reproduction in many species (fish: Liley and Stacey, 1983; Ros et al., 2004; birds: Rohwer et al., 1978; Rohwer and Wingfield 1981; mammals: Bouisou, 1983; Monaghan

MA

et al., 1992), including some lizards (Crews, 1975; Moore and Marler, 1987; Moore et al., 1998; Wikelski et al., 2005; Mills et al., 2008). However, causal interactions between

TE

D

hormones and male aggression can operate in the opposite direction also. Aggression with same-sex rivals appears to elicit secretion of androgens (Arlet et al., 2011; Oliveira

AC CE P

et al., 2009; Smith and John-Alder, 1999), as well as corticosteroids that may reflect the social stress induced by aggression (Earley and Hsu, 2008; Gilmour et al., 2005; Knapp et al., 1995). Secretion of androgens and corticosteroids might also depend on the outcome of contests with rivals, with both positive and negative associations among secretion of each hormone, social status, and winning or losing being reported (reviewed by Hsu et al., 2006). Moreover, the possibility that testosterone and corticosterone act antagonistically to one another (Moore and Jessop 2003) may complicate matters further. If aggression prompts secretion of either hormone, levels of the other may be suppressed.

3

ACCEPTED MANUSCRIPT The challenge hypothesis is the leading theoretical framework proposed to explain secretion of androgens in response to aggression (Wingfield et al., 1990). This hypothesis

PT

predicts that patterns of androgen secretion will be related to variation in demands on

RI

males linked to social structure and their involvement in parental activities. The challenge hypothesis was first developed for and tested in endotherms in which males are generally

SC

more aggressive during settlement of reproductive territories, but then aggression

NU

diminishes as parental activities commence (Wingfield et al., 1990). Secretion of additional androgens is expected in response to challenges from rivals when hormone

MA

levels are below physiological maximum (Wingfield et al., 1990). Because most lizards do not exhibit parental care yet it is common for males to defend territories throughout

TE

D

prolonged reproductive seasons during which females produce multiple clutches of eggs (Baird et al., 2001; Fox et al., 2003), they offer an interesting opportunities to test the

AC CE P

applicability of the challenge hypothesis as a general model. If territory defense requires lizard males to maintain androgen secretion at physiologically maximal levels throughout the season, pulses of androgen secretion are not expected in response to individual challenges by rivals.

Meta-analysis of male vertebrate androgen secretion in response to territorial challenges revealed only moderately strong effects (Hirschenhauser and Oliveira, 2006). Variability in the influence of aggression on androgen levels was likely influenced by differences in the context during which aggression occurred across the diversity of species sampled (Hirschenhauser and Oliveira, 2006), and a number of other ecological and behavioral variables (Goyman, et al., 2007). Tests of the influence of aggression on

4

ACCEPTED MANUSCRIPT secretion of adrenocorticoids have also generated mixed results, with both positive and negative associations being reported (reviewed by Hsu et al., 2006). Hormone responses

PT

by male lizards to contest outcome and heightened aggression are particularly variable.

RI

Male green anoles that won contests staged in the laboratory had higher androgen levels than contest-losers or control males (Greenberg and Crews, 1990). However, this

SC

response may have been influenced by the presence of females in enclosures, or the low

NU

pre-trial androgen levels of male winners. Aggression increased testosterone levels in male marine iguanas (Wikelski et al., 2005), but not in either territorial male tree lizards

MA

(Knapp and Moore, 1995), or male mountain spiny lizards (Moore, 1987). The influence of winning on both hormones depended on social status in male tree lizards, with winners

TE

D

having decreased testosterone, but elevated corticosterone (Knapp and Moore, 1995). These mixed results clearly indicate that data from lizard studies are insufficient to

AC CE P

evaluate the extent to which patterns of hormone secretion in response to heightened aggression and contest outcome support theoretical predictions, and additional such studies are necessary, especially those that involve manipulations of social conditions in the field.

We studied relationships among aggression, androgens, and corticosterone in the eastern water dragon, Intellagama [Physignathus] lesueurii. In our study population, males and females exhibit markedly different levels of aggression, and mature males utilize one of two alternative social tactics (territorial versus non-territorial) that are characterized by territorial males displaying at rates that were over six times that of nonterritorial males, which usually fled or were repelled by territory owners (Baird et al., 2012; 2013a). On three occasions, however, non-territorial males challenged and defeated

5

ACCEPTED MANUSCRIPT a territory owner and usurped their territories (Baird et al., 2012; 2013a). In the present study, we used this fact to induce social changes in levels of aggression and male social

PT

tactics by temporarily removing and later reinstating individual territorial males (Baird et

RI

al., 2012). We collected blood samples from free-ranging lizards to test whether testosterone and corticosterone differ with sex, maturity, and male social status. We also

SC

examined the extent to which hormone levels were correlated with the frequency of male

NU

behavior patterns involved in the advertisement and control of territories, and used temporary removal/reinstatement field experiments to test whether hormone levels

MA

changed in response to heightened social stress and winning or losing aggressive

Methods

TE

D

encounters.

AC CE P

Study population and general methods Intellagama lesueurii is a large (up to 1 kg) semiaquatic, diurnal agamid that occurs in riparian areas throughout eastern New South Wales and Queensland (Cogger, 1986). They are strong swimmers that enter the water routinely to escape predators, and at night (Courtice, 1981). Water dragons feed primarily on terrestrial insects (Harlow, 2001). We conducted this study on the grounds of the Flynn’s Beach Resort in Port Macquarie, NSW (31o 26´ S latitude, 152o 55´ E longitude) from 12 Sept–30 Nov, 2009, during the Austral spring when eastern water dragons are reproductively active (Cuervo and Shine, 2007; Harlow, 2001; Thompson,1993). Water dragons occur throughout the resort grounds (8980 m2) but are most concentrated along 200 m of Wright’s Creek which bisects this site (Baird et al., 2012). Microhabitat used by water dragons consisted

6

ACCEPTED MANUSCRIPT of naturally vegetated riparian areas as well as lawns, plant beds, sidewalks, and even swimming pool decks. Lizards were habituated enough to humans that they tolerated

PT

approach to within 2 m, but retreated when approached more closely. We captured 111

RI

lizards by noose from 15–30 Sept 2009. Lizards were marked by painting numbers on each side of the dorsal torso using white water proof nail polish. Worn numbers were

NU

SC

retouched during recaptures, but none of the lizards lost their numbers through molting.

