Copyright 
by 
Ryan Ying Wong 
2011 
 
 
  
The Dissertation Committee for Ryan Ying Wong Certifies that this is the approved 
version of the following dissertation: 
 
 
Investigating the female mate preference brain: identifying molecular 
mechanisms underlying variation in mate preference in specific regions 
of a swordtail (Xiphophorus nigrensis) brain  
 
 
 
 
 
Committee: 
 
Molly E. Cummings, Supervisor 
Hans A. Hofmann, Co-Supervisor 
David Crews 
Michael J. Ryan 
Harold H. Zakon 
 
Investigating the female mate preference brain: identifying molecular 
mechanisms underlying variation in mate preference in specific regions 
of a swordtail (Xiphophorus nigrensis) brain 
 
 
 
 
by 
Ryan Ying Wong, B.S. 
 
 
 
Dissertation 
Presented to the Faculty of the Graduate School of  
The University of Texas at Austin 
in Partial Fulfillment  
of the Requirements 
for the Degree of  
 
Doctor of Philosophy 
 
 
The University of Texas at Austin 
May 2011 
Dedication 
 
To my mother, father and brother 
 
 
 v
Acknowledgements 
 
I would like to say thanks to Molly Cummings and Hans Hofmann for their 
invaluable support and guidance as mentors. Thanks to the rest of my committee, David 
Crews, Michael Ryan, and Harold Zakon for their encouragement and keen eyes 
throughout this process. Thanks to Cummings and Hofmann Lab members of past and 
present for critical and constructive discussions. I am especially grateful to Mary 
Ramsey, who taught me nearly all the molecular techniques that made this thesis 
possible. Thanks to all of the undergraduates who have helped in various parts of my 
research. My family and friends have also had immense contributions through emotional 
support and by keeping me grounded and sane. The Mexican government, Brackenridge 
Field Lab and ICMB Core Facilities have all provided generous access to their facilities 
and organisms. Parts of this research would not have happened without the generous 
support from Anthony J. Weido, MD and research fellowships from the Department of 
Ecology, Evolution, Behavior here at UT Austin. Lastly, thanks to the swordtail fish for 
being widely variable in their behaviors. 
 vi
Investigating the female mate preference brain: identifying molecular 
mechanisms underlying variation in mate preference in specific regions 
of a swordtail (Xiphophorus nigrensis) brain 
 
