effects of water restriction on reproductive physiology and affiliative behavior in an...

Hormones and Behavior 63 (2013) 462–474

Contents lists available at SciVerse ScienceDirect

Hormones and Behavior

j ourna l homepage: www.e lsev ie r .com/ locate /yhbeh

Effects of water restriction on reproductive physiology and affiliative behavior in anopportunistically-breeding and monogamous songbird, the zebra finch

Nora H. Prior b,⁎, Sarah A. Heimovics a, Kiran K. Soma a,b,c

a Department of Psychology, University of British Columbia, Vancouver, BC, Canadab Department of Zoology, University of British Columbia, Vancouver, BC, Canadac Graduate Program in Neuroscience, University of British Columbia, Vancouver, BC, Canada

⁎ Corresponding author at: 2136West Mall, UniversityBC, Canada V6T 1Z4.

E-mail address: [email protected] (N.H. Prior)

0018-506X/$ – see front matter © 2013 Elsevier Inc. Alhttp://dx.doi.org/10.1016/j.yhbeh.2012.12.010

a b s t r a c t
a r t i c l e i n f o
Article history:Received 12 June 2012Revised 7 December 2012Accepted 10 December 2012Available online 26 December 2012

Keywords:AffiliationCorticosteroneDHEAEstradiolMonogamyNeurosteroidPair bondSex differenceSongTestosterone

Wild zebra finches form long-term monogamous pair-bonds that are actively maintained year-round, evenwhen not in breeding condition. These desert finches are opportunistic breeders, and breeding is highlyinfluenced by unpredictable rainfall. Their high levels of affiliation and complex breeding patterns makezebra finches an excellent model in which to study the endocrine regulation of affiliation. Here, we comparedzebra finch pairs that were provided with water ad libitum (control) or water restricted. We examined(1) reproductive physiology, (2) pair-maintenance behaviors in several contexts, and (3) circulating andbrain steroid levels. In females, water restriction profoundly reduced largest ovarian follicle size, ovary size,oviduct size, and egg laying. In males, water restriction had no effect on testes size but decreased systemictestosterone levels. However, in the hypothalamus, local testosterone and estradiol levels were unaffectedby water restriction in both sexes. Systemic and local levels of the androgen precursor dehydroepiandroster-one (DHEA) were also unaffected by water restriction. Lastly, in three different behavioral paradigms, we ex-amined a variety of pair-maintenance behaviors, and none were reduced by water restriction. Takentogether, these correlational data are consistent with the hypothesis that local production of sex steroids inthe brain promotes the expression of pair-maintenance behaviors in non-breeding zebra finches.

© 2013 Elsevier Inc. All rights reserved.

Introduction

Affiliative, pro-social behaviors are expressed in a wide range of con-texts, including courtship, pair-bonding, parental behavior, reconcilia-tion, coordination of group movements, and cooperation (Abartz andHollander, 2006; Buck, 2002; Carter, 1998; Pellegrini, 2008; Penner etal., 2005). These behaviors include physical contact, allopreening orgrooming, and vocal communication (Elie et al., 2010, 2011a; Penner etal., 2005; St-Pierre et al., 2009). Themotivation to engage in affiliative be-haviors is physiologically regulated (Buck, 2002; Goodson et al., 2005;Lim and Young, 2006; Robinson et al., 2005), and engagement in thesebehaviors has physiological effects on individuals (Kikusui et al., 2006;Neumann, 2009). While affiliative behaviors are expressed in variouscontexts, research examining the physiological regulation of these be-haviors has been largely restricted to parental behaviors and pair-bondformation in monogamous animals.

Zebra finches are an excellent model in which to study affiliationand its endocrine regulation. They are gregarious and have a large rep-ertoire of affiliative behaviors (Birkhead et al., 1988; Elie et al., 2010;Zann, 1994, 1996). As juveniles, they engage in these behaviors with

of British Columbia, Vancouver,

.

l rights reserved.

their cohorts and parents. After sexual maturity, however, they engagein these behaviors almost exclusively with their pair-bonded partner(Birkhead et al., 1988). These long-term pairs are socially and sexuallymonogamous, and the pair maintains this bond year-round (Birkheadet al., 1988, 1990; Griffith et al., 2010; Zann, 1994).

Zebra finches are opportunistic breeders and integrate many envi-ronmental cues to time breeding in arid habitats, includingwater avail-ability, food availability, presence of green grasses, and photoperiod(Perfito et al., 2007, 2008; Zann, 1994). Water availability is one of themost important cues for wild zebra finches, and water restriction canbring captive male zebra finches out of breeding condition (Morton,2009; Perfito et al., 2008; Vleck and Priedkalns, 1985). The hormonalprofile of non-breeding zebra finches has not been fully characterized,but evidence to date suggests that circulating sex steroids are reducedin non-breeding zebra finches (Perfito et al., 2006, 2007).

While courtship and pair-bond formation in zebrafinches have beenwell studied, the maintenance of pair bonds has not. Courtship behav-iors such as male song and sexual displays are regulated by sex steroids(Arnold, 1975; Harding and Rowe, 2003; Hill et al., 2005). These behav-iors are also important in pair-maintenance. Classically, sex steroidshave been thought to be produced in the gonads and then travel tothe brain to regulate behavior. However, there is now abundant evi-dence that sex steroids can also be produced locally in the brain, eitherde novo from cholesterol or from circulating prohormones such as

http://dx.doi.org/10.1016/j.yhbeh.2012.12.010

mailto:[email protected]

http://dx.doi.org/10.1016/j.yhbeh.2012.12.010

http://www.sciencedirect.com/science/journal/0018506X

463N.H. Prior et al. / Hormones and Behavior 63 (2013) 462–474

dehydroepiandrosterone (DHEA) (Balthazart and Ball, 2006; Forlanoet al., 2006; Schlinger and Remage-Healey, 2012). Behaviors that areregulated by gonadally-produced sex steroids during the breeding sea-son can be regulated by neurally-produced sex steroids during the non-breeding season, when gonadal production of sex steroids is low orabsent (Pradhan et al., 2010; Soma et al., 1999, 2000).

Taken together, these studies suggest that pair-maintenance be-haviors in zebra finches might be regulated by gonadally-producedsex steroids while pairs are in breeding condition and regulated byneurally-produced sex steroids while pairs are in non-breeding con-dition. As a first step towards testing this hypothesis, we examinedthe effects of water restriction on (1) male and female reproductivephysiology, (2) pair-maintenance behaviors in a variety of behavioralparadigms, and (3) circulating and brain levels of estradiol, testoster-one, and DHEA.

Materials and methods

Subjects

These experiments were carried out under a University of BritishColumbia Animal Care Committee protocol and followed the guide-lines of the Canadian Council on Animal Care. Subjects were adult(>120 d old) captive zebra finches housed in a colony maintainedon a 14:10 h light:dark cycle with an average temperature of 22 °Cand an average relative humidity of 31%. All zebra finches had adlibitum access to seed (50/50, Panicum millet/white millet, Just ForBirds, Langley BC), cuttlefish bone, and grit. Prior to experimentalwater restriction, all subjects had ad libitum access to water. Male–female dyads were housed together in cages (38 1/2″×19 3/4″×19″,Corner's Cages) that had a nestbox (5 1/2 ″×5 1/2 ″×7 1/2 ″) and a cen-ter groove into which a divider could be placed. Dyads were housed to-gether for aminimumof 2 months prior to the start of the experimentalmanipulation. All pairs engaged in affiliative, courtship, and/or nestingbehaviors, and were thus considered pair-bonded.

Pairs were then assigned to one of two treatment groups: control(CON, n=10 pairs) or water restriction (WR, n=11 pairs). Treat-ment groups were counterbalanced with respect to the number ofeggs laid and chicks hatched per pair during the previous 2 months.Water-restricted subjects were given decreasing amounts of waterover the course of 5 weeks, to a minimum of 1 mL per subject perweek, which they continued to receive for the duration of the exper-iment (Table 1). Water-restricted subjects always had access toempty water towers. When water was administered to the water-restricted pairs, a specific amount of water (Table 1) was added tothe water tower for a limited period of time (30 to 120 min), andthe amount of water consumed was estimated to the nearest0.25 mL by measuring (and removing) the remaining water with a pi-pette (Table 1). Control pairs received water ad libitum from theirwater towers. After the start of the experimental manipulation, alleggs laid were removed from all pairs within 48 h of laying, to pre-vent parental behavior from being a confound.

General timeline

A timeline for the experiment is shown in Fig. 1. To assess the effectsof water restriction on baseline behaviors and circulating steroid levels,

Table 1Water consumption by water-restricted female (n=11) and male (n=11) zebra finches.