Collection of blood samples

MA

To determine baseline hormone levels, we collected blood samples from 1000−1400 h within 5 min of beginning to approach, and within 1 min of capturing lizards (mean

TE

D

pursuit + handling time = 1.9 ± 0.18 min) by inserting a heparinized micro-capillary tube (50 ul) beneath the lower eyelid into the orbital sinus (see Ethics statement). We collected

AC CE P

100−200 ul of whole blood, and then staunched the blood flow by applying gentle pressure to the closed eye using a clean cloth. Blood samples were kept on ice in the field for a maximum of 3 h, centrifuged, and the plasma was frozen. Frozen plasma samples were transported to the University of Sydney where they were freeze-dried on Dec 10 2009. Freeze-dried plasma was later transported to Oklahoma State University where all samples were radioimmunoassayed for testosterone and corticosterone (see below). Freeze drying human blood plasma samples did not degrade steroids measured using radioimmunoassay (Das et al., 1983), and freeze dried lizard yolk samples have also been successfully used in quantitative measurements in lizard studies (Warner et al., 2007; 2008).

7

ACCEPTED MANUSCRIPT Baseline samples included 36 females (SVL = 146–223) that we determined were reproductively active by palpation of enlarging ovarian follicles. Fifty-one lizards were

PT

males (162–271 mm SVL) as evidenced by their eversible hemipenes, and 38 of these

RI

(SVL = 200–271 mm) were secreting seminal fluid. Eight of the reproductively active males moved off or our study site after initial marking, whereas the remaining 30 males

SC

(SVL = 238–271) were present throughout our behavioral studies (Baird et al., 2012), and

NU

most of these were involved in removal experiments (described below). We also collected blood samples from 21 lizards (SVL = 192–214 mm) that we classified as

MA

immature because they lacked enlarging follicles or eversible hemipenes.

TE

D

Social and spatial behavior of male dragons We mapped the study site to scale by recording distance and compass measurements

AC CE P

among prominent, permanent landmarks (sidewalks, fence and light posts, trees, creek shore-line). The location of each mapping marker was determined using measurements among a minimum of five adjacent markers to yield a composite map accurate to the nearest 2 m.

One of us (TAB) recorded the social behavior of subject lizards on scale-drawn maps during focal observations (sensu Altmann, 1974). Focal observations involved recording all of the displays and aggressive encounters with conspecifics of both sexes initiated by subject lizards on scale-drawn maps (see similarly Baird et al., 2007; 2012; 2013a). Although lizards at this site were not obviously affected by human presence, we recorded focal observations when human disturbance was minimal. Each observation session lasted 20 min, except for a few (< 5 %) that were terminated earlier because

8

ACCEPTED MANUSCRIPT subjects were lost from view in thick vegetation. To quantify behavior during baseline social conditions on 30 mature males, we recorded an average of 9.2 observation

PT

sessions/male ( ± 0.4 SE) on separate days, which yielded an average of 204 min of

RI

observations/male (± 17.7 SE) (Baird et al., 2012; 2013a). From focal observations, we determined the per min frequency of all social displays pooled. Most of these behavior

SC

patterns involved movements of the head and dewlap, occasionally movements of the tail,

NU

and lateral compression and elevation of the torso. Males also initiated aggressive contests with same-sex rivals, including bipedal chases, head and torso displays at close

MA

range (1 body length), and shoving, wrestling, and biting (Baird et al., 2012). Contests (especially those with escalated physical activity) were rare under baseline conditions,

TE

D

but increased in frequency and intensity during unstable social conditions that we prompted by temporary removal and reinstatement of males that controlled territories

AC CE P

(see below).

We used focal observations to calculate the frequency (per min) that subject males initiated contests, and to determine the percentage of contests that males won or fought to a tie. Winners of contests were readily determined because opponents withdrew and fled (often running bipedally at high speed) to refuge (usually the water). A few contests between owners of adjacent territories with shared boundary ended in ties characterized by both contestants returning to their respective territories, and no change in subsequent use of space by these contesting males (Baird et al., 2012).

9

ACCEPTED MANUSCRIPT Removal experimental protocol and male behavioral responses We conducted 11 trials where we removed a territorial male (a different individual in

PT

each trial) at 1400 –1600 h and kept him off site for 2 d. Removed males were maintained

RI

in damp, low light conditions to mimic cool, cloudy/rainy field conditions when water dragons are inactive. From 0730 to 1500 h during the 2 d removal, we recorded at least

SC

three, 20-min focal observations on all of the mature males adjacent to the territory of the

NU

removed male. Removals prompted intense contests from 800 to 1500 h on the following day among non-territorial males on the removal site (Fig. 1; Baird et al., 2012). In ten of

MA

these trials, by 1500 h of the first removal day one of these formerly non-territorial males had established dominance over the territory (winners), whereas the other contenders

TE

D

were defeated and resumed subordinate behavior (losers) (Baird et al., 2012; 2013a). In the remaining trial, two non-territorial males established dominance over a portion of the

AC CE P

removed male’s territory (two winners). We reinstated the original territory owner at the exact location where we captured him from 2000−2100 h (after dark) of the second removal day. From 800 to 1300 h on the following day (day 3) we recorded at least three, 20-min focal observations on the males that had been present during the removal, plus on the reinstated territorial male, until either the former owner re-established occupancy, or one of his rivals defeated him. The twelve males that we included as winners were previously non-territorial males that fought successfully to establish control of a territory when the original owner was being held off-site. These males were blood sampled from 1300−1500 on their second day of successful territory acquisition/control. We sampled a total of 17 males that we classified as losers because they were defeated during intense contests (one non-