 
Publication No._____________ 
 
 
Ryan Ying Wong, Ph. D. 
The University of Texas at Austin, 2011 
 
Supervisors:  Molly E. Cummings and Hans A. Hofmann 
 
Choosing with whom to mate is one of the most important decisions a female 
makes in her lifetime and inter-individual variation of these preferences can have 
important evolutionary consequences. In order to get a complete understanding of why 
and how females choose a mate, we must identify factors that can contribute to variation 
of female mate choice. Many decades of research sought to understand ultimate 
mechanisms of female mate choice with proximate mechanisms receiving a lot more 
attention in recent years. For my thesis, I identify intrinsic and extrinsic factors that 
correlate with individual variation of female Xiphophorus nigrensis mate preference. I 
provide evidence that a female’s size (e.g. age and sexual experience) as well as male 
behavioral displays can predict female mate preference. Using genes associated with 
female mate preference (neuroserpin, neurologin-3), I identify four brain regions (Dl, 
Dm, HV, POA) that show significant differences in gene expression between females 
 vii
exhibiting high preference for males relative to females displaying little mate preference. 
Neuroserpin and neuroligin-3 gene expression within these brain regions are also 
positively correlated with female mate preference behavior. Two of these brain regions 
(Dm and Dl) integrate multisensory information and are found in the putative teleost 
mesolimbic reward circuitry; the other two regions (HV and POA) are involved in sexual 
behaviors. With the implication of the reward circuitry, I assess whether there are 
changes in dopamine synthesis (via tyrosine hydroxylase, TH) in dopaminergic brain 
regions associated with the degree of mate preference. I do not find evidence of rapid 
changes (within 30 minutes) of TH expression (i.e. dopamine synthesis) in dopaminergic 
brain regions related to variation in female mate preference. Collectively my results 
suggest that mate preference behavior in the brain may be coordinated not just through 
regions associated with sexual response but also through forebrain areas that may 
integrate primary sensory information, with no associated changes of a proxy for 
dopamine synthesis in dopaminergic brain regions.  
 viii
Table of Contents 
List of Tables ...........................................................................................................x 
List of Figures ....................................................................................................... xii 
Introduction ..............................................................................................................1 
Chapter 1: How female size and male displays influence female mate preference 
behavior in the northern swordtail, Xiphophorus nigrensis ............................7 
Abstract ...........................................................................................................7 
Introduction .....................................................................................................7 
Methods.........................................................................................................11 
Female mate choice trials .....................................................................11 
Male behavior ......................................................................................12 
Statistical Analysis ...............................................................................12 
Results: ..........................................................................................................14 
Discussion .....................................................................................................16 
Chapter 2: Neural correlates of female mate preference behavior: a candidate gene 
approach ........................................................................................................30 
Abstract .........................................................................................................30 
Introduction ...................................................................................................30 
Material and Methods ...................................................................................34 
Behavioral Paradigm ............................................................................34 
Estradiol Measurements .......................................................................37 
Tissue Processing .................................................................................37 
Probe Synthesis ....................................................................................38 
in situ hybridization .............................................................................39 
Gene expression quantification ............................................................41 
Digoxigenin quantification .........................................................42 
DIG validation with S-35 riboprobe ...........................................44 
Statistics ...............................................................................................44 
Results ...........................................................................................................45 
 ix
Discussion .....................................................................................................48 
Chapter 3: Characterizing the roles of tyrosine hydroxylase and neuroligin-3 in 
female mate preference behavior in a teleost ................................................67 
Abstract .........................................................................................................67 
Introduction ...................................................................................................68 
Material and Methods ...................................................................................73 
Behavioral Paradigm ............................................................................73 
Estradiol Measurements .......................................................................75 
Cloning tyrosine hydroxylase ..............................................................76 
Tissue Processing .................................................................................77 
Probe Synthesis ....................................................................................77 
in situ hybridization .............................................................................78 
Gene expression quantification ............................................................78 
Digoxigenin quantification .........................................................80 
Statistics ...............................................................................................82 
Results ...........................................................................................................83 
Discussion .....................................................................................................86 
Summary .......................................................................................................93 
References ............................................................................................................107 
 x
List of Tables 
Table 1.1:  Definitions of stereotyped male behaviors .......................................28 
Table 1.2:  Significant effects of male behaviors and preference measures. ......29 
Table 2.1:  Comparisons between in situ hybridization (ISH) quantification 
methods. ............................................................................................60 
Table 2.2:  Neuroserpin optical density (mean ± SE) comparisons between “high” 
(> median) and “low” (< median) preference score. ........................61 
Table 2.3:  Neuroserpin optical density (mean ± SE) comparisons between “high” 
(> median) and “low” (< median) behaviors. ...................................62 
Table 2.4:  Correlations between preference score and gene expression in Dm, Dl, 
POA in non-sexual contexts. .............................................................63 
Table 2.5:  Correlations between glides, transits and gene expression in Dm, Dl, 
POA in male exposed environments. ................................................64 
Table 2.6:  Correlation between a proxy for circulating estradiol levels and 
preference score, glides, transits, and gene expression in different brain 
regions in male exposed environments. ............................................65 
Table 2.7:  Comparison of degree centrality between candidate nuclei (Dm, Dl, 
POA) and other nuclei in each treatment group for females used to 
localize neuroserpin. .........................................................................66 
Table 3.1:  Tyrosine hydroxylase optical density (mean ± SE) comparisons between 
“high” (> median) and “low” (< median) behavioral patterns. .......102 
Table 3.2:  Correlation between a proxy for circulating estradiol levels and 
association bias, glides, transits, and gene (neuroligin-3 and tyrosine 
hydroxylase) expression in different brain regions. ........................103 
 xi
Table 3.3:  Neuroligin-3 optical density (mean ± SE) comparisons between “high” 
(> median) and “low” (< median) association bias. ........................104 
Table 3.4:  Neuroligin-3 optical density (mean ± SE) comparisons between “high” 
(> median) and “low” (< median) transits. .....................................105 
Table 3.5:  Neuroligin-3 optical density (mean ± SE) comparisons between “high” 
(> median) and “low” (< median) glides. .......................................106 
 xii
List of Figures 
Figure 1.1:  Relationship between female size and (a) strength of preference or (b) 
preference score toward large males. ................................................22 
Figure 1.2:  Correlations between female standard length and female behavioral 
displays. ............................................................................................23 
Figure 1.3:  Significant correlation (n = 122, r = 0.68, p = 7.77 * 10-18) between 
preference measures toward large males (strength of preference and 
preference score). ..............................................................................24 
Figure 1.4:  Time spent with large and small size class males for (A) sexually 
experienced and (B) virgin females. .................................................25 
Figure 1.5:  Box and whisker plots of the frequency of male behaviors. .............26 
Figure 1.6:  Significant main and interaction effects between female strength of 
preference and male behaviors. .........................................................27 
Figure 2.1:  Preference measures by stimuli and social group. ............................54 
Figure 2.2:  Correlation between neuroserpin expression in situ hybridization (ISH) 
quantification methods on adjacent series. .......................................55 
Figure 2.3:  Neuroserpin expression in females exposed to a mate preference context 
(LL, LS, and SS). ..............................................................................56 
Figure 2.4:  Individual variation of preference score and neuroserpin expression.57 
Figure 2.5:  Correlation between neuroserpin expression in Dm and preference score 
using S35 labeled riboprobes. ...........................................................58 
Figure 2.6:  Neuroserpin expression network by context. ....................................59 
Figure 3.1:  Distribution of tyrosine hydroxylase across all female X. nigrensis 
tested. ................................................................................................94 
 xiii
Figure 3.2:  Distribution of tyrosine hydroxylase in brains of female X. nigrensis.95 
Figure 3.3:  Tyrosine hydroxylase expression in select brain regions across treatment 
groups. ...............................................................................................96 
Figure 3.4:  Tyrosine Hydroxylase expression in females exposed to a large vs. small 
male (LS) context. .............................................................................97 
Figure 3.5:  Neuroligin-3 expression in females exposed to a large vs. small male 
(LS) context. .....................................................................................98 
Figure 3.6:  Individual variation of association bias and neuroligin-3 expression in 
LS exposed females. .........................................................................99 
Figure 3.7:  Neuroligin-3 expression network by context. .................................100 
Figure 3.8:  Tyrosine hydroxylase expression network by context. ...................101 
  