Week 1 2 3

Water received 3×4 mL 2×3 mL 2×2 mLWater consumed, females (mL) 5.5±0.41 2.6±0.22 2.5±0.2Water consumed, males (mL) 5.2±0.36 2.9±0.22 2.5±0.14

Note: Water received is given in number of times per week×volume at each “watering” (e

Baseline Behavior Sessions were recorded and blood samples werecollected before (Pre) and after (Post) water restriction (Fig. 1). Next,we conducted two behavioral tests that elicited pair-maintenance be-haviors under different conditions: the ‘Partner Preference Test’ andthe ‘Partner Reunion Test’ (see below for details). Immediately follow-ing the Partner Reunion Test, blood and brain tissue were collected forquantification of circulating and brain steroid levels (Fig. 1).

Baseline behavior and circulating steroid levels

Baseline behaviors were assessed during two 20-min sessions(40 min total) both before (Pre) and after (Post) water restriction.Pairs were recorded in their home cages in the colony room, between09:00 and 13:00 h, using a digital camcorder.

To measure the effects of water restriction on circulating steroidlevels under normal conditions, blood samples were collected from thebrachial vein before (Pre) and after (Post) water restriction (Fig. 1).The male and female of each pair were caught simultaneously fromtheir home cage, ~2 d after the second Baseline Behavior Session.Approximately 150 μL of blood was collected into heparinized capillarytubes within 10 min (6.8±0.2 min) of entering the colony room andstored on wet ice. After centrifugation (10 min at 10,000 g), plasmawas collected and stored at −20 °C.

Partner Preference Test and Partner Reunion Test

Both the Partner Preference Test and the Partner Reunion Testtook place in a separate testing room (i.e., not the colony room). Be-tween 11:00 and 13:00 h on the day prior to testing, the pairs weremoved to a testing room.

For the Partner Preference Test, the home cage was placed in be-tween two smaller stimulus cages. Opaque partitions separated thestimulus cages and the home cage. The male or female in the pairwas randomly assigned to be the focal animal, and the other individ-ual was the partner stimulus. The pair was separated immediatelyprior to lights out (21:00 h), the night before the test. The focal ani-mal remained in the home cage, and the partner stimulus was placedin one of the stimulus cages. At the same time, a novel stimulus indi-vidual was placed in the other stimulus cage. Note that the novelstimulus was the same sex and in the same condition (control orwater restricted) as the partner stimulus. The sides of the partnerand novel stimuli were counterbalanced between the two treatmentgroups. The following morning, at the start of the test, the opaquepartitions were removed, and behavior was recorded for 20 minusing a digital camcorder. Note that during the overnight separation,the pair remained in acoustic contact, allowing them to maintaintheir pair bond (Zann, 1996). Previous studies have used longer sep-aration periods, without disruption of the pair bond (Remage-Healeyet al., 2003).

For the Partner Reunion Test, the pairs were again separated imme-diately prior to lights out (21:00 h), the night before the test. The maleand female were placed on opposite sides of the home cage, and theywere physically and visually isolated by a wire partition and an opaquepartition, which were both inserted into a groove in the center of thecage. The following morning, at the start of the test, the opaque parti-tion only was removed (the wire partition remained in place), so thepair was physically but not visually isolated for 10 min. Next, the wire

4 5 6 7 8

1×2 mL 1×1 mL 1×1 mL 1×1 mL 1×1 mL1.2±0.08 0.8±0.02 0.8±0.03 0.7±0.05 0.7±0.061.2±0.06 0.8±0.05 0.8±0.01 0.6±0.04 0.7±0.06

.g., in week 1, individuals received 4 mL of water on 3 separate days).

00

Plasma Sampling

Water RestrictionnB B B B PP PR

Tissue Collection

10 20 30 40 50 60Day

Plasma Sampling

Fig. 1. Experimental timeline. B=Baseline Behavior Sessions, in which subjects were recorded in the home cage in the colony room. PP=Partner Preference Test, and PR=PartnerReunion Test. Plasma samples were collected from the brachial vein from subjects in the home cage in the colony room. Brain tissue and trunk plasma were collected immediatelyafter the Partner Reunion Test. Each tick mark indicates 1 d.

464 N.H. Prior et al. / Hormones and Behavior 63 (2013) 462–474

partition was removed, so the pair was physically reunited for 10 min.Behavior was recorded using a digital camcorder.

Scoring behavior

All behaviors were scored by one researcher (NHP) who wasblind to treatment condition. During the Baseline Behavior Sessions,behaviors that were quantified included general activity (feeding,drinking, and self-preening), nesting behaviors (time spent in thenestbox, and number of trips collecting nestingmaterials) and affiliativebehaviors (clumping [i.e., sitting, facing the same direction, touchingeach other], allopreening, and proximity time [time spent within10 cm of each other]) (Elie et al., 2011a; Goodson et al., 1999). We didnot quantify songs in the colony room because we were unable to reli-ably distinguish the focal male's vocalizations from other males'vocalizations.

In the Partner Preference Test, the cage was divided into threeareas: adjacent to the partner, center/neutral area, or adjacent tothe novel stimulus animal. Time spent adjacent to either the partneror novel stimulus was considered ‘contact time.’ The ‘preference forpartner’ was quantified as time spent adjacent to partner, as a per-centage of contact time (Goodson et al., 2004).

During the Partner Reunion Test, behaviors were scored separatelyduring the first and second halves of the test. Time spent on the centerperch (close to partner) was scored in the first half of the test (visualreunion only), and proximity time within 10 cm was scored duringthe second half of the test (physical reunion). General activity andmale song were also scored during both parts of the test.

Plasma and tissue collection

Systemic and/or local steroid levels might be rapidly and transientlyincreased after affiliative behavior is expressed. To measure circulatingand brain steroid levels immediately after engaging in affiliative behav-ior, we captured both male and female of a pair immediately followingthe Partner Reunion Test. Individuals were euthanized via rapid decap-itation within 3 min (1.2±0.07 min) of entering the testing room.Trunk blood, whole brain, gonads, and oviduct were collected. Thelength and width of the testes, diameter of the largest follicle of theovary, and oviduct length were recorded by a researcher who wasblind to treatment. Tissues were frozen on powdered dry ice and storedat −80 °C. Masses of the testes, whole ovary, and oviduct were mea-sured at a later date.

Brain dissection

Brains were sectioned in the coronal plane at 300 μm on a cryostatat−12 °C. Major neuroanatomical landmarks were used to divide thebrain into several regions of interest (Fig. 2). Specifically, a scalpelwas used to dissect six regions: rostral telencephalon (rTEL), centraltelencephalon (ceTEL), caudal telencephalon (caTEL), hypothalamus(HYP), mid/hindbrain (M/HB), and cerebellum (CB) (Fig. 2). Tissue

of the same region (within an individual) was pooled across multiplesections. The optic tectum was not collected. Brain samples remainedfrozen during this process and were stored at −80 °C.

Steroid extraction and measurement

Brachial plasma samplesPlasma collected from the brachial vein was used to examine the

effect of water restriction on circulating estradiol, testosterone, andcorticosterone levels. Corticosterone was measured in unextractedplasma, as described previously (Newman et al., 2008b). For estradioland testosteronemeasurements, steroids were extracted from plasmausing solid phase (Newman et al., 2008a; Taves et al., 2010, 2011).Plasma (~40 μL) was diluted in 10 mL water and loaded onto C18columns (Agilent Bond-Elut OH, 500 mg, cat # 12113045) that hadbeen primed with 3 mL HPLC-grade methanol and equilibrated with10 mL de-ionized water. Samples were then washed with 10 mL40% HPLC-grade methanol, and steroids were eluted with 5 mL 90%HPLC-grade methanol. The eluted samples were dried at 40 °C in avacuum centrifuge (ThermoElectron SPD111V Speedvac) and storedat −20 °C until assayed.

Extracted steroid samples were resuspended in PBSG (phosphate-buffered saline containing 0.1% gelatin) with absolute ethanol (0.8%) toaid resuspension (Newman et al., 2008a). These resuspended sampleswere used to measure estradiol and/or testosterone using sensitive andspecific radioimmunoassays (RIAs) (Table 2). Plasma samples ≥30 μLwere resuspended in 450 μL; 300 μL was used to quantify estradiol and75 μL was used to quantify testosterone. Plasma samples between 20and 30 μL were resuspended in 350 μL and used to quantify estradiolonly. Plasma samples ≤20 μL were resuspended in 200 μL and used toquantify testosterone only. All samples were measured as singletons(Charlier et al., 2011). All samples from an individual were run in thesame assay. All values were corrected for recovery (Table 2). All brachialplasma samples had detectable levels of steroids.