10

ACCEPTED MANUSCRIPT territorial male on the site was involved in two removal experiments that were conducted a month apart). Loser males included nine non-territorial males that fought during the two

PT

days when territorial males were removed, but were defeated by another non-territorial male that acquired the territory temporarily. These males were blood-sampled from

RI

1300−1500 on the second day of removal experiments. We also included as loser males,

SC

eight of the formerly non-territorial males that were successful in acquiring a territory for

NU

two days when the original owner was removed, but then were defeated and displaced by the reinstated owner on the third day. Blood samples were collected from these

MA

individuals from 1100 to 1400 h on the day that they were defeated. To examine the influence of high stress from being held in captivity plus the

TE

D

heightened aggression that ensued upon reinstatement on testosterone and corticosterone, we compared levels in nine of 11 territorial males during baseline social conditions with

AC CE P

those taken from 1100 to 1400 h on the day that we reinstated them to their territories. We were not able to capture two of these territory owners for blood sampling on the day that they were reinstated.

Radioimmunoassay

Plasma concentrations of testosterone and corticosterone were measured using radioimmunoassay after extraction and isolation by column chromatography (Wingfield and Farner, 1975). For each sample, 3–10 mg of freeze-dried plasma (recorded to the nearest mg for each sample) was mixed in 0.5 ml of ddH2O. We equilibrated samples overnight at 4 °C with ~ 1000 cpm of each 3H-testosterone (NET-370, 70 Ci/mmol) and 3

H-corticosterone (NET-399, 71 Ci/mmol) from PerkinElmer Life Sciences, Inc. (Boston,

11

ACCEPTED MANUSCRIPT MA) for individual recovery determinations. Each sample was extracted twice with 2 ml diethyl ether, dried under nitrogen gas in a 37°C water bath, and then reconstituted in 0.5

PT

ml of 10% ethyl acetate in isooctane for chromatography. Columns consisted of a diatomaceous earth:ethylene glycol:propylene glycol upper phase (4:1:1 m:v:v) and a

RI

diatomaceous earth:ddH2O (3:1 m:v) lower phase. Diatomaceous earth (“Celpure P300”)

SC

was purchased from Sigma-Aldrich Catalog #525243 (St. Louis, MO). Neutral lipids and

NU

dihydrotestosterone were removed from the columns with 1.5 ml isooctane and 2.0 ml 10% ethyl acetate in isooctane, respectively, and discarded. Testosterone and

MA

corticosterone were eluted with 2.0 ml 20% and 2.5 ml 52% ethyl acetate in isooctane, respectively, then dried under nitrogen gas in a 37°C water bath, resuspended in

TE

D

phosphate buffer, and refrigerated at 4 °C overnight. Competitive binding RIAs were performed using tritiated steroid tracer (see

AC CE P

above), antibodies from Research Diagnostics, Inc., for testosterone (T-3003; Flanders, NJ) and Sigma-Aldrich for corticosterone (C8784) and testosterone and corticosterone standards from Sigma-Aldrich (T1500 and C2505, respectively). We ran samples across two assays, one for the baseline samples of lizards not involved in experiments (see below), and a second assay for all males sampled during different phases of the removal experiment. Standard curves ranged from 1.95 to 500 pg for testosterone and 3.91 to 1000 pg for corticosterone and were run in duplicate. Samples were run singly and adjusted for recovery and sample mass. Average recovery for testosterone in the two assays was 50 and 62%, respectively, and for corticosterone was 38 and 58%, respectively. Intra-assay coefficients of variation, based on four aliquots from a standard pool for each steroid, were 7 and 12% for testosterone and 8 and 8% for corticosterone.

12

ACCEPTED MANUSCRIPT The inter-assay coefficient of variation was 3% for testosterone and 18% for

PT

corticosterone.

RI

Statistical analyses

We used simple regression to examine potential relationships between aggressive

SC

behavior during baseline social conditions and hormone levels in both territorial and non-

NU

territorial males pooled (N = 30), and in non-territorial males only (N = 29) during unstable social conditions; either the second day when territory owners were being held

MA

off site, the day when original owners had been reinstated, or both days. Baseline hormone levels among mature males, mature females, and immature lizards were

TE

D

compared using Kruskal-Wallis tests because these data were heteroscedastic even after log -transformation. Two-group comparisons were made using t-tests. We used the

AC CE P

program R, version 3.01 (R Development Core Team, 2013) to calculate effect sizes (Cohen’s d) for both Kruskal-Wallis and t-tests. We used two-factor ANOVA to examine the influence of social condition (baseline versus unsettled) and male social status (winner versus loser) at the end of removal/reinstatement phases on testosterone and corticosterone, and calculated effect sizes (eta-squared) using the program R. Data sets that were heteroscedastic as determined by F-tests were log-base 10 transformed for analyses.

Ethics statement This research was conducted with permission of the New South Wales National Parks and Wildlife Service (permit # S12905), and the Animal Ethics Committee, University of

13

ACCEPTED MANUSCRIPT Sydney (L04/9-2009/1/5063). All study subjects were in good health at the end of our study. Capturing lizards by noosing does not prevent their ventilation because the hyoid

PT

bone prevents any pressure on the trachea. Blood samples were collected by inserting a heparinized micro-capillary tube (50 ul) beneath the lower eyelid into the orbital sinus.

RI

We collected 100−200 ul of whole blood within 30 secs, and then quickly staunched the

SC

blood flow by applying gentle pressure to the closed eye using a clean cloth. Because

NU

water dragons are large (0.3−1.0 kg), the volume of blood sampled for hormone assays was very small relative to their total volume, especially for large males. Lizards resumed

MA

their normal behavior as soon as we released them following blood sampling, including the males involved in experiments from which we collected up to three samples only few

lizards.