 1
Introduction 
 Animal behavioral displays are a product of the integration of both intrinsic and 
extrinsic mechanisms. To help answer the question of “Why (and how) do animals 
behave the way they do?”, Nikolaas Tinbergen proposed a framework [Tinbergen 1963] 
that consisted of four non-mutually exclusive levels of analyses to study animal behavior: 
causation, ontogeny, survival value, and evolution. A typical animal can encounter a 
diverse array of challenges and opportunities during its life [O'Connell and Hofmann 
2010]. From birth, the ecological and social environment it was raised in may influence 
how the animal behaves as it ages (ontogeny). In order to survive and pass along its genes 
(survival value), the animal must obtain and defend resources and mates. Appropriate 
behavioral responses for reproduction may have been selected over time and may have 
diverged in form or function in closely related species (evolution). All behavioral 
displays can be influenced by a number of different physiological mechanisms such as 
gene expression (see [Wong and Hofmann 2010] for review of genomic responses in 
natural behaviors), neural activity, and hormones (causation). Only by understanding the 
contributions of both proximate (causation, ontogeny) and ultimate (survival value, 
evolution) mechanisms to an animal’s behavior, can we begin to get a complete 
understanding of why and how an animal behaves the way it does. 
One critical behavior for an animal is the process of finding and evaluating 
prospective mates. Why and how a female chooses to mate with one male over another 
has been a central question in sexual selection research [Andersson 1994; Darwin 1871]. 
 2
In species with mating rituals, females must both perceive and evaluate male cues prior to 
reaching a mating decision. The process and the outcome of mate choice can be 
influenced by factors within an individual’s lifetime as well as the species’ evolutionary 
history. Furthermore the perception, evaluation and mate choice decision rules can differ 
between females of the same species. Variation in female preference functions and/or 
choosiness can facilitate our understanding of (i) rate and direction of evolution by sexual 
selection, (ii) evolutionary history of female choice and its benefits, (iii) the origin or 
maintenance of secondary sexual characteristics, and (iv) environmental, social, and 
proximate mechanisms underlie female mate choice [Jennions and Petrie 1997]. Many 
studies have documented the adaptive value of female preferences and others have 
proposed hypotheses for the evolution of female mate choice in multiple taxa [Andersson 
1994; Jones and Ratterman 2009]. Studies have also documented how ecological and 
social environment prior to sexual maturation can also influence mate choice [Westneat 
et al. 2000]. The role of causal mechanisms of female mate preference, however, has 
historically not been as well studied. Of the major causal mechanisms (genes, hormones, 
and brain), studies have shown organizational and activational effects of hormones on 
female preference functions and choosiness [Adkins-Regan 2011; Wilczynski and Lynch 
2010]. The neural mechanisms of female mate preference in vertebrates have identified 
auditory processing and hypothalamic nuclei as areas showing increased neural activity 
(via expression of an immediate early gene) in sexual contexts [Gentner et al. 2001; Hoke 
et al. 2004; Hoke et al. 2005; Maney et al. 2003; Sockman et al. 2002; Woolley and 
Doupe 2008]. Only recently has a study expanded the search for brain regions underlying 
 3
female mate assessment to areas outside of primary sensory processing such as those 
found in the Social Behavior Network [Hoke et al. 2005; Desjardins et al. 2010]. 
The molecular mechanisms of female mate preference in vertebrates are not well 
understood. To date, only one study has explored the genomic response to sexual and 
social environments in the brains of females [Cummings et al. 2008]. Using a 
combination of microarray and quantitative real time PCR analyses, we have identified 
genes (neuroserpin, neuroligin-3, and others) that show context specific differential 
whole brain expression patterns and whose expressions are significantly correlated with 
individual variation of female Xiphophorus nigrensis preference. These candidate genes 
associated with mate preference present an opportunity to identify neural and molecular 
mechanisms potentially underlying female mate preference in a more specific manner. 
Given the general lack of knowledge of the neural and molecular mechanisms underlying 
female mate preference, the central theme of my thesis is to identify intrinsic as well as 
extrinsic mechanisms underlying variation in female mate preference in the northern 
swordtail, Xiphophorus nigrensis. 
Some species are better suited to study the proximate mechanisms of female mate 
preferences than others. Characteristics of a good system include indicators of receptivity 
and preference in the lab that reflect natural behaviors in the wild, as well as the ability to 
evaluate female preferences without the confounding physical interactions with males. 
One such system is the Poecillid fishes (guppies, swordtails, mollies, and platyfish). 
Poecillid fishes are a useful system for exploring sexual selection by female choice 
 4
[Basolo 1990, 1995; Houde 1997; Houde and Endler 1990; Ryan and Rosenthal 2001]. In 
my focal swordtail species, Xiphophorus nigrensis, the males mature into genetically 