Brain and trunk plasma samplesBrain and trunk plasma samples were homogenized prior to extrac-

tion. Tissue was homogenized in 2 mL polypropylene microcentrifugetubes with 225 μL ice-cold de-ionized water and 1200 μL HPLC-grademethanol, using a bead homogenizer (Omni Bead Ruptor 24). Threesmall ceramic beads were added, and samples were homogenized for1 min at a speed of 4 m/s. Homogenateswere left at 4 °C overnight. Fol-lowing centrifugation, supernatants were diluted in 10 mL of water,and samples were loaded onto primed and equilibrated C18 columns(~60 μL of plasma and no more than 50 mg of brain tissue per C18column). Samples were washed, eluted, and dried as described above.Steroids were resuspended using 400 μL of PBSG with 1% absoluteethanol. From each sample, 150 μL, 80 μL, 100 μL, and 20 μL weretaken as singletons for estradiol, testosterone, DHEA, and corticosteroneRIAs, respectively. All samples from an individual were run in the sameassay. All values were corrected for recovery (Table 2). Non-detectable

rTEL ceTEL

ceTEL

HYP

HYP

caTEL

caTELcaTEL

M/HB

M/HB

CB

CB

caTEL

caTELcaTEL

M/HB

CB

A B

C D

E F

Fig. 2. Diagram of brain dissections. Coronal sections (300 μm) were made on a cryostat and a scalpel blade was used to dissect brain tissue. Prominent neuroanatomical landmarkswere used to identify regions and make cuts with the scalpel blade (indicated by the dashed lines). The diagram in A is the most rostral section, and the diagram in F is the mostcaudal section. Tissue from the same region was pooled within an individual across sections. caTEL=caudal telencephalon, CB=cerebellum, ceTEL=central telencephalon, HYP=hypothalamus, M/HB=mid/hindbrain, rTEL=rostral telencephalon.


samples were set to zero (estradiol, 97% detectable; testosterone, 100%detectable; DHEA, 93% detectable; corticosterone, 96% detectable).

Statistical analyses

To test for an effect of Treatment (control vs. water restriction) on re-productive physiology (size of gonads and oviduct, number of eggs laid),we used Welch's t-tests. As necessary, data were log-transformed toachieve homogeneity of variances.

Baseline behaviors and steroid levels in brachial plasma were an-alyzed using three-way repeated measures ANOVAs, with Treatmentand Sex as between-subjects factors, and Session (pre- vs. post-water restriction) as a within-subjects factor. When there was a sig-nificant interaction, we used a model reduction technique (simplemain effects) and conducted follow-up ANOVAs within levels of oneof the factors in the interaction. For baseline behaviors and brachialsteroid levels, we conducted follow-up two-way ANOVAs separatelyby Session (pre- vs. post-water restriction). In cases where thesetwo-way ANOVAs yielded a significant Treatment×Sex interaction,

Table 2Radioimmunoassay specifications.

Steroid RIA kit Modification Dete(pg/

17β-Estradiol Beckman-Coulter, DSL-4800 Charlier et al. (2010) 0.2Testosterone MP Biomedicals, cat. 07189102 Overk et al. (2013) 0.3DHEA Beckam-Coulter, DSL 8900 Granger et al. (1999) 2.0Corticosterone MP Biomedicals, cat. 07120103 Washburn et al. (2002) 3.1

we then we conducted one-way ANOVAs to test for an effect of Treat-ment separately in males and females.

In Baseline Behavior Sessions, for infrequent behaviors (allopreening,clumping, copulation, and number of trips carrying nesting materials),Chi-squared tests were used on the frequency of occurrences in pairs.

For the Partner Preference Test, we measured partner preference(% contact time), total contact time, and total time spent adjacent topartner. The data from both sexes were pooled to increase samplesizes (as there were no sex differences) and then analyzed withWelch's t-tests.

For the Partner Reunion Test, time spent on the center perch dur-ing the first 10 min was analyzed by two-way ANOVA with Treat-ment and Sex as between-subjects factors. Time spent singing (bythe male) during the first and second halves of the test and proximitytime during the second half of the test were all analyzed usingWelch's t-tests. The proportion of males singing was analyzed usingChi-squared tests.

Steroid levels in brain regions and trunk plasma were analyzed bytwo-way ANOVAswith Treatment and Sex as between-subjects factors.

ction limittube)

%Recovery Intra-assay variation(mean %CV)

Inter-assay variation(mean %CV)

Plasma Brain

72.3 81.6 5.1 5.190.2 78.9 12.0 11.085.0 90.0 12.2 12.578.4 87.7 10.5 11.0


Steroid levels were analyzed by region because our primary interestwas not to identify regional differences. When there was a significantTreatment×Sex interaction,we conducted follow-up ANOVAs to exam-ine the effect of water restriction within each sex.

All statistics were run in Cran R Statistics 2.14. All data presentedin tables and graphs represent means±SEM.

Results

Reproductive physiology and steroid levels in brachial plasma

Water-restricted females and males consumed similar amounts ofwater (Table 1), but water restriction had very different effects on thetwo sexes. Female reproductive physiology was significantly affectedby water restriction. The sizes of the largest ovarian follicle, ovary,and oviduct were significantly decreased in water-restricted females(Table 3 and Fig. 3A). Further, the total number of eggs laid was signif-icantly lower in water-restricted females (Fig. 3C; t=5.67, pb0.0001).In males, however, there was no significant effect of water restrictionon testes size (Table 3 and Fig. 3B).

We also measured steroids in plasma collected from the brachialvein, both before and after water restriction (Fig. 1). There was noeffect of water restriction on circulating estradiol levels in either sex(Table 4). Water restriction significantly decreased circulating testos-terone levels in males (Table 4, F1,18=6.16, p=0.02) but not infemales (Table 4). The decrease in plasma testosterone levels inmales was not associated with an increase in plasma corticosteronelevels (Table 4). Due to logistical limitations, the brachial plasma sam-ples were not collected within 3 min (6.8±0.2 min), so the cortico-sterone levels in Table 4 do not represent baseline levels.

General activity, nesting and affiliative behavior

Baseline Behavior SessionsGeneral activity was unaffected by water restriction. Feeding and

self-preening levels were similar in control and water-restricted indi-viduals (Table 5, all p values>0.05). As expected, water-restricted in-dividuals stopped visiting the water tower in their cage (which wasempty during Baseline Behavior Sessions) (Table 5, p valuesb0.001).Water restriction significantly decreased time spent in the nest boxby both sexes (Fig. 3D, F1,38=13.23, p=0.0008).

However, water restriction had no effect on baseline affiliative be-haviors. Proximity time was similar in control and water-restrictedpairs (Fig. 4A, all p values>0.05). Additionally, there was no signifi-cant effect of water restriction on allopreening (Table 6, X2 (1)=2.01, p=0.16) or clumping (Table 6, X2 (1)=3.18, p=0.07). Thereis a trend for more clumping in water-restricted pairs.

Table 3Effects of water restriction (WR) on female and male reproductive physiology.

Control WR

FemalesLargest follicle Diameter (mm) 3.9±0.7 1.8±0.4**Ovary Mass (mg) 119.6±10.6 29.9±1.2**Oviduct Length (mm) 81.2±5.6 43.0±5.6***

Mass (mg) 230.7±13.4 65.5±4.1**

MalesLeft testis Length (mm) 4.0±0.2 4.0±0.2

Width (mm) 3.3±0.2 3.1±0.2Volume (mm3) 23.1±3.0 21.5±3.4

Right testis Length (mm) 3.4±0.1 3.5±0.2Width (mm) 2.9±0.2 2.7±0.1Volume (mm3) 15.4±1.8 13.3±1.3

Total testes Mass (mg) 41.3±0.8 40.7±1.0

Note: Bold values indicate significant difference between CON and WR. **p≤0.01,***p≤0.001.

Partner Preference TestWater restriction had no effect on partner preference (Fig. 4B, t=

1.71, p=0.11). There was also no effect of water restriction on con-tact time (CON: 797±110 s, WR: 931±41 s; t=−1.14, p=0.28)or absolute time spent adjacent to the partner (CON: 516±105 s,WR: 425±65 s; t=0.73, p=0.47).

Partner Reunion TestDuring the first 10 min (visual reunion only), there was no effect

of water restriction on time on the center perch in either males or fe-males (Fig. 4C, Treatment, F1,36=1.29, p=0.26; Sex, F1,36=0.0043,p=0.95; Treatment×Sex, F1,36=1.004, p=0.31). Almost all malessang during the first half of the test: 9/10 control males and 8/10water-restricted males. There was no effect of water restriction ontime spent singing during the first half of the test (CON: 15.1±4.7 s,WR: 11.6±3.4 s; t=0.79, p=0.44).

During the second 10 min (physical reunion) there was no effectof water restriction on proximity time (Fig. 4D, t=0.83, p=0.42).Relative to the first half of the test, fewer males sang during the sec-ond half of the test: 4/10 control males and 2/10 water-restrictedmales. There was no effect of water restriction on time spent singingduring the second half of the test (CON: 13.7±9.11 s, WR: 0.82±0.55 s; t=1.41, p=0.19).

Steroid levels in trunk plasma and brain

We collected trunk plasma and the brain immediately after thePartner Reunion Test. All subjects were euthanized within 3 min ofentering the testing room.

EstradiolWater restriction had no effect on plasma levels of estradiol inmales

or females (Fig. 5, Table 7). There was a significant main effect of Treat-ment on estradiol levels in ceTEL, M/HB, and CB (CON>WR, Fig. 5,Table 7). Therewas a significantmain effect of Sex in CB (female>male,Fig. 5, Table 7). There was a significant Treatment×Sex interaction incaTEL; examining the effect of water restriction within each sex yieldeda significant effect of Treatment in males only (Fig. 5, Table 7, F1,19=12.75, p=0.002). There was no effect of Treatment or Sex in anyother region.