Results

AC CE P

TE

D

days apart. There was no indication of infection from blood sampling in any of the

Hormones and male behavior during baseline social conditions Testosterone levels varied by sex and social class (H 2, 108 = 58.33, P = 0.0001, Cohen’s d = 0.54), with the mean for mature males being 12.5 times that of females and 11.6 times that of lizards that we could not sex because they were immature (Fig. 2). By contrast, baseline levels of corticosterone did not differ in these three groups of lizards (H 2, 108 = 2.26, P = 0.323, Cohen’s d = 0.02). Neither testosterone (t 1, 29 = 0.87, P = 0.39, Cohen’s d = 0.32), nor corticosterone (t 1, 29 = 0.20, P = 0.83, Cohen’s d = 0.22) levels were statistically different in territorial and non-territorial males during baseline conditions (Fig. 3). There were also no statistically significant relationships between the three behavioral variables that we 14

ACCEPTED MANUSCRIPT measured (displays/min, aggressive contests initiated/min, and % contests won or tied) with either hormone for all males pooled (F’s = 0.002–2.33, P’s = 0.139 – 0.996), or for

PT

territorial males considered separately (F’s =0.007–2.31, P’s = 0.36–0.93). Although not

RI

statistically significant, there were positive relationships between testosterone and corticosterone when all males were pooled (F 1, 29 = 2.40, P = 0.13, r = 0.28), and in

NU

SC

territorial males considered separately (F 1, 29 = 3.39, P = 0.095, r = 0.46, Fig. 4).

Hormones and behavior in non-territorial males during unsettled social conditions

MA

There were no significant regression relationships between the three male behavioral variables (same as above) and testosterone levels in non-territorial males during unsettled

TE

D

social conditions (Table 1). By contrast, corticosterone increased with all three behavioral variables (Table 1). Levels of corticosterone and testosterone in non-

(Table 1).

AC CE P

territorial males were positively correlated with one another during unsettled conditions

The interaction between social condition (baseline versus unsettled) and male contest status (winner versus loser) at the end of the experimental phase was marginally significant (F 1, 57 = 4.07, P = 0.049, Ƞ 2 = 0.070). Testosterone tended to decrease during unsettled conditions, more so in losers than winners (Fig. 5A), but this interaction explained only 7% of the variance in testosterone. There was a strong interaction between social condition (baseline versus unsettled) and male contest status (F1, 57 = 5.76, P = 0.019, Ƞ 2 = 0.852) and elevated corticosterone in male winners (Fig. 5B). Nine of the 11 removed territorial males regained their original territories by 1500 h of the day that we reinstated them, whereas two males did not regain their territories

15

ACCEPTED MANUSCRIPT (see below). Mean testosterone level after the heightened aggression that ensued upon reinstatement (394.3 pg/mg ± 58.9) was not different (paired t-test; t 1, 8 = 0.47, P = 0.65,

PT

Cohen’s d = 0.22) from that before these males were removed (570.7 pg/mg ± 141.3). By contrast, the average concentration of corticosterone following reinstatement and

RI

aggression (51.05 pg/mg ± 8.87) was 3.4 times higher (t 1, 9 = 2.34, P = 0.042, Cohen’s d

SC

= 1.10) the level during baseline conditions (174.9 pg/mg ± 43.7).

NU

One of the two removed males that did not regain his territory did not fight upon reinstatement. When we removed this male, we discovered that he had a fractured lower

MA

jaw that was infected. Instead of fighting, this male remained on his original area by adopting subordinate social tactics. His testosterone level before removal and after

TE

D

reinstatement decreased over 4-fold (1216 to 320 pg/mg), whereas his corticosterone levels remained nearly the same (67 versus 82 pg/mg). The other instance when the

AC CE P

reinstated male lost his former territory was drastically different. This male was nearly identical in size to the male that took over his territory during the removal, and these two males fought intensely from 0730−1130 on the day that the original owner was reinstated (Baird et al., 2012). At the end of this long fight, the original territory owner withdrew to the periphery of his former territory. Until the end of our study, this male remained there using non-territorial tactics, whereas the male that won ownership actively controlled this territory until the end of our study (Baird et al., 2012). The new owner male’s testosterone level increased 2.3 times that of his baseline level (182 versus 423 pg/mg), whereas his level of corticosterone increased 7.8 times (41.8 versus 326.3 pg/mg). For the male that lost his territory, levels of testosterone and corticosterone after removal were

16

ACCEPTED MANUSCRIPT 1.8, and 1.7 times higher, respectively, than those before removal (testosterone, 282 vs.

PT

498 pg/mg, corticosterone = 69.8 vs. 119.3 pg/mg).

RI

Discussion

As expected, sexually mature male water dragons had much higher testosterone levels

SC

than did mature females and immature lizards. However, although frequencies of

NU

aggressive behavior were much higher in males that defended territories than those relegated to non-territorial social status (Baird et al., 2012; 2013a), neither baseline

MA

testosterone nor corticosterone differed depending on male social tactics. Even during unsettled social conditions when male aggression was heightened, testosterone was not

TE

D

correlated with the frequency of aggressive acts. Moreover, males prevailing in the intense contests that we prompted experimentally had lower (not higher) testosterone

AC CE P

levels than males that were defeated. In marked contrast, corticosterone was positively correlated with male aggression when social conditions were unsettled, and was higher in males that won contests. Corticosterone was also higher in territorial males after they had experienced the combined stressors of being held off the study site followed by the combat that ensued upon their reinstatement. During unsettled conditions, corticosterone and testosterone were positively correlated. There is abundant theory to address bi-directional causal relationships between aggression and hormone levels, particularly for androgens. Male lizards that display plastic alternative reproductive tactics are important study models for tests of these theoretical predictions (reviewed by Baird, 2013b). For species that display alternative male tactics that are plastic, organization-activation theory predicts that the transition

17

ACCEPTED MANUSCRIPT from one behavioral tactic to another is activated by secretion of androgens (Moore, 1991). Coupled with literature showing that androgens increase with heightened

PT

aggression (e.g., Bouissou, 1983; Rohwer and Wingfield, 1981; Wikelski et al., 2005),

RI

males utilizing low aggression tactics might be expected to secrete less androgen. However, androgen levels are expected to increase when males shift to more aggressive

SC

tactics, such as defense of territories (relative plasticity hypothesis, Baird and Hews,

NU

2007; Moore, 1991).