determined sizes related to the copy number of melanocortin receptor 4 B alleles whereas 
females have indeterminate growth and lack melanocortin receptor 4 B alleles [Kallman 
1984, 1989; Lampert et al. 2010]. The two larger size classes of males court females 
while the smallest size class exhibits force copulation [Ryan and Causey 1989]. Most 
large and intermediate sized males possess sexually dimorphic traits such as large body 
size, courting behavior, and ultraviolet ornamentation that females prefer to varying 
degrees [Cummings et al. 2003; Ryan et al. 1990; Ryan and Rosenthal 2001]. Females 
also exhibit a putative receptive display (glide displays), which are generally exhibited 
prior to copulation events in free-ranging experiments [Cummings and Mollaghan 2006]. 
These displays are not unique to my focal species, but similar receptive behaviors have 
been described across poeciliids [Liley 1965]. It has been found that visual stimuli are 
sufficient to elicit female preferences in this system [Crapon de Caprona and Ryan 1990; 
Ryan et al. 1990; Ryan and Wagner 1987]. Specifically, the preference for large body 
size seen under laboratory settings is consistent with greater reproductive success of 
larger males in the wild as assessed by paternity analysis of young from wild caught 
females [Ryan et al. 1990; Ryan et al. 1992]. Of particular importance is the apparent 
decoupling of reproductive state from preferences because it allows us to investigate the 
neural and molecular mechanisms without influences of reproductive state. We have 
recently documented that a female’s preference is not predicted by her reproductive cycle 
status [Ramsey et al. 2011]. Furthermore, we identified significant patterns between 
 5
female preference behavior and gene expression in the brain [Cummings et al. 2008]. In 
sum, swordtails are a powerful system to isolate the neural and molecular mechanisms of 
preference behavior because they (1) do not require physical contact to elicit preference, 
(2) their lab behaviors mirror those observed in the field, (3) preference behavior appears 
decoupled from reproductive state, and (4) our initial research has identified candidate 
genes at the whole brain level that underlie individual variation in behavior. 
My thesis characterizes additional extrinsic and intrinsic factors that underlie 
variation in female mate preference in X. nigrensis. In Chapter 1, I look at some external 
mechanisms and demonstrate that a female’s standard length and male behavioral 
displays can predict female preference. Sexual experience may also influence variation in 
female preference. In Chapters 2 and 3, I focus on identifying neural mechanisms that 
may underlie female mate preference behavior. Using a context-specific marker 
associated with mate assessment, neuroserpin, in addition to an immediate early gene 
(egr-1), I quantify and localize the expression of these genes for Chapter 2 to describe a 
putative female mate preference network. I test the hypothesis that the neural circuitry 
associated with female mate preference behavior is constrained to brain regions 
governing sexual response (nodes of the Social Behavior Network, SBN [Goodson 2005; 
Newman 1999]) as well as an alternative hypothesis that female mate preference behavior 
is associated with brain regions distinct from those associated with sex. Using 
neuroserpin I show that some brain regions (Dm, Dl) associated with mate preference are 
also involved in multisensory processing and possibly the homologous mesolimbic 
reward pathway. Others (HV, POA) are associated with sexual behaviors (SBN or 
 6
hypothalamic brain region). In Chapter 3, I assess if changes in dopamine production (via 
tyrosine hydroxylase expression) is related to female mate preference behavior in 
addition to characterizing the expression of a different gene associated with mate 
preference, neuroligin-3. I do not have evidence showing rapid changes in dopamine 
synthesis related to female mate preference, but I do have evidence that the brain exhibits 
molecular “signatures” of mate preference looking at neuroligin-3 expression. Results of 
the neuroligin-3 expression patterns were consistent with neuroserpin expression patterns 
shown in Chapter 2. I show that neuroligin-3 expression is differentially expressed by 
context and related to individual variation in female mate preference in the same brain 
regions (Dm, Dl, POA, HV) identified in Chapter 2. In addition to supporting a putative 
network of brain regions associated with mate preference, I show that the density of the 
neuroligin-3 expression network varies with the presumed complexity of the mate 
preference environment, which I also observed with the neuroserpin expression network 
in Chapter 2. 
Evolutionary consequences of sexual selection can be driven in part by variation 
in female mate preferences [Jennions and Petrie 1997]. This thesis and other studies 
where I have contributed, which are not presented here [Cummings et al. 2008; Ramsey 
et al. 2011], represent the very first steps in identifying molecular, neural, and hormonal 
mechanisms underlying variation in female mate preference. This foundational work will 
hopefully guide other studies to an insightful understanding of what is going on in the 
brains of female X. nigrensis as they choose a mate. 
 7
Chapter 1: How female size and male displays influence female mate 
preference behavior in the northern swordtail, Xiphophorus nigrensis 
  