TestosteroneAs expected, testosterone levels in plasma were higher in males

than in females (Fig. 6, Table 7). Water restriction significantlydecreased plasma testosterone levels in males (Fig. 6, F1,17=13.61,p=0.002) but not in females. There were significant main effects ofTreatment and Sex on testosterone levels in the rTEL (Fig. 6,Table 7). There was a significant Treatment×Sex interaction on tes-tosterone levels in caTEL and CB. Examining the effect of water re-striction within each sex yielded significant effects of Treatment inmales only (Fig. 6, Table 7, caTEL, F1,19=4.94, p=0.04; CB, F1,19=9.42, p=0.006). There was no effect of Treatment or Sex in anyother region.

DHEAWater restriction had no effect on plasma DHEA levels in males or

females (Fig. 7, Table 7). DHEA levels in plasma were higher in malesthan in females (Fig. 8, Table 7). There were no significant maineffects of Treatment or Sex, or a significant Treatment×Sex interactionin any brain region for DHEA (Fig. 7, Table 7).

CorticosteroneWater restriction had no effects on plasma corticosterone levels

(Fig. 8, Table 7). Additionally, there were no significant main effects ofTreatment or Sex, or a significant Treatment×Sex interaction in anybrain region (Fig. 8, Table 7).

0

150

300

450

***

***

Tim

e Sp

ent

in N

estb

ox (

s)

Pre Post

F CON

F WR

M CON

M WR

CON WR

CON WR

CON WR

Lar

gest

Ova

rian

F

ollic

le V

olum

e (m

m3 ) 5

4

3

2

1

**

0

3

6

9

Num

ber

of E

ggs

Tot

al T

este

s V

olum

e (m

m3 )

50

40

30

20

10

0

***

A B

C D

Fig. 3. Effects of water restriction (WR) on reproductive physiology in females (F) and males (M). Effects of WR on (A) largest ovarian follicle volume, (B) total testes volume, (C) thenumber of eggs laid, and (D) the time spent in the nestbox. Relative to control subjects (CON) with ad libitum water, WR strongly reduced largest ovarian follicle volume (t=3.25,p=0.004) but had no effect on total testes volume (t=0.52, p=0.61). WR also decreased the number of eggs laid (t=5.67, pb0.0001) and time spent in the nestbox in the post-WRperiod (F1,38=13.23, p=0.0008). **p≤0.01, ***p≤0.001.


Discussion

These data demonstrate for the first time in this important modelspecies (1) the robust effects of experimental water restriction on fe-male reproductive physiology, (2) the lack of an effect of water restric-tion on pair-maintenance behavior, and (3) the differential effects ofwater restriction on systemic and local sex steroid levels. Importantly,although water restriction strongly affected the ovary and oviduct infemales and plasma testosterone levels in males, water restriction didnot significantly reduce estradiol levels in the hypothalamus or testos-terone levels in the hypothalamus and central telencephalon. Whilethese data are strictly correlational, they are consistent with thehypothesis that neurosteroids promote pair-maintenance behaviors,including male song, in non-breeding zebra finches, a hypothesis thatwill be tested further in future studies.

Effects of water restriction on reproductive physiology

Zebra finches are opportunistic breeders and can exist in a rangeor continuum of reproductive states (Perfito, 2010). We thereforedid not expect to see dichotomous “breeding” and “non-breeding”

Table 4Steroid concentrations in plasma collected from the brachial vein.

Estradiol (pg/mL) Testo

Pre Post Pre

Female CON 34.7±3.1 (8) 40.3±7.8 (10) 0.2±WR 35.3±4.8 (6) 31.4±2.9 (11) 0.2±

Male CON 35.8±1.4 (7) 34.0±2.6 (9) 2.9±WR 32.3±2.0 (9) 28.9±2.1 (11) 2.2±

Note: Sample sizes are in parentheses. Bolded value indicates significant difference betwee

conditions, as one would see in a temperate seasonally-breeding spe-cies. Nonetheless, control and water-restricted subjects did exhibitdifferent physiological and endocrine profiles, consistent with fielddescriptions of breeding and non-breeding zebra finches, respectively(Perfito, 2010; Perfito et al., 2007).

The present water restriction protocol is intermediate betweenprevious studies, which range from complete water deprivation(Sossinka, 1974) to slowly decreasing water over 11 weeks (Perfitoet al., 2006). Most previous research has focused on the effects ofwater restriction on isolated males. In early studies, there were no ef-fects of water restriction or deprivation on testis size (Oksche et al.,1963; Sossinka, 1974). However, in wild male zebra finches, testisvolume is smaller in non-breeding males (15 to 40% reduction)(Perfito et al., 2007, 2011). Further, in captive male zebra finches, ifa within-subjects design is used, then subtle effects of water restric-tion on testis size are detectable (Perfito et al., 2006; Vleck andPriedkalns, 1985). Here, water restriction had no significant effectson testis volume or mass. Possibly whenmale zebra finches are pairedwith females, a longer or more severe water restriction is necessary,or perhaps we did not detect a subtle effect of water restriction be-cause we used a between-subjects design.

sterone (ng/mL) Corticosterone (ng/mL)

Post Pre Post

0.01 (5) 0.3±0.02 (10) 21.7±3.4 (4) 13.5±2.8 (10)0.02 (7) 0.2±0.02 (10) 15.5±3.4 (4) 17.2±2.6 (10)0.2 (9) 1.9±0.4 (9) 11.4±3.6 (2) 9.4±0.9 (7)0.6 (8) 0.7±0.2* (11) 14.1±4.3 (6) 17.2±1.8 (11)

n CON and WR in the Post-WR time period. * p≤0.05.

Table 5Effect of water restriction on general activity during Baseline Behavior Sessions.

# visits to the water tower % time spent feeding % time spent self-preening

Pre Post Pre Post Pre Post

Female CON 1.75±0.4 2.00±0.5 4.39±1.1 5.98±1.8 8.53±3.6 5.67±2.3WR 2.09±0.6 0.00±0.0⁎⁎⁎ 5.65±1.0 5.56±0.8 6.63±2.0 2.92±1.5

Male CON 1.45±0.4 1.80±0.5 1.96±0.5 3.43±1.2 6.80±2.9 2.59±0.6WR 1.64±0.3 0.05±0.1⁎⁎⁎ 2.34±0.2 4.34±1.0 7.37±2.3 3.68±1.6

Note: Bolded values indicate significant difference between CON and WR in the Post-WR time period.⁎⁎⁎ p≤0.001.


In contrast, we saw a strong effect of water restriction on femalereproductive physiology. Few previous studies have examined the ef-fects of water restriction on female zebra finches. In wild female zebrafinches, follicular volume is greatly reduced in non-breeding females(90% reduction) (Perfito et al., 2007), which is consistent with our re-sults. Here, there was also a strong inhibitory effect of water restric-tion on oviduct size and egg laying. Overall, water restriction clearlyreduced reproductive readiness in females.

We also report for the first time the effects of water restriction oncirculating sex steroid levels. Circulating luteinizing hormone levelsare decreased in wild non-breeding male and female zebra finches(Perfito et al, 2007). However, in captivity, water restricting males andfemales do not necessarily reduce luteinizing hormone levels (Vleckand Priedkalns, 1985). Additionally, Vleck and Priedkalns (1985) mea-sured circulating androgens in males and estrogens in females, usingpooled plasma samples. However, their steroid assays were not as sen-sitive, andmost samples from control andwater-restricted subjects hadnon-detectable values. Here, using ultra-sensitive assays, we saw that

0

20

40

60

80

100

PRE POSTCON WR CON WR

F MF MCON WR

% T

ime

on t

he C

ente

r P

erch

% P

roxi

mit

y T

ime

< 10

cm

0

20

40

60

80

100

A) Baseline Behavior Sessions

C) Partner Reunion Test (1st Half)

Fig. 4. No effects of water restriction (WR) on pair-maintenance behaviors. WR had noparadigms: (A) Baseline Behavior Sessions, (B) Partner Preference Test, (C) 1st half of theTest (physical reunion).

water restriction decreased circulating testosterone levels in malesbut not in females. Interestingly, this effect on plasma testosterone isthe opposite of what we observedwith gonad size, wherewater restric-tion had a greater effect in females than in males.

There were no effects of water restriction on circulating estradiollevels in males or females. The primary source of circulating estradiol inmales is the brain, whereas the primary source in females is the ovary(Schlinger and Arnold, 1993). The lack of an effect of water restrictionon female plasma estradiol levels is surprising, given the clear effects ofwater restriction on the ovary and oviduct. The estradiol assay usedhere is highly accurate, precise, specific, and sensitive (can detect as littleas 0.2 pg of estradiol) (Charlier et al., 2010, 2011; Taves et al., 2010, 2011).It is possible that plasma estradiol levels are generally low in controlfemales, with only transient increases at specific times (e.g., ovulation).Alternatively, estrone, progesterone, prolactin, or arginine vasotocin(AVT) might be modulated by water restriction and affect female repro-ductive physiology (Li et al., 2011; Liu and Bacon, 2005; Srivastava andChaturvedi, 2010), but these hormones were not measured here.