We found no evidence to support predictions of the relative plasticity hypothesis

MA

in I. lesueurii. Baseline levels of testosterone did not differ in territorial and nonterritorial water dragon males, and testosterone was not correlated with the frequency of

TE

D

any of the behavior patterns that territorial males used to advertise and defend territories (Baird et al., 2012). Maintenance of high testosterone by all male water dragons may be

AC CE P

related to the high levels of aggression in this dense population, and especially the chronically high potential for subordinate males to challenge territory owners and usurp territories by decisively winning a single prolonged contest (Baird et al., 2012; 2013a). We documented two such instances of spontaneous territory take-over (and anecdotally witnessed a third instance) among 30 males. Baseline androgen levels also did not differ in territorial and non-territorial male collared lizards, another iguanid in which alternative social tactics are highly plastic (Baird and Hews, 2007). Hormone profiles of both water dragons and collared lizards differ markedly from green anoles and Galapagos marine iguanas. Average testosterone levels were lower in non-territorial male anoles than in older males that controlled territories (Husak et al., 2007). In Galapagos marine iguanas, another long-lived species, males adopt one of three plastic social tactics conditional on

18

ACCEPTED MANUSCRIPT the local densities of same-sex competitors. Consistent with predictions of the relative plasticity hypothesis, testosterone levels in males that defended territories were higher

PT

than those of males utilizing subordinate social tactics (Wikelski et al., 2005).

RI

Whether or not heightened aggression (such as that during challenges to territory ownership), prompts testosterone secretion may also depend upon the temporal pattern of

SC

baseline androgen secretion throughout the reproductive season (challenge hypothesis;

NU

Wingfield et al., 1990). In territorial endotherms with male parental care, high baseline levels of androgens are expected when males settle territories early in the breeding

MA

season, but then testosterone secretion subsides once mutual boundaries have been established, eggs/offspring are produced, and males become involved in parental

TE

D

activities (Wingfield et al., 1990). Hence, intrusions are not expected to prompt androgen secretion when production is already near maximum levels, but later challenges should

AC CE P

cause a pulse of androgen production to stimulate the aggressive responses necessary to repel intruders (Wingfield et al., 1990). Territorial lizards provide an interesting comparative test of the general explanatory power of the challenge hypothesis because few species have parental care, and advertisement/defense of territories continues throughout the breeding season, especially if females produce multiple clutches of eggs (Baird et al., 2001; Husak et al., 2009; Baird, 2013c). Male lizards (including water dragons), exhibit prolonged defense of territories are expected to maintain testosterone near maximal levels throughout the breeding season. Social challenges, therefore, are not expected to prompt secretion of additional testosterone (Baird and Hews, 2007; reviewed by Baird, et al., 2013b).

19

ACCEPTED MANUSCRIPT Consistent with this expectation for territorial lizards, testosterone levels in male water dragons tended to decrease rather than increase following two days of intense

PT

contests. A similar dissociation between androgen secretion and heightened aggression

RI

has been reported in several other iguanian lizards (e.g., Curtis, 2010; Klukowski and Nelson, 2001; Moore, 1986; Moore et al., 1998), including another agamid (Watt et al.,

SC

2003), as well as one snake (Schuett et al., 1996). Other studies, however, are less

NU

consistent with predictions of the challenge hypothesis. Results of field studies include significant within-season fluctuations in androgen levels of male tree- (Moore, 1986) and

MA

fence-lizard (Klukowski and Nelson, 2001), and increased testosterone secretion by male fence lizards in response to successive intrusion over four days (Smith and John-Alder,

TE

D

1999). When inhibition of testosterone was discontinued in male marine iguanas, males exhibited very high levels of aggression, which prompted secretion of testosterone at

AC CE P

even higher levels than before the inhibitory treatment (Wikelski et al., 2005). Together, these mixed findings suggest that the influence of aggression on androgen secretion in lizards is more complex and varied than the explanation posed by the challenge hypothesis (reviewed by Baird, 2013b; Hirschenhauser and Oliviera, 2006; Goyman, 2007).

In contrast with testosterone, corticosterone levels increased in male water dragons following heightened contests, and levels were higher in males that won contests. Our observed positive relationship between corticosterone and heightened aggression, which usually characterizes males that win contests, is consistent with results of some other studies (Hannes et al., 1984; Knapp and Moore, 1995; Overli et al., 1999a, Ramos-

20

ACCEPTED MANUSCRIPT Fernandez et al., 2000). However, it is more common for corticosterone to increase following defeat (reviewed by Hsu et al., 2006).

PT

The observed pattern of hormone secretion in mature male water dragons appears

RI

to reflect the intense social dynamics in this population described by Baird et al (2012). A high density of mature males necessitated high levels of patrol and display to maintain

SC

territory ownership in the face of a chronically high risk of challenge by numerous large

NU

male rivals. Because over one-half of these males had been prevented from acquiring territories, they were well rested and poised to fight. All mature males, regardless of their

MA

social status, maintained high levels of testosterone. We suggest that the high testosterone levels of territorial males support the minimum display and patrol activity necessary to

TE

D

maintain territory ownership under such high chronic pressure from intruders. Equally high testosterone levels may be advantageous in non-territorial males also, allowing them

AC CE P

to initiate intense fighting as soon as they detect an opportunity to challenge a territory owner that is injured or fatigued (see similarly, Husak et al., 2007). We observed three cases when non-territorial males took over territories by defeating territory owners during a single prolonged bout of intense fighting, and all of our temporary removals of territory owners prompted intense aggression by nearby non-territorial males. Because both categories of males maintained elevated levels of testosterone, albeit for different functions, it is not surprising that levels did not increase in response to heightened aggression. Indeed, testosterone decreased in male water dragons that eventually won long, intense contests. Decreased testosterone in response to winning has been observed in some earlier studies, including those on other iguanian lizards (Hannes et al., 1984; Knapp and Moore, 1995). Non-territorial collared lizard males maintained testosterone

21

ACCEPTED MANUSCRIPT levels as high as those of territory owners as in water dragons. Because non-territorial male collared lizard males were younger and smaller, high testosterone production may

PT

have supported their high rates of growth as well as readiness for aggression (Baird and

RI

Hews, 2007). Increased growth was unlikely to be a factor in male water dragons, because most of the non-territorial males were as large or even larger than territory

SC

owners.