ABSTRACT 
Building a complete model of animal behavior requires knowing both the internal 
and external factors that influence behavior. As recent genomic and neural studies begin 
to establish the northern swordtail, Xiphophorus nigrensis, as a model for proximate 
explorations of female mate preference, it is important to recognize the complexity of 
other factors that can drive variation in female preference behavior. In this study we 
determine how different experiential and social factors correlate with inter-individual 
variation in female mate preference: female body size, sexual experience, and male 
behavioral displays. We show that variation in female mate preferences is influenced by 
her size (a proxy for age) and specific male behavioral displays. We also demonstrate that 
copulating at least once may be sufficient to increase preference for large males. Our 
identification of these relationships emphasizes the multiple and diverse types of 
mechanisms that can influence female mate preference variation and encourage their 
consideration in future studies when trying to assess why and how females choose males. 
INTRODUCTION 
In order to respond appropriately to a variety of social and environmental 
contexts, an animal must amalgamate intrinsic and extrinsic information. How this 
information is processed can be influenced by the animal’s developmental conditions and 
evolutionary history of the species. Such mechanisms of behavior (e.g. causation, 
 8
ontogeny, adaptive value, and evolution) all integrate to explain how and why an animal 
behaves in specific contexts [Tinbergen 1963]. Thus, to completely understand an 
animal’s behavioral display, we must consider both ultimate (e.g. adaptive value and 
evolution) and proximate mechanisms (e.g. causation and ontogeny). Female mate 
choice, one of the key behaviors in sexual selection studies, has been studied at each of 
the four levels of analyses across a number of different taxa [Andersson 1994], however, 
integrating across proximate and ultimate mechanisms has proven challenging. 
The northern swordtail, Xiphophorus nigrensis, represents one species where 
studies can tractably examine both the proximate and ultimate mechanisms of female 
mate choice. Studies from our lab have identified molecular, neural, and hormonal 
mechanisms correlated with individual variation of female mate preference [Cummings et 
al. 2008; Ramsey et al. 2011]. Furthermore, a rich history of experimental work has 
dissected how specific male attributes such as size [Cummings and Mollaghan 2006; 
Ryan et al. 1990; Ryan and Rosenthal 2001; Ryan and Wagner 1987], sword [Rosenthal 
et al. 2002], and UV ornamentation [Cummings et al. 2003; Cummings et al. 2006] 
influence female preference response. In order to get a complete understanding of the 
processes driving female mate choice within this species, however, we must extend our 
explorations into how the experiential state of the female as well as how the more 
dynamic traits (e.g. behavior) of the male influence a female’s response. In this study we 
explore if an X. nigrensis female’s body size (a proxy for age), sexual experience (virgins 
 9
versus non-virgins), and male behavioral displays influence variation in her mate 
preference for large males. 
The role of experience and social influences on female mate choice is common 
across taxa [Westneat et al. 2000]. As a female sexually matures, her rearing environment 
can alter her preference function and choosiness of future mates [Adkins-Regan 2011; 
Holveck and Riebel 2010; Rutledge et al. 2010; Walling et al. 2008; Woodgate et al. 
2010]. After reaching sexual maturity, however, female mate preferences may continue to 
be plastic as they gain sexual experience [Coleman et al. 2004; Kodric-Brown and 
Nicoletto 2001b; Morris et al. 2006; Rebar et al. 2011; Tudor and Morris 2009]. Some 
females make mate choice decisions by utilizing social information from other females 
by mate-choice copying [Dugatkin 1996; Godin et al. 2005; Schlupp et al. 1994; Servedio 
et al. 2009; Westneat et al. 2000] or by observing sexual displays [Andersson 1994; Elias 
et al. 2006; Kodric-Brown and Nicoletto 2001a; Patricelli et al. 2002]. As such, sexual 
and social experiences represent potentially important sources of variation of female mate 
preference and knowing their contribution (if any) will give us a more comprehensive 
view of mate choice mechanisms and their evolutionary consequences for a particular 
species.  
In this study we determine whether female X. nigrensis preference response is 
predictable based on her size, sexual experience and/or specific male behavioral displays. 
Poeciliid fishes (guppies, mollies, platyfish and swordtails) are a popular taxon to 
examine the evolution of female mate choice (reviewed in [Houde 1997; Ryan and 
 10
Rosenthal 2001]. Although poeciliid (e.g. X. nigrensis) females will show, on average, a 
preference towards particular male phenotypes under certain contexts, there is ample 
individual variation amongst females, which facilitates identifying underlying 
mechanisms [Brooks 2002; Brooks and Endler 2001; Cummings and Mollaghan 2006; 
Morris et al. 2010; Morris et al. 2006]. While female X. nigrensis have indeterminate 
growth [Kallman 1989; Morris and Ryan 1990],  males in this species mature into 
genetically influenced size classes where the copy number of the melanocortin receptor 4 
(mc4r) B alleles positively correlates with the terminal size of males [Kallman 1989; 
Lampert et al. 2010]. The larger classes of males typically court females while the 
smallest size class exhibits force copulation [Ryan and Causey 1989].  X. nigrensis 
females generally prefer an assortment of secondary sexual characteristics [Cummings et 
al. 2003; Ryan et al. 1990; Ryan and Rosenthal 2001; Ryan and Wagner 1987] and a 
female’s preference for larger males does not significantly vary across her reproductive 
cycle [Ramsey et al. 2011]. Variation in gene expression in the whole brain, specific 
brain regions, and circulating estradiol levels are significantly correlated with female 
mate preference [Cummings et al. 2008; Ramsey et al. 2011; Wong et al. in review]. 
Understanding how experiential (i.e. female size and sexual experience) and social 
factors (e.g. male behaviors) contribute to variation in female mate preferences will allow 
us to build a more complete model for female mate choice behavior.  
 11
METHODS  
Female mate choice trials 
In total, we observed preference behavior from 176 sexually mature X. nigrensis 