0

20

40

60

80

100

CON WR

CON WR

% P

roxi

mit

y T

ime

< 10

cm

Pre

fere

nce

for

Par

tner

(% C

onta

ct T

ime)

0

20

40

60

80

100

B) Partner Preference Test

D) Partner Reunion Test (2nd Half)

effects on physical proximity, the primary measure of affiliation, in the 3 behavioralPartner Reunion Test (visual reunion only) and (D) 2nd half of the Partner Reunion

Table 6Number of pairs that engaged in infrequent affiliative and reproductive behaviorsduring Baseline Behavior Sessions.

Allopreening Clumping Copulations Carrying nestingmaterials

Pre Post Pre Post Pre Post Pre Post

CON 1 1 0 0 2 1 4 5WR 6 4 0 3 2 0 6 1⁎

Note: Bolded value indicates significant difference between CON and WR in thePost-WR time period. n=10 CON pairs and 11 WR pairs.⁎ p≤0.05.


While stress is known to affect reproductive physiology (Kellyet al., 2011; Lynn et al., 2010; Perfito, 2010; Roberts et al., 2007),the effects of water restriction seen here do not appear to be the resultof stress. Consistent with previous research (Perfito et al., 2007), therewere no effects of water restriction on circulating or brain corticoste-rone levels, body mass (data not shown), or general activity.

Measures of pair-bonding and pair-maintenance

Zebra finch social behavior has been extensively studied (Goodsonand Kingsbury, 2011; Griffith and Buchanan, 2010; Healy et al., 2010;Zann, 1996); however, this is one of the few studies to focus onpair-maintenance behaviors in this species with long-term monoga-my (Dunn and Zann, 2010; Elie et al., 2011a; Tomaszycki andAdkins-Regan, 2006). Wild zebra finch pairs remain pair-bonded re-gardless of breeding condition (Zann, 1994). However, how captivepairs would perform in traditional behavioral paradigms was un-known. Therefore, we examined a variety of affiliative behaviorsunder different paradigms. We also ensured that all of our pairswere bonded prior to the start of water restriction: pairs were housedfor a minimum of 2 months, had laid eggs, and nearly all had chicks(Zann, 1994).

0

15

30

45

60

75

90

0

15

30

45

60

75

90

0

15

30

45

60

75

90

0

15

30

45

60

75

90

Est

radi

ol (

pg/m

L)

Females

Males

Est

radi

ol (

pg/m

L)

Est

radi

ol (

pg/g

)E

stra

diol

(pg

/g)

Plasma HYP

Fig. 5. Effects of water restriction (WR) on plasma and brain estradiol levels. Samples weestradiol levels in males or females, but decreased estradiol levels in some brain regions.and WR groups pooled) for females and for males. *p≤0.05, **p≤0.01, ***p≤0.001.

We used physical proximity as a primary measure of pair-maintenance behavior in all three behavioral testing paradigms.Physical proximity is a combination of coordination and following be-haviors, both of which are very reliable measures of pair-bonding andattachment (Zann, 1996). Water restriction had no effect on physicalproximity in any testing paradigm. First, there was no effect of waterrestriction on physical proximity (time spent within 10 cm) duringBaseline Behavior Sessions, in which we observed normal behaviorin the home cage within the colony. Second, during the Partner Prefer-ence Test, there was no effect of water restriction on physical proximity(time spent adjacent to partner). Third, during the Partner ReunionTest, there was no effect of water restriction on physical proximityduring the visual reunion phase (time spent on the center perch) orduring the physical reunion phase (time spent within 10 cm). Asexpected, the amount of time spent within 10 cm was higher duringthe Partner Reunion Test than during Baseline Behavior Sessions.Taken together, these data indicate that water restriction does not re-duce pair-maintenance behavior, consistent with observations of wildzebra finches (Zann, 1996).

Choice testing paradigms have been effectively used to determinezebra finch mate preferences and group size preference (Adkins-Regan and Leung, 2006; Goodson et al., 2009; Svec and Wade, 2009;Svec et al., 2009; Tomaszycki and Adkins-Regan, 2005), but partnerpreference was lower than what is typically seen in prairie voles(Goodson et al., 2004; Young et al., 2008). Zebra finches are highlygregarious and do not engage in high levels of extra pair copulations;thus one might expect the focal bird to interact with both the partnerand the novel stimulus animal (Birkhead et al., 1988, 1990; Elie et al.,2011a,b).

In addition to physical proximity, we also measured other affiliativebehaviors such as clumping, allopreening, and male song, all of whichoccurred infrequently here.While clumping and allopreening are highlyreliablemeasures of pair-bonding and attachment, levels vary consider-ably from study to study (Alger et al., 2011; Elie et al., 2011b; Goodson

***

***

**

**

*

***

F CON

F WR

M CON

M WR

rTEL ceTEL caTEL M/HB CB

re collected immediately after the Partner Reunion Test. WR had no effect on plasmaThe dashed lines on the brain graphs indicate the mean plasma estradiol level (CON

Table 7ANOVA results from trunk plasma and brain steroid levels.

Region Factor Estradiol Testosterone DHEA Corticosterone

F P F P F P F P

Plasma WR 1.07 0.31 10.58 0.002 1.15 0.29 0.07 0.80Sex 1.08 0.31 27.58 0.000 4.28 0.05 1.79 0.19WR∗sex 0.34 0.57 16.08 0.000 0.51 0.48 2.19 0.15

Hypothalamus WR 2.80 0.10 0.73 0.40 1.59 0.22 0.02 0.87Sex 1.82 0.19 0.20 0.66 1.19 0.28 0.20 0.65WR∗sex 0.00 1.00 2.83 0.10 0.06 0.81 0.68 0.42

Rostral telencephalon WR 2.31 0.14 4.77 0.04 0.12 0.74 1.60 0.22Sex 0.01 0.93 4.33 0.04 0.01 0.91 3.90 0.06WR∗sex 0.44 0.51 0.14 0.71 0.53 0.82 1.20 0.29

Central telencephalon WR 10.20 0.003 1.69 0.20 0.06 0.80 0.72 0.40Sex 0.22 0.64 7.75 0.008 1.59 0.22 3.82 0.06WR∗sex 0.01 0.91 0.54 0.47 0.01 0.93 0.05 0.83

Caudal telencephalon WR 16.52 0.000 0.97 0.33 0.37 0.55 0.90 0.35Sex 10.50 0.002 6.14 0.02 0.09 0.77 3.14 0.08WR∗sex 6.13 0.02 5.80 0.02 0.00 0.97 0.19 0.67

Mid/hindbrain WR 6.60 0.01 3.46 0.07 0.28 0.60 0.77 0.39Sex 0.01 0.92 1.24 0.27 0.09 0.76 0.00 1.00WR∗sex 0.83 0.37 2.84 0.10 0.42 0.52 0.00 0.95

Cerebellum WR 19.57 0.000 9.94 0.003 0.65 0.43 1.40 0.24Sex 7.32 0.01 12.12 0.001 0.04 0.84 0.60 0.45WR∗sex 0.14 0.71 5.23 0.03 0.83 0.37 0.00 0.96

Note: Bold values indicate significant effects of water restriction (WR), sex, or the interaction. Significant interactions were followed up with ANOVAs to examine the effect of waterrestriction within each sex.


et al., 2004; Tomaszycki et al., 2006). In any case, water restriction didnot reduce these behaviors in Baseline Behavior Sessions or the PartnerReunion Test, or reducemale song during the Partner Reunion Test. Thisis consistent with our data on physical proximity in these behavioraltests and previous work on wild zebra finches (Zann, 1996).

Effects of water restriction on brain steroid levels

Plasma and brain were collected immediately after the PartnerReunion Test. Thus, the steroid concentrations measured in the plas-ma and brain might not represent “baseline” levels but rather levelsthat are transiently elevated in response to reunion with the partner.Indeed, several studies have shown that social cues rapidly regulateneurosteroid synthesis (Dickens et al., 2012; Pradhan et al., 2010;Remage-Healey et al., 2008).

Plasma corticosterone levels increase following partner separationand return to baseline levels between 24 and 48 h following reunion(Remage-Healey et al., 2003). Consistent with this, circulating levels ofcorticosterone in control subjects were elevated here, relative to previouswork ( Remage-Healey et al., 2003; Taves et al., 2010). Also, circulatingDHEA and testosterone levels in control males were reduced comparedto levels under normal conditions (see Table 4 here and Taves et al.,2010). However, circulating levels of estradiol were similar to levelsunder normal conditions (see Table 4 here and Taves et al., 2010).