NU

The positive effect of heightened aggression and winning contests on corticosterone levels in water dragons supports the hypothesis that mature males in this

MA

dense population incur very high chronic competition costs to acquire and maintain territories over the long term (Baird et al. 2012; 2013a). It was relatively common (three

TE

D

instances in a two month study) for non-territorial males to challenge and defeat territorial males (Baird et al., 2012). Defeated territory owners abruptly shifted to non-

AC CE P

territorial tactics characterized by low rates of display and patrol and remained on an area adjacent to the territory that they once controlled. Corticosterone increased in response to aggressive behavior, in formerly non-territorial males that temporarily won contests so essential for territory defense, and both a male that fought unsuccessfully to regain his territory, and the male usurped this territory. Maintenance of territories in this dense population requires frequent advertisement using patrol and display, and aggression directed toward non-territorial males which are sometimes able to defeat and displace territory owners (Baird et al., 2012; 2013a). Elevated corticosterone levels in response to such heightened aggression suggests that the chronic cost of territory maintenance in this population may eventually wear down territory owners physiologically to a point that they are not able to fend off

22

ACCEPTED MANUSCRIPT challenges by better-rested rivals. Because water dragons are long-lived (Baird et al., 2012), males in dense populations such as this one probably must cycle between

PT

territorial and non-territorial tactics in response to the dynamics of their energetic

RI

budgets. The markedly different energy expenditures required by territorial versus nonterritorial social tactics thus seem to drive a cycle whereby individual males oscillate

SC

between periods of territorial and non-territorial status. This scenario may also explain

NU

the decrease in male water dragon testosterone in response to winning contests. If high levels of corticosterone inhibit production of testosterone by the testes (Pankhurst and

MA

Van Der Kraak, 2000; Consen et al., 2001), then the stress of fighting intensely enough to

Acknowledgements

TE

D

win contests may depress the circulating levels of testosterone.

AC CE P

This research was supported by the Office of Research and Grants, University of Central Oklahoma (T.A.B), and the Australian Research Council Grant FF056365 (R.S.). We thank M. Elphick, M. Greenles, A. Haythornwaite, S. LaFave, D. Pike, T. Shine, W. Unsell, and J.R. York for their assistance on logistics, N. Soran and S. Simpson for freeze-drying our water dragon plasma samples, and T.D. Baird for help in the field, and her photography. We are very grateful to the entire staff of the Flynn’s Beach Resort and Blue Water Bar and Restaurant for access to the study site. Special thanks to A. Greenway and P. Kemsley whose enthusiasm and keen appreciation of water dragons contributed greatly to this project.

23

ACCEPTED MANUSCRIPT Table 1. Summary of regression statistics for the relationships between the behavior of non-territorial male water dragons (N = 29) during unstable social conditions with

Independent

Dependent

Variable

Variable

Displays/min

T

Contests/min

T

% wins + ties

T

Displays/min

C

Contests/min

C

% wins + ties

C

r

0.252

0.620

0.100

0.437

0.515

0.131

1.810

0.193

0.790

7.969

0.009

0.492

5.064

0.034

0.410

C

5.330

0.031

0.450

T

6.820

0.015

0.456

SC

P

AC CE P

TE

D

MA

NU

F

RI

PT

testosterone (T) and corticosterone (C), and the relationship between the two hormones.

24

ACCEPTED MANUSCRIPT Figure captions Fig. 1. Contesting male water dragon males engaged in side-to-side head pushing. This

PT

behavior was common during intense combat over territory ownership such as occurs

RI

between rival non-territorial males in response to temporary removal of male territory

SC

owners.

NU

Fig. 2. Baseline plasma hormone levels in immature lizards of unknown sex (N = 21), mature female (N = 36), and mature male (N = 51) water dragons. Data are means ± SE.

MA

The asterisk indicates statistically higher (P = 0.0001) testosterone levels in males.

TE

D

Fig. 3. Baseline plasma hormone levels in levels in territorial (N = 14) and non-territorial

AC CE P

(N = 16) water dragon males.

Fig. 4. The relationship between plasma corticosterone and testosterone (pg/mg) for (A) all male water dragons pooled, and (B) the relationship between plasma corticosterone and testosterone (pg/mg) for territorial males alone.

Fig. 5. Influence of social condition and male contest outcome on plasma (A) testosterone and (B) corticosterone. Hatched bars indicate baseline social conditions, solid bars indicate unsettled conditions. Data are mean ± SE.

25

ACCEPTED MANUSCRIPT References Altmann, J., 1974. Observational study of behaviour: sampling methods. Behav. 49,

PT

227–267.

RI

Arlet, M.E., Kaasik, A., Molleman, F., Isbell, L., Carey, J.R., Mänd, R., 2011. Social factors increase fecal testosterone levels in wild male gray-cheeked mangabeys

SC

(Lophocebus albibena). Horm. Behav. 59, 605–611.

NU

Baird, T.A., Hranitz, J.M., Timanus, D.K., Schwartz, A.M., 2007 Behavioral attributes influence annual mating success more than morphological traits in male collared

MA

lizards. Behav. Ecol. 18, 1146–1154.

Baird, T.A., Sloan, C.L., Timanus, D.K., 2001. Intra- and interseasonal variation in the

TE

D

socio-spatial behavior of adult male collared lizards, Crotaphytus collaris (Reptilia, Crotaphytidae). Ethol. 106, 1–19.