females. Some females were lab reared in sexual isolation (n = 13) and others were either 
wild caught from Nacimento de Rio Choy in the state of San Luis Potosi, Mexico or 
progeny of wild caught females (n = 163). We followed established methodology for 
measuring female preference in a dichotomous choice paradigm [Cummings et al. 2008]. 
From September 2005 – 2010, we subjected females to 30 minutes in an experimental 
tank with a large vs. small size class male stimuli behind UV transparent barriers. The 
center portion of the tank was subdivided into three 24 cm regions: a middle ‘neutral’ 
zone with an “association” zone adjacent to each stimulus. Under lighting conditions 
found in the wild [Cummings et al. 2003], females were initially placed into the central 
region of the tank and allowed 5 minutes for acclimation within an opaque cylinder 
before starting the trial . We switched the sides of the stimuli halfway through each trial 
to avoid side bias. We measured the amount of time in the association zones for all 
females, along with glide displays (a proxy for receptivity, n = 172, [Cummings and 
Mollaghan 2006; Liley 1965]) and the number of transits into the neutral zone (a proxy 
for general locomotor activity, n = 132) for a subset of these females. We measured 
preference for the large or the small male by three measures (i) association bias, (ii) the 
strength of preference [Morris et al. 2010], and (iii) a preference score [Cummings et al. 
2008]. Association bias is defined as the time spent with each (large or small) divided by 
the total association time (sum of time spent in both association zones). Strength of 
 12
preference is calculated as the absolute difference in association time between males (e.g. 
Strength of preferencelarge male = Timelarge – Timesmall). And preference score is calculated 
as in Cummings et al. 2008, but now calculated according to each male stimulus (e.g. 
preference scorelarge male = association biaslarge male + log((1 + glides displayed toward large 
male)/ total number of transits ). Prior to behavior trials, we measured the standard length 
(the length between the tip of the snout to the end of caudal peduncle) of the majority of 
females (n = 166). 
Male behavior  
For a subset of sexually experienced females (n = 75), a single observer (P.S.) 
identified and quantified stereotyped male behavior patterns from video (see Table 1.1 
for definitions). For previously described male behaviors, we followed original 
definitions for quantification [Cummings and Gelineau-Kattner 2009; Ryan and Causey 
1989]. To optimize video resolution, we only video-taped and quantified male stimuli 
behaviors when the female was in association zone adjacent to that male. Due to a limited 
number of males, some large (n = 14) and small (n = 13) males were reused for multiple 
trials.    
Statistical Analysis  
All statistics were calculated using SPSS (ver. 18). We used a t-test to assess 
differences in time spent in the association zones between the male stimuli. We used a 
Shapiro-Wilk to test for normality. The female standard length, association biaslarge male, 
glides and transits were not normally distributed even after a log transformation or a 
arcsine(square root of association bias). Hence, we used Spearman’s rank correlations to 
 13
assess relationships between female standard length and preference score, strength of 
preference, association bias, glides, and transits. To assess the effect of sexual experience 
on the strength of preference we used a general linear model (GLM) with female standard 
length as a covariate. To account for multiple hypotheses testing, we only report 
significant values after a Benjamini-Hochberg correction [Benjamini et al. 2001].  
To analyze the effect of male behaviors on female preference, we used a 
generalized estimating equation (GEE) and controlled for male size-class and female 
standard length.  We used a GEE for repeated measures of the female’s preference (a 
measure of preference was calculated for both large and small male, see description 
above). Although all behavioral displays with the exception of foraging were exhibited 
by at least two different males in each size class, any behavioral displays exhibited by 
less than four males in each size class were considered rare across males and were not 
statistically analyzed (parallel swimming, sigmoid display, sporadic movement, foraging, 
inactivity pause). Due to low power, we were not able to reliably account for effects of 
individual males in our models. For our two main measures of preference (strength of 
preference and preference score), we ran two GEE models with male size-class as a 
cofactor and female standard length as a covariate that consisted of (a) the measure of 
preference and individual male behaviors or (b) the measure of preference and total 
activity of the male (sum of all behavioral displays). Significant interaction effects are 
relative to the small male class.  To compare number of individual behavioral displays 
between large and small males we averaged the number of behavioral displays for each 
male and then assessed the difference in number of display using a Mann-Whitney U test. 
 14
RESULTS:  
Female mate preference and female size: There were highly significant positive 
relationships between female size (standard length) and the three measures of preference 
towards the large male class stimulus: the strength of preferencelarge male (n = 168, ρ = 
0.44, p = 2.3 * 10-9, Figure 1.1A), preference scorelarge male (n = 123, ρ = 0.364, p = 3.5 * 
10-5, Figure 1.1B), and association biaslarge male (n = 168, ρ = 0.454, p = 1.4 * 10-10). There 
were no significant correlations between female size and total glides (n = 164, ρ = -0.027, 
p = 0.728, Figure 1.2C) or transits (n = 123, ρ = -0.128, p = 0.157, Figure 1.2D). There 
was also no significant correlation between female size and total association time (n = 
168, ρ = -0.105, p = 0.177). Strength of preferencelarge male and preference scorelarge male 
were significantly and positively correlated with each other (n = 122, r = 0.68, p = 7.77 * 
10-18, Figure 1.3). Preference scorelarge male was also significantly and positively correlated 
with the total number of glides (n = 131, ρ = 0.45, p = 6.8 * 10-8) and negatively 
correlated with the number of transits (n = 131, ρ = -0.233, p = 0.007) whereas strength 
of preferencelarge male was only positively correlated with total glides (n = 172, ρ = 0.19, p 
= 0.01). 
Female mate preference and sexual experience: As expected, sexually experienced 
females spent significantly more time next to the large male relative to the small male (n 