DHEA is a prohormone and a precursor to testosterone. Here, cir-culating and brain levels of DHEA were not affected by water restric-tion in males and females. Given that circulating testosterone levelswere low in water-restricted males and females, the maintenance ofDHEA levels could allow for local synthesis of testosterone and estra-diol in brain regions that express the necessary steroidogenic en-zymes (Pradhan et al., 2010; Schlinger et al., 2008). Similarly, insong sparrows, circulating and brain levels of DHEA are high in thebreeding and non-breeding seasons (Newman and Soma, 2009,2011; Newman et al., 2008b; Soma and Wingfield, 2001).

Water restriction affected brain testosterone and estradiol levels in aregion-specificmanner, sometimes in parallel with and sometimes inde-pendent of changes in plasma levels. Further, in some regions, such asthe hypothalamus, brain testosterone and estradiol levels were farhigher than plasma levels (especially in water-restricted subjects).Brain steroid levels are result of several factors, including circulating ste-roid levels, circulating steroid binding proteins, local steroid synthesis

and catabolism, and tissue sequestration by steroid receptors or bindingproteins (Schmidt et al., 2008; Taves et al., 2011).

Circulating testosterone levels in male zebra finches were stronglyreduced by water restriction; plasma testosterone levels were ex-tremely low in all females. Water restriction decreased testosteronelevels in some brain regions: the female rTEL and the male rTEL,caTEL, and CB. In males, the above decreases in regional testosteronelevels are likely the result of the decrease in circulating testosteronelevels. One the other hand, the maintenance of testosterone levelsin some regions of the male brain could be the result of increasesin local testosterone production and/or sequestration. Testosteronelevels in the hypothalamus are particularly intriguing. In femalesand males, testosterone levels were far higher in the hypothalamusthan in the plasma. In addition, water restriction did not significantlyreduce hypothalamic testosterone levels in males and even showed atrend to increase hypothalamic testosterone levels in females. Thismaintenance of testosterone levels in the male hypothalamus (andcentral telencephalon) might be due to an up-regulation of neuralsteroidogenic enzymes such as 3β-hydroxysteroid dehydrogenase(3β-HSD), which plays a critical role in DHEA metabolism in thezebra finch brain (Schlinger et al., 2008; Soma et al., 2004). In malesong sparrows, 3β-HSD activity in the forebrain is upregulated duringthe non-breeding season, when circulating testosterone levels arelow (Pradhan and Soma, 2012; Pradhan et al., 2010). Future studiescan measure 3β-HSD and other steroidogenic enzymes in brain re-gions from control and water-restricted zebra finches.

Water restriction had no effect on circulating estradiol levels in eithermales or females, but did reduce brain estradiol levels in some regions.Water restriction decreased estradiol levels in males in 4 regions (ceTEL,caTEL,M/HBandCB) anddecreased estradiol levels in females in 3 regions(ceTEL, M/HB, and CB). Note that in some regions (e.g., central telenceph-alon), local estradiol levels, but not local testosterone levels, were reducedbywater restriction. This pattern of results suggests that water restrictionreduces aromatase activity or estrogen receptors in specific regions, andfuture studies can examine aromatase and estrogen receptors.

The hypothalamus contains nuclei that are part of the “social be-havior network” (e.g., preoptic area, ventromedial hypothalamus)and likely important in pair-maintenance behavior (Goodson, 2005;Newman, 1999). Hypothalamic testosterone and estradiol levelswere similar in males and females, and they were not significantly de-creased by water restriction. Additionally, testosterone and estradiol

0

0.2

0.4

0.6

0.8

1

1.2

0

0.2

0.4

0.6

0.8

1

1.2

0

0.2

0.4

0.6

0.8

1

1.2

0

0.2

0.4

0.6

0.8

1

1.2

Tes

tost

eron

e (n

g/m

L)

Tes

tost

eron

e (n

g/m

L)

Tes

tost

eron

e (n

g/g)

Tes

tost

eron

e (n

g/g)

Females

Males

***

F CON

F WR

M CON

M WR

***

*

HYP rTEL ceTEL caTEL M/HB CBPlasma

Fig. 6. Effects of water restriction (WR) on plasma and brain testosterone levels. Samples were collected immediately after the Partner Reunion Test. WR had no effect on plasmatestosterone levels in females but significantly decreased plasma testosterone levels in males. Additionally, WR decreased testosterone levels in some brain regions. The dashed lineon the female brain graph indicates the mean plasma testosterone level (CON and WR groups pooled) for females. The solid line on male brain graph indicates the mean plasmatestosterone level in CON males, and the dotted line indicates the mean plasma testosterone level in WR males. *p≤0.05, **p≤0.01.


levels were higher in the hypothalamus than in the plasma, whichsuggests local sex steroid synthesis (Taves et al., 2011). Previous stud-ies have shown that the adult zebra finch hypothalamus expresses

0

0.5

1

1.5

2

0

0.5

1

1.5

2

0

0.5

1

1.5

2

0

0.5

1

1.5

2

DH

EA

(ng

/mL

)D

HE

A (

ng/m

L)

Females

Males

DH

EA

(ng

/g)

DH

EA

(ng

/g)

HYP rPlasma

Fig. 7. No effects of water restriction (WR) on plasma and brain dehydroepiandrosterone (Ddashed lines on the brain graphs indicate the mean plasma DHEA level (CON and WR grou

several steroidogenic enzymes, including CYP11A1, CYP17, 3β-HSD,17β-HSD, and aromatase (London et al., 2006; Schlinger andRemage-Healey, 2012) and thus has the capacity to synthesize sex

F CON

F WR

M CON

M WR

TEL ceTEL caTEL M/HB CB

HEA) levels. WR had no effect on plasma or brain DHEA levels in males or females. Theps pooled) for females and for males.

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

0

2

4

6

8

10

Females

Males

F CON

F WR

M CON

M WR

Plasma

Cor

tico

ster

one

(ng/

mL

)C

orti

cost

eron

e (n

g/m

L)

Cor

tico

ster

one

(ng/

g)C

orti

cost

eron

e (n

g/g)

HYP rTEL ceTEL caTEL M/HB CB

Fig. 8. No effects of water restriction (WR) on plasma and brain corticosterone levels. WR had no effect on plasma or brain corticosterone levels in males or females. The dashedlines on the brain graphs indicate the mean plasma corticosterone level (CON and WR groups pooled) for females and for males.


steroids de novo from cholesterol. Future studies can focus on mea-suring sex steroids and steroidogenic enzymes within specific hypo-thalamic nuclei and portions of the social behavior network, as wellas their roles in the expression of affiliative behavior.

In zebra finches, the roles of sex steroids in the regulation ofpair-maintenance behaviors have not been studied. Numerous stud-ies show that sex steroids promote courtship behavior and malesong in zebra finches (Ball et al., 2002; Cynx et al., 2005; Hardinget al., 1983; Walters et al., 1991). However, there is also evidencethat sex steroids may not play a role in pair-bond formation(Tomaszycki et al., 2006). This study used the steroidal aromataseinhibitor ATD (Tomaszycki et al., 2006), which is much less potentthan non-steroidal aromatase inhibitors such as fadrozole (Wadeet al., 2004). Future studies should examine the effects of fadrozoleon pair-bond formation and maintenance. Moreover, overtly similarbehaviors can be regulated differently across contexts (Wingfieldet al., 2001), and therefore it is possible that pair-bond formationand pair-bond maintenance are regulated differently.

Conclusions

Water restriction affects the physiological and endocrine profilesof both male and female zebra finches, consistent with an inductionof non-breeding condition, particularly in females. Nonetheless,pair-maintenance behavior and sex steroid levels in some brain re-gions, such as the hypothalamus, are not reduced by water restriction.While correlational, these data are nonetheless consistent with thehypothesis that local production of sex steroids in the brain promotesthe expression of pair-maintenance behaviors in non-breeding zebrafinches, a hypothesis that will be directly tested in future experimentsby pharmacologically inhibiting sex steroid synthesis or action inwater-restricted individuals.

Acknowledgments

We thank Pavendeep Gill, Annika Sun, Alice Chan, and AnneCheng for help with animal husbandry and data collection; Dr. H.Bobby Fokidis for steroid recovery data; and Matthew D. Taves andBenjamin A. Sandkam for feedback on the manuscript. This study issupported by an Operating Grant from the Canadian Institutes ofHealth Research (CIHR) to KKS, postdoctoral fellowships from CIHRand the Michael Smith Foundation for Heath Research to SAH, anda graduate research fellowship from the National Science Foundationto NHP.

References

Abartz, J., Hollander, E., 2006. The neuroscience of affiliation: forging links betweenbasic and clinical research on neuropeptides and social behavior. Horm. Behav.50, 518–528.

Adkins-Regan, E., Leung, C.H., 2006. Sex steroids modulate changes in social and sexualpreference during juvenile development in zebra finches. Horm. Behav. 50,772–778.

Alger, S.J., Juang, C., Riters, L.V., 2011. Social affiliation relates to tyrosine hydroxylaseimmunolabeling in male and female zebra finches (Taeniopygia guttata). J. Chem.Neuroanat. 42, 45–55.