AC CE P

Baird, T.A., Hews, D.K., 2007. Hormone levels in territorial and non-territorial male collared lizards. Physiol. Behav. 92, 755–763. Baird, T.A., Baird, T.D., Shine, R., 2012. Aggressive transition between alternative male social tactics in a long-lived Australian dragon (Physignathus lesueurii) living at high density. PLoS ONE, 7 (8), e41819.doi.10.137/journal.pone.0041819. Baird, T.A., Baird, T.D., Shine, R. 2013a. Showing red: male coloration signals same-sex rivals in an Australian water dragon. Herpetol. 69, 436−444. Baird, T.A. 2013b. Lizards and other reptiles as model systems for the study of contest behavior, In: Hardy, I.C.W., Briffa, M. (Eds.), Animal Contests. Cambridge Univ. Press, pp. 258–286.

26

ACCEPTED MANUSCRIPT Baird, T.A., 2013c. Social life on the rocks: Behavioral diversity and sexual selection in collared lizards, In: Lutterschmidt, W.I. (Ed.), Reptiles in Research:

PT

Investigations of Ecology, Physiology, and Behavior from Desert to Sea. Nova

RI

Biomedical Press, pp. 213−245.

Bouissou, M.F. 1983. Androgens, aggressive behavior and social relationships in higher

SC

mammals. Horm. Res.18, 43–61.

NU

Cogger, H.G., 1986. Reptiles and Amphibians of Australia. Reed Books, Australia. Courtice, P.G., 1981. Respiration in the eastern water dragon, Physignathus lesueurii

MA

Agamidae). Comp. Biochem. Physiol. 68A, 429–436. Consten, D., Lambert, J.G.D., Goos, H.J.Th., 2001. Cortisol affects testicular

TE

D

development in male carp, Cyprinus carpio L. but not via an effect on LH secretion. Comp. Biochem. Physiol. B 129, 671–677.

AC CE P

Crews, D., 1975. Psychobiology of reptilian reproduction. Sci. 189, 1059–1065. Cuervo, J.J., Shine, R., 2007. Hues of a dragon’s belly: morphological correlates of ventral colouration in water dragons. J. Zool. 273, 298–304. Curtis, J., 2010. Social modulation of androgens and glucocorticoids in male collared lizards.Unpublished Masters Thesis University of Central Oklahoma, Edmond, OK. Das, R.E.G., Calam, D.H., Mitchell, F.L., Woodford, F.P., Cadwood, M., Gaskell, S.J., Joyce, B., McMillan, M., and Wheeler, M.J., 1983. Stability of four steroids in lyophilized human serum. Ann. Clin. Biochem. 20, 364−368. Earley, R.L., Hsu, Y., 2008. Reciprocity between endocrine state and contest behavior in the killifish, Kryptolebius marmoratus. Horm. Behav. 53, 442–451.

27

ACCEPTED MANUSCRIPT Fox, S.F., McCoy, J.K., Baird, T.A., 2003. Lizard Social Behavior. Johns Hopkins. Gilmour, K.M., DiBattista, J.D., Thomas, J.B., 2005. Physiological causes and

PT

consequences of social status in salmonid fish. Integr. Comp. Biol. 45, 263–273. Goyman, W., Landys, M.M., and Wingfield, J.C. 2007. Distinguishing seasonal

SC

hypothesis. Horm. Behav. 51, 463–476.

RI

responses from male-male androgen responsiveness – revisiting the challenge

NU

Greenberg, N., Crews, D., 1990. Endocrine and behavioral responses to aggression and social dominance in the green anole lizard, Anolis carolinensis. Gen. Comp.

MA

Endocrinol. 77, 1–10.

Harlow, P.S., 2001. The ecology of sex-determining mechanisms in Australia agamid

TE

D

lizards. Unpublished Ph.D Thesis, Macquarie University, Sydney, Australia. Hannes, R.P., Franck, D., Liemann, F., 1984. Effects of rank-order fights on whole body

AC CE P

and blood concentrations of androgens and corticosteroids in the male swordtail (Xiphophorus helleri). Z. für Tierpsychol. 65, 53–65. Hirschenhauser, K., Oliveira, R.F., 2006. Social modulation of androgens in male vertebrates: Meta-analysis of the ‘challenge hypothesis’. Anim. Behav. 71, 265– 277.

Hsu, Y., Earley, R.L., Wolf, L.L., 2006. Modulation of aggressive behavior by fighting experience: mechanisms and contest outcomes. Biol. Rev. 81, 33–74. Husak, J.F., D.J. Irschick, J.J. Meyers, S.P. Lailvaux, and I.T. Moore. 2007. Hormones, sexual signals and performance of green anole lizards (Anolis carolinensis). Horm. Behav. 52, 360−367. Husak, J.F., D.J. Irschick, J.P. Henningsen, K.S., Kirkbride, S.P. Lailvaux, and I.T.

28

ACCEPTED MANUSCRIPT Moore. 2009. Hormonal response of male green anole lizards (Anolis carolinensis) to GnRH challenge. J. Exp. Zool. 311A, 105−114.

PT

Klukowski, M., Nelson, C.E., 2001. The challenge hypothesis and seasonal changes in

RI

aggression and steroids in male northern fence lizards (Sceloporus undulatus hyacinthinus). Horm. Behav. 33, 197–204.

SC

Knapp, R., Moore, M.C., 1995. Hormonal responses to aggression vary in different types

NU

of agonistic encounters in male tree lizards. Horm. Behav. 29: 85–105. Liley, N.R., Stacey, N.E., 1983. Hormones, pheromones, and reproductive behavior in

MA

fish. In: Hoar, W.S., Randall, D.S., Donaldson, E.M. (Eds.) Fish Physiology vol. IXB. Academic Press, New York. Pp 1–63.