= 163, t = 14.34, p = 1.79 * 10-36, Figure 1.4A). Virgins, however, showed no significant 
difference in time spent between the males (n = 13, t = 1.06, p = 0.29, Figure 1.4B). 
Virgins had a significantly lower strength of preference than non-virgins (mean strength 
of preference ± SE: virgins, 170.6 ± 168.8, non-virgins, 518 ± 44.9; t = -2.08, p = 0.038). 
 15
Virgins were significantly smaller than sexually experienced females (t = 1.99 p = 0.047). 
After controlling for female size, the strength of preference for the large male between 
sexually experienced and virgins was not significant (GLM, F = 1.606, p = 0.2).   
Female mate preference and male behavior: Of all the behaviors, large males performed 
lateral displays (Z = -3.328, p = 0.0004) and circles swims (Z = -2.132, p = 0.33) 
significantly more than small males but circle swims do not pass a multiple hypothesis 
correction (Figure 1.5). As expected, there was a significant main effect on female’s 
strength of preference and preference score between large and small males (strength of 
preference, χ2 = 96.426, p < 10-19; preference score, χ2 = 49.072, p = 2.4 *10-12). Female 
size, however showed no significant main effect on her strength of preference (χ2 = 
0.568, p = 0.451) or preference score (χ2 = 3.293, p = 0.07). One male behavior with a 
significant main effect that occurs with both measures of preference was up and downs 
(Table 1.2). There was also a significant interaction effect between male size class and up 
and downs for strength of preference (Table 1.2, Figure 1.6C). Circle swims and solo 
swims had significant main effects with the female strength of preference (Table 1.2, 
Figure 1.6A, B). There were significant interaction effects of male size class between 
elongated male swimming with preference score and circle swim, bobbing, and solo 
swims with strength of preference (Table 1.2). There was no significant main (χ2 = 0.475, 
p = 0.491) or interaction effects (χ2 = 0.035, p = 0.851) between male activity and 
preference score; or sexual displays and preference score (main effect: χ2 = 0.245, p = 
0.621; interaction effect: χ2 = 0.783, p = 0.376). There were, however, significant main 
(χ2 = 9.942, p = 0.001) and interaction (χ2 = 4.801, p = 0.028) effects between male 
 16
activity with strength of preference (Figure 1.6d) but not between sexual displays and 
strength of preference (main effect: χ2 = 1.117, p = 0.291; interaction effect: χ2 = 0.592, p 
= 0.442).  
DISCUSSION 
Female experience and male behavioral displays influence female response 
towards males in X. nigrensis. In this species, females of larger size (and presumably 
greater age and experience) show stronger preference for larger males (Figure 1.1). Male 
behaviors, some of which may indicate a male’s motivation or vigor in these 
experimental conditions (e.g. up-down movements and total activity), can also strongly 
predict female response. These results suggest that preference behavior is not hard-wired 
in this species but rather is molded over a female’s lifetime and can be modulated by 
specific male behaviors.   
 A female swordtail’s size can be taken as a proxy of her age [Morris et al. 2006; 
Morris and Ryan 1990] and may represent the amount of social and sexual experience. In 
this study, the size of a female X. nigrensis significantly predicted both her strength of 
preference and preference score for large males. This is consistent with a sister species, 
X. multilineatus [Morris et al. 2010; Rios-Cardenas et al. 2007]. Although many 
interpretations of the observation are possible, we will focus on two with support from 
data in this study: size-assortative preferences and age/sexual experience. Assortative 
mating preferences are widely documented across taxa where females can prefer to mate 
with males of similar size/body shape or coloration [Elmer et al. 2009; McKinnon et al. 
2004; Nagel and Schluter 1998; Salzburger et al. 2006; Schluter 2009]. In our study there 
 17
is a significant positive relationship between female size and the proportion of time she 
spends with a large male (association biaslarge male) while no significant correlation 
between female size and total association time (Figure 1.2A & B). This suggests that 
motivation does not vary with age, but discrimination between male size classes does. 
Specifically, larger females tend to spend more time associating with large males while 
smaller females exhibit little discrimination. While small males gain some reproductive 
advantage by sexually maturing earlier than large males [Morris and Ryan 1992], if 
smaller females do not actively discriminate against them this may suggest that small 
males may gain much of their mating success with smaller females. The possibility that 
female size-based variation in tolerance of the force-copulating morphotype may 
contribute to the maintenance of alternative mating strategies in this system, and may also 
play a similar role in its sister taxa, X. multilineatus [Morris et al. 2010].  
The role of age and sexual and social experience on female preference functions 
or choosiness is documented across a wide range of taxa from crickets [Rebar et al. 
2011], to guppies [Kodric-Brown and Nicoletto 2001b], to birds [Coleman et al. 2004]. 
Some mechanisms that allow females to modify or refine preferences through time are 
physiological changes associated with copulation, learning and remembering which 
males are best (e.g. male-male competition), or copying choices of other females [Bailey 
and Zuk 2008; Bailey and Zuk 2009; Kodric-Brown and Nicoletto 2001b; Rosenqvist and 
Houde 1997; Schlupp et al. 1994; Servedio et al. 2009; Tudor and Morris 2009; Westneat 
et al. 2000; Wong and Candolin 2005]. In this study, smaller (i.e. presumably younger 
and less experienced) females do not display a strong preference towards large males, 
 18
whereas larger (i.e. presumably older and more experienced) do (Figure 1.1). In X. 
nigrensis, females may refine their preference for larger male with age through learning 
and memory of male quality or copulation. Whether this learning is based on aspects of 
reward via mating events with large class males, or whether it is driven by aversive 
learning from negative interactions with small, harassing males is not known. 
Sexual and social experience in shaping female preferences is common across 
poeciliids [Breden et al. 1995; Kodric-Brown and Nicoletto 2001b; Marler et al. 1997; 
Tudor and Morris 2009; Walling et al. 2008; Walling et al. 2007]. In the congener, X. 
malinche, copulating at least once appears to be sufficient to alter female preference for 
vertical bar symmetry [Morris et al. 2006; Tudor and Morris 2009]. In X. nigrensis a 
similar scenario may occur as there were no significant differences in time spent with a 
large or small male for virgin females but there is a clear preference for large males in 
sexually experienced females (Figure 1.4A & B). Although this difference may be due to 
virgins being smaller in size, with a small number of virgins tested in this study we had 
low power to detect a difference. It is interesting to consider whether male behavior, and 
perhaps male preference for larger females, influenced the apparent low preference for 
large males in smaller females. The body size of a poecillid female is highly correlated 
with her fecundity [Herdman et al. 2004; Ojanguren and Magurran 2004] and in many 
species, including other poecillids, males prefer to mate with larger and more fecund 
females [Andersson 1994; Herdman et al. 2004; Liao and Lu 2009; Plath et al. 2006]. It is 
conceivable that large males displayed differentially to larger females more so than 
smaller females in this study. Our statistical analysis do not allow us to evaluate this 
 19
question, however, a more direct examination of how male behavior differs by female 
size in this species would be enlightening.   
 Behavioral displays by the male (e.g. courtship) influencing female preferences 
are widely documented [Andersson 1994; Elias et al. 2006; Kodric-Brown and Nicoletto 
2001a; Patricelli et al. 2002]. The interaction between behavioral displays and physical 
attributes (e.g. size of ornamentation or coloration) can lead to different levels of 
choosiness depending on the combination of traits [Cummings et al. 2006; Kodric-Brown 
and Nicoletto 2001a; Patricelli et al. 2002]. For example, in guppies there appears to be a 
synergistic effect of male courtship display rates and coloration on female preference  
[Kodric-Brown and Nicoletto 2001a]. Similarly there is an interaction effect between 
swordtail male UV ornamentation and male behavioral activity on female preference 
where females prefer large, UV ornamented males with lower total behavioral activity 
levels relative to non-UV ornamented males with higher total activity levels [Cummings 
et al. 2006]. In the current study, we also find a significant effect of total male activity 
(sum of all behavioral displays in Figure 1.6d) and strength of preference. Females tend 
to spend more time associating with males (e.g. higher strength of preference) that are 
more active overall (Figure 1.6D). It is interesting to note that small males benefit more 
from increased activity than large males who garner much higher female preference 
responses than small males even in the absence of activity (Figure 1.6D).  
While a male’s total activity is positively related to the strength of preference, 
only one behavior displayed by the males, up and downs, significantly explained 
variation in female strength of preference and preference score (Table 1.2, Figure 1.6C). 
 20
Up and downs could be interpreted as motivation for the male to gain access to the 
female by swimming vigorously against the barrier causing an up and down motion. The 
attractiveness of this behavior to females could stem from it being a direct indicator of a 
male’s vigor and/or indicate a male’s level of interest in the female.  It is interesting to 
note that the most frequently displayed male behavior (lateral display, Figure 1.5) had no 
significant effect on female preference.  This suggests that females are more responsive 
to male displays that are more vigorous (e.g. up and down; circle swim, total activity, 
Figure 1.6) than passive displays that allow for viewing of ornamentation with little 
physical activity. It was surprising that there was not a significant main effect of sexual 
displays on female preference. However, given that full expression of a male swordtail 
sexual display requires close contact with females, we cannot rule out that some of our 
results are artifacts of a lab testing environment. Studying free-ranging behavioral 
interactions between the sexes and resulting chances of copulation either in the lab (e.g. 
[Cummings and Mollaghan 2006; Ryan and Causey 1989] or in the wild will help 
elucidate biological relevance.  
When trying to understand proximate mechanisms (e.g. sensory, neural, 
molecular, hormonal) of female mate preference, it is important to consider that variation 
of these internal mechanisms can be due to a variety of extrinsic influences. For example, 
while studies in female birds have identified brain regions showing differential neural 
activity in response to attractive versus less attractive songs [Gentner et al. 2001], her 
recent acoustic experiences to particular male songs can also influence neural activity 
patterns [Sockman et al. 2002, 2005]. Other studies in songbirds have also shown that 
 21
neural activity and genomic responses can habituate to repeated songs [Dong et al. 2009; 
Mello et al. 1995] and constant changes in song features may serve to keep females 
engaged [Searcy 1992].  An X. nigrensis female’s size, sexual and social (e.g. male 
behavioral displays) experiences are constantly changing throughout her life. As these 
factors have an effect on female mate preference behavior, it is likely they influence 
underlying proximate mechanisms. Future studies investigating intrinsic mechanisms of 
female mate preference should consider female ages, sexual, and social experience as 
sources of variation and factors that should be controlled. 
There can be many factors that contribute to individual variation in female mate 
choice. In poeciliids, ecological and social environments, female condition, age, 
experience, male physical and behavioral phenotypes, gene expression and hormones are 
just some of the factors that may affect female mate choice. In this study we provided 
evidence that female size and some male behaviors correlate with female preference. 
These factors can allow us to build a more complete model of female mate choice and its 
effects on male behavioral or phenotypic trait evolution in X. nigrensis. In this species, 
females may vary their preferences in response to extrinsic factors such as male 
secondary sexual characteristics (e.g. body size, UV ornamentation), male behavioral 
displays, and intrinsic factors like size (e.g. age/experience), and physiological cues (e.g. 
regulation of gene expression and hormones). It is likely that combinations of these 
mechanisms help to shape differences in choosiness