Arnold, A., 1975. The effects of castration and androgen replacement on song, court-ship, and aggression in zebra finches (Poephila guttata). J. Exp. Zool. 191, 309–326.

Ball, G.F., Riters, L.V., Balthazart, J., 2002. Neuroendocrinology of song behavior andavian brain plasticity: multiple sites of action of sex steroid hormones. Front.Neuroendocrinol. 23, 137–178.

Balthazart, J., Ball, G.F., 2006. Is brain estradiol a hormone or a neurotransmitter?Trends Neurosci. 29, 241–249.

Birkhead, T., Clarkson, K., Zann, R., 1988. Extra-pair courtship, copulation and mateguarding in wild zebra finches Taeniopygia guttata. Anim. Behav. 36, 1853–1855.

Birkhead, T., Burke, T., Zann, R., Hunter, F., Krupa, A., 1990. Extra-pair paternity and in-traspecific brood parasitism in wild zebra finches Taeniopygia guttata, revealed byDNA fingerprinting. Behav. Ecol. Sociobiol. 27, 315–324.

Buck, R., 2002. The genetics and biology of true love: prosocial biological affects and theleft hemisphere. Psychol. Rev. 109, 739–744.

Carter, C., 1998. Neuroendocrine perspectives on social attachment and love.Psychoneuroendocrinology 23, 779–818.


Charlier, T.D., Po, K.W.L., Newman, A.M.E., Shah, A.H., Saldanha, C.J., Soma, K.K., 2010.17β-Estradiol levels in male zebra finch brain: combining Palkovits punch and anultrasensitive radioimmunoassay. Gen. Comp. Endocrinol. 167, 18–26.

Charlier, T.D., Newman, A.E.M., Heimovics, S.A., Po, K.W.L., Saldanha, C.J., Soma, K.K., 2011.Rapid effects of aggressive interactions on aromatase activity and oestradiol in discretebrain regions of wild male white-crowned sparrows. J. Neuroendocrinol. 23, 742–753.

Cynx, J., Bean, N.J., Rossman, I., 2005. Testosterone implants alter the frequency rangeof zebra finch songs. Horm. Behav. 47, 446–451.

Dickens, M.J., Balthazart, J., Cornil, C.A., 2012. Brain aromatase and circulating cortico-sterone are rapidly regulated by combined acute stress and sexual interaction ina sex-specific manner. J. Neuroendocrinol. 24, 1322–1334.

Dunn, A.M., Zann, R., 2010. Undirected song encourages the breeding female zebrafinch to remain in the nest. Ethology 102, 540–548.

Elie, J.E., Mariette, M.M., Soula, H.A., Griffith, S.C., Mathevon, N., Vignal, C., 2010. Vocalcommunication at the nest between mates in wild zebra finches: a private vocalduet? Anim. Behav. 80, 597–605.

Elie, J.E., Mathevon, N., Vignal, C., 2011a. Same-sex pair-bonds are equivalent to male–female bonds in a life-long socially monogamous songbird. Behav. Ecol. Sociobiol.65, 2197–2208.

Elie, J.E., Soula, H.A., Mathevon, N., Vignal, C., 2011b. Dynamics of communal vocaliza-tions in a social songbird, the zebra finch (Taeniopygia guttata). J. Acoust. Soc. Am.129, 4037–4046.

Forlano, P.M., Schlinger, B.A., Bass, A.H., 2006. Brain aromatase: New lessons from non-mammalian model systems. Front. Neuroendocrinol. 27, 247–274.

Goodson, J.L., 2005. The vertebrate social behavior network: evolutionary themes andvariations. Horm. Behav. 48, 11–22.

Goodson, J.L., Kingsbury, M.A., 2011. Nonapeptides and the evolution of social groupsizes in birds. Front. Neuroanat. 5, 13.

Goodson, J.L., Eibach, R., Sakata, J., Adkins-Regan, E., 1999. Effect of septal lesions onmale song and aggression in the colonial zebra finch (Taeniopygia guttata) andthe territorial field sparrow (Spizella pusilla). Behav. Brain Res. 98, 167–180.

Goodson, J.L., Lindberg, L., Johnson, P., 2004. Effects of central vasotocin and mesotocinmanipulations on social behavior in male and female zebra finches. Horm. Behav.45, 136–143.

Goodson, J.L., Evans, A., Lindberg, L., Allen, C., 2005. Neuro-evolutionary patterning ofsociality. Proc. R. Soc. B Biol. Sci. 272, 227–235.

Goodson, J.L., Schrock, S.E., Klatt, J.D., Kabelik, D., Kingsbury, M.A., 2009. Mesotocin andnonapeptide receptors promote estrildid flocking behavior. Science 325, 862–866.

Granger, D.A., Schwartz, E.B., Booth, A., Curran, M., Zakaria, D., 1999. Assessing dehy-droepiandrosterone in saliva: a simple radioimmunoassay for use in studies ofchildren, adolescents and adults. Psychoneuroendocrinology 24, 567–579.

Griffith, S.C., Buchanan, K.L., 2010. The zebra finch: the ultimate Australian supermodel.Emu 110, v.

Griffith, S.C., Holleley, C.E., Mariette, M.M., Pryke, S.R., Svedin, N., 2010. Low level ofextrapair parentage in wild zebra finches. Anim. Behav. 79, 261–264.

Harding, C.F., Rowe, S.A., 2003. Vasotocin treatment inhibits courtship in male zebra finches;concomitant androgen treatment inhibits this effect. Horm. Behav. 44, 413–418.

Harding, C.F., Sheridan, K., Walters, M.J., 1983. Hormonal specificity and activation ofsexual behavior in male zebra finches. Horm. Behav. 17, 111–133.

Healy, S.D., Haggis, O., Clayton, N.S., 2010. Zebra finches and cognition. Emu 110,242–250.

Hill, W.L., Ballard, S., Coyer, M.J., Rowley, T., 2005. The interaction of testosterone andbreeding phase on the reproductive behavior and use of space of male zebrafinches. Horm. Behav. 47, 452–458.

Kelly, A.M., Kingsbury, M.A., Hoffbuhr, K., Schrock, S.E., Waxman, B., Kabelik, D., Thompson,R.R., Goodson, J.L., 2011. Vasotocin neurons and septal V1a-like receptors potentlymodulate songbird flocking and responses to novelty. Horm. Behav. 60, 12–21.

Kikusui, T., Winslow, J.T., Mori, Y., 2006. Social buffering: relief from stress and anxiety.Philos. Trans. R. Soc. B-Biol. Sci. 361, 2215–2228.

Li, W.L., Liu, Y., Yu, Y.C., Huang, Y.M., Liang, S.D., Shi, Z.D., 2011. Prolactin plays a stim-ulatory role in ovarian follicular development and egg laying in chicken hens.Domest. Anim. Endocrinol. 41, 57–66.

Lim, M.M., Young, L.J., 2006. Neuropeptidergic regulation of affiliative behavior andsocial bonding in animals. Horm. Behav. 50, 506–517.

Liu, H.K., Bacon, W.L., 2005. Changes in egg production rate induced by progesteroneinjection in broiler breeder hens. Poult. Sci. 84, 321–327.

London, S.E., Monks, A., Wade, J., Schlinger, B.A., 2006. Widespread capacity for steroidsynthesis in the avian brain and song system. Endocrinology 147, 5975–5987.

Lynn, S.E., Stamplis, T.B., Barrington, W., Weida, N., Hudak, C.A., 2010. Food, stress, andreproduction: short-term fasting alters endocrine physiology and reproductive be-havior in the zebra finch. Horm. Behav. 58, 214–222.

Morton, S., 2009. Rain and grass: lessons in how to be a zebra finch. In: Robin, L.,Heinsohn, R., Joseph, L. (Eds.), Boom and Bust. Csiro Publishing.

Neumann, I.D., 2009. The advantage of social living: brain neuropeptides mediate the bene-ficial consequences of sex and motherhood. Front. Neuroendocrinol. 30, 483–496.

Newman, S.W., 1999. The medial extended amygdala in male reproductive behavior anode in the mammalian social behavior network. Ann. N. Y. Acad. Sci. 877, 242–257.

Newman, A.E.M., Soma, K.K., 2009. Corticosterone and dehydroepiandrosterone insongbird plasma and brain: effects of season and acute stress. Eur. J. Neurosci. 29,1905–1914.

Newman, A.E.M., Soma, K.K., 2011. Aggressive interactions differentially modulate localand systemic levels of corticosterone and DHEA in a wild songbird. Horm. Behav.60, 389–396.

Newman, A.E.M., Chin, E.H., Schmidt, K.L., Bond, L., Wynne-Edwards, K.E., Soma, K.K.,2008a. Analysis of steroids in songbird plasma and brain by coupling solid phaseextraction to radioimmunoassay. Gen. Comp. Endocrinol. 155, 503–510.

Newman, A.E.M., Pradhan, D.S., Soma, K.K., 2008b. Dehydroepiandrosterone and corti-costerone are regulated by season and acute stress in a wild songbird: jugular ver-sus brachial plasma. Endocrinology 149, 2537–2545.

Oksche, A., Farner, D., Serventy, D., Wolff, F., Nicholls, C., 1963. The hypothalamo-hypophysial neurosecretory system of the zebra finch, Taeniopygia Castanotis. Z.Zellforsch. Mikrosk. Anat. 58, 846–914.

Overk, C.R., Perez, S.E., Ma, C., Taves, M.D., Soma, K.K., Mufson, E.J., 2013. Sex steroidlevels and AD-like pathology in 3xTgAD mice. J. Neuroendocrinol. 25, 131–144.

Pellegrini, A., 2008. The roles of aggressive and affiliative behaviors in resource control:a behavioral ecological perspective. Dev. Rev. 28, 461–487.

Penner, L.A., Dovidio, J.F., Piliavin, J.A., Schroeder, D.A., 2005. Prosocial behavior:multilevel perspectives. Annu. Rev. Psychol. 56, 365–392.

Perfito, N., 2010. The reproductive and stress physiology of Zebra Finches in context:integrating field and laboratory studies. Emu 110, 199.

Perfito, N., Bentley, G.E., Hau, M., 2006. Tonic activation of brain GnRH immunoreactivitydespite reduction of peripheral reproductive parameters in opportunistically breed-ing zebra finches. Brain Behav. Evol. 67, 123–134.

Perfito, N., Zann, R.A., Bentley, G.E., Hau, M., 2007. Opportunism at work: habitat pre-dictability affects reproductive readiness in free-living zebra finches. Funct. Ecol.21, 291–301.

Perfito, N., Kwong, J.M.Y., Bentley, G.E., Hau, M., 2008. Cue hierarchies and testiculardevelopment: Is food a more potent stimulus than day length in an opportunisticbreeder (Taeniopygia g. guttata)? Horm. Behav. 53, 567–572.

Perfito, N., Zann, R., Ubuka, T., Bentley, G.E., Hau, M., 2011. Potential roles for GNIH andGNRH-II in reproductive axis regulation of an opportunistically breeding songbird.Gen. Comp. Endocrinol. 173, 20–26.

Pradhan, D.S., Soma, K.K., 2012. Regulation of 3β-HSD activity in the songbird brain.J. Ornithol. 153, S227–S234.

Pradhan, D.S., Newman, A.E.M., Wacker, D.W., Wingfield, J.C., Schlinger, B.A., Soma, K.K.,2010. Aggressive interactions rapidly increase androgen synthesis in the brain duringthe non-breeding season. Horm. Behav. 57, 381–389.

Remage-Healey, L., Adkins-Regan, E., Romero, L., 2003. Behavioral and adrenocorticalresponses to mate separation and reunion in the zebra finch. Horm. Behav. 43,108–114.

Remage-Healey, L., Maidment, N.T., Schlinger, B.A., 2008. Forebrain steroid levels fluc-tuate rapidly during social interactions. Nat. Neurosci. 11, 1327–1334.

Roberts, M.L., Buchanan, K.L., Bennett, A.T.D., Evans, M.R., 2007. Mate choice in zebrafinches: does corticosterone play a role? Anim. Behav. 74, 921–929.

Robinson, G.E., Grozinger, C.M., Whitfield, C.W., 2005. Sociogenomics: social life in mo-lecular terms. Nat. Rev. Genet. 6, 257–270.

Schlinger, B.A., Arnold, A.P., 1993. Estrogen synthesis in-vivo in the adult zebra finch -additional evidence that circulating estrogens can originate in brain. Endocrinology133, 2610–2616.

Schlinger, B.A., Remage-Healey, L., 2012. Neurosteroidogenesis: insights from studiesof songbirds. J. Neuroendocrinol. 24, 16–21.

Schlinger, B.A., Pradhan, D.S., Soma, K.K., 2008. 3beta-HSD activates DHEA in the song-bird brain. Neurochem. Int. 52, 611–620.

Schmidt, K.L., Pradhan, D.S., Shah, A.H., Charlier, T.D., Chin, E.H., Soma, K.K., 2008.Neurosteroids, immunosteroids, and the Balkanization of endocrinology. Gen.Comp. Endocrinol. 157, 266–274.

Soma, K.K., Wingfield, J.C., 2001. Dehydroepiandrosterone in songbird plasma: seasonal reg-ulation and relationship to territorial aggression. Gen. Comp. Endocrinol. 123, 144–155.

Soma, K.K., Sullivan, K.A., Wingfield, J.C., 1999. Combined aromatase inhibitor andantiandrogen treatment decreases territorial aggression in a wild songbird duringthe nonbreeding season. Gen. Comp. Endocrinol. 115, 442–453.

Soma, K.K., Tramontin, A.D., Wingfield, J.C., 2000. Oestrogen regulates male aggressionin the non-breeding season. Proc. Biol. Sci. 267, 1089–1096.

Soma, K.K., Alday, N.A., Hau, M., Schlinger, B.A., 2004. Dehydroepiandrosterone metab-olism by 3β-hydroxysteroid deydrogenase/Δ5–Δ4 isomerase in adult zebra finchbrain: sex difference and rapid effect of stress. Endocrinology 145, 1668–1677.

Sossinka, R., 1974. Der Einfluss von Durstperioden auf die Schilddruesen-undGonadenaktivitaet und ihre Bedeutung fuer die Brutperiodik des Zebrafinken(Taeniopygia castanotis Gould). J. Ornithol. 115, 128–141.

Srivastava, R., Chaturvedi, C.M., 2010. Effect of estrogen and tamoxifen on the shellgland AVT and VT3R of scotosensitive and scotorefractory Japanese quail. Gen.Comp. Endocrinol. 167, 104–112.

St-Pierre, A., Larose, K., Dubois, F., 2009. Long-term social bonds promote cooperationin the iterated Prisoner's Dilemma. Proc. R. Soc. B Biol. Sci. 276, 4223–4228.

Svec, L.A., Wade, J., 2009. Estradiol induces region-specific inhibition of ZENK but doesnot affect the behavioral preference for tutored song in adult female zebra finches.Behav. Brain Res. 199, 298–306.

Svec, L.A., Licht, K.M., Wade, J., 2009. Pair bonding in the female zebra finch: a potentialrole for the nucleus taeniae. Neuroscience 160, 275–283.

Taves, M.D., Schmidt, K.L., Ruhr, I.M., Kapusta, K., Prior, N.H., Soma, K.K., 2010. Steroidconcentrations in plasma, whole blood and brain: effects of saline perfusion to re-move blood contamination from brain. PLoS One 5, e15727.

Taves, M.D., Ma, C., Heimovics, S.A., Saldanha, C.J., Soma, K.K., 2011.Measurement of steroidconcentrations in brain tissue: methodological considerations. Front. Endocrinol. 2, 39.

Tomaszycki, M.L., Adkins-Regan, E., 2005. Experimental alteration of male song qualityand output affects female mate choice and pair bond formation in zebra finches.Anim. Behav. 70, 785–794.

Tomaszycki, M.L., Adkins-Regan, E., 2006. Is male song quality important in maintainingpair bonds? Behaviour 143, 549–567.

Tomaszycki, M.L., Banerjee, S.B., Adkins-Regan, E., 2006. The role of sex steroids incourtship, pairing and pairing behaviors in the socially monogamous zebra finch.Horm. Behav. 50, 141–147.


Vleck, C., Priedkalns, J., 1985. Reproduction in zebra finches — hormone levels and ef-fect of dehydration. Condor 87, 37–46.

Wade, J., Schlinger, B.A., Hodges, L., Arnold, A.P., 2004. Fadrozole: a potent and specificinhibitor of aromatase in the zebra finch brain. Gen. Comp. Endocrinol. 94, 53–61.

Walters, M.J., Collado, D., Harding, C.F., 1991. Oestrogenic modulation of singing inmale zebra finches: differential effects on directed and undirected songs. Anim.Behav. 42, 445–452.

Washburn, B.E., Morris, D.L., Millspaugh, J.J., Faaborg, J., Schulz, J.H., 2002. Using a com-mercially available radioimmunoassay to quantify corticosterone in avian plasma.Condor 104, 558.

Wingfield, J.C., Lynn, S.E., Soma, K.K., 2001. Avoiding the 'costs' of testosterone: ecolog-ical bases of hormone–behavior interactions. Brain Behav. Evol. 57, 239–251.

Young, K., Liu, Y., Wang, Z., 2008. The neurobiology of social attachment: a comparativeapproach to behavioral, neuroanatomical, and neurochemical studies. Comp.Biochem. Physiol. C Toxicol. Pharmacol. 148, 401–410.

Zann, R., 1994. Reproduction in a zebra finch colony in South-Eastern Australia — thesignificance of monogamy, precocial breeding and multiple broods in a highlymobile species. Emu 94, 285–299.

Zann, R.A., 1996. The Zebra Finch: A Synthesis of Field and Laboratory Studies. OxfordUniversity Press, USA.

effects of water restriction on reproductive physiology and affiliative behavior in an...

Documents