TE

D

Monaghan, E.P., Glickman, S.E. 1992. Hormones and aggressive behavior. In: Becker, J.B., Breedlove, S.M., Crews, D (Eds.) Behavioral Endocrinology. MIT Press.

AC CE P

Cambridge, MA. pp. 261–285. Moore, M.C., 1986. Circulating steroid hormones during rapid aggressive responses of territorial male mountain spiny lizards, Sceloporus jarrovi. Horm. Behav. 21, 511–521.

Moore, M.C. 1987. Circulating steroid hormones during rapid aggressive responses of territorial mountain spiny lizards, Sceloporus jarrovi. Horm. Behav. 21, 511−521. Moore, M.C., 1991. Application of the organization-activation theory to alternative male reproductive strategies: a review. Horm. Behav. 25, 154–179. Moore, M.C., Marler, C.A., 1987. Effects of testosterone manipulations on nonbreeding season territorial aggression in free-living male lizards, Sceloporus jarrovi. Gen. Comp. Endocrinol. 65, 225–232.

29

ACCEPTED MANUSCRIPT Moore, M.C, Hews, D.K., Knapp, R., 1998. Hormonal control and evolution of alternative male phenotypes: generalizations of models for sexual

PT

differentiation. Am. Zool. 38, 133–151.

RI

Moore, I.T., Jessop, T.S., 2003. Stress, reproduction, and adrenocortical modulation in amphibians and reptiles. Horm. Behav.43, 39–47.

SC

Overli, O., Harris, C.A., and Winberg, S., 1999. Short-term effects of fights for social

NU

dominance and the establishment of dominant-subordinate relationships on brain monamines and cortisol in rainbow trout. Brain Behav. Evol., 54, 263−275.

MA

Oliveira, R.F., Silva, A., Canário, A.V.M., 2009. Why do winners keep winning?

TE

Soc. B. 276, 2249–2256.

D

Androgen mediation of winner but not loser effects in a cichlid fish. Proc. Roy.

Pankhurst, N.W., Van Der Kraak, G., 2000. Evidence that acute stress inhibits ovarian

AC CE P

steroidogenesis in rainbow trout in vivo through the action of cortisol. Gen. Comp. Endocrinol. 117, 225–237. R Development Core Team, 2013. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing http://www.//R-project.org/. Ramos-Fernandez, G., Nunez-De La Mora, A., Wingfield, J.C., Drummond, H., 2000. Endocrine correlates of dominance in chicks of the blue-footed booby (Sula nebouxii): testing the challenge hypothesis. Ethol. Ecol. Evol. 12, 27−34. Rohwer, S., Rohwer, F.C., 1978. Status signaling in Harris' sparrows: experimental deceptions achieved. Anim Behav. 26, 1012–1022. Rohwer, S., Wingfield, J.C., 1981. A field study of social dominance, plasma levels of

30

ACCEPTED MANUSCRIPT luteinizing hormone and steroid hormones in wintering Harris' sparrows. Z. für Tierpsychol. 57, 173–183.

PT

Ros, A.F.H., Bruintjes, R., Santos, R.S., Canario, A.V.M., Oliveira, R.F., 2004. The role of

RI

androgens in the trade-off between territorial and parental behavior in the Azorean rock pool blenny, Parablennius parvicornis. Horm. Behav. 46, 491–497.

SC

Schuett, G.W., Harlow, H.J., Rose, J.D., Van Kirk, E.A., Murdoch, W.J., 1996. Levels of

NU

plasma testosterone in male copperheads Agkistrodon contortrix following staged encounters. Horm. Behav. 30, 60–68.

MA

Smith, L.C., John-Alder, H.B., 1999. Seasonal specificity of hormonal, behavioral, and coloration responses to within- and between-sex encounters in male lizards

TE

D

Sceloporus undulatus. Horm. Behav. 36, 39–52. Thompson, M.B., 1993. Estimates of population structure of the eastern water dragon

AC CE P

Physignathus lesueurii (Reptilia: Agamidae) along riverside habitat. Wildlife Res. 20, 613–619.

Warner, D.A., Lovern, M.B., and Shine, R. 2007. Maternal nutrition affects reproductive output and sex allocation in a lizard with environmental sex determination. Proc. Royal Soc. B: Biological Sciences. 274, 883−890. Warner, D.A., Lovern, M.B., and Shine, R. 2008. Maternal influences on offpring phenotypes and sex ratios in a multi-clutching lizard with environmental sex determination. Biol. J. Linn. Soc. 95, 256−266. Watt, M.J., Joss, J.M.P. 2003. Structure and function of visual displays produced by male jacky dragons, Amphibolurus muricatus, during social interactions. Brain Behav. Evol. 61: 172–183.

31

ACCEPTED MANUSCRIPT Wikelski, M., Steiger, S.S., Gall, B., Nelson, K.H., 2005 Sex, drugs and mating role: testosterone-induced phenotype-switching in Galapagos marine iguanas. Behav.

PT

Ecol. 16, 260–268.

RI

Wingfield, J.C., Farner, D.S. 1975. The determination of five steroids in avian plasma by radioimmunoassay and competitive protein binding. Steroids, 26, 311–327.

SC

Wingfield, J.C., Hegner, R.E., Ball, G.F., Dufty, A.M. 1990. The ‘challenge hypothesis’:

NU

Theoretical implications for patterns of testosterone secretion, mating systems,

AC CE P

TE

D

MA

and breeding strategies. Am. Nat. 136, 829–846.

32

AC CE P

Figure 1

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

33

Figure 2

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

34

Figure 3

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

35

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 4

36

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 5

37

ACCEPTED MANUSCRIPT Highlights Some water dragon males were aggressively territorial against subordinate males.



Baseline hormone levels were not related to social status or aggressive behaviors.



Intense aggression and winning decreased testosterone but increased

RI

PT





SC

corticosterone.

During aggression, these hormones were positively correlated in non-territorial

Costly competition may cause males to cycle between dominant and subordinate

TE

D

MA

tactics.

AC CE P



NU

males.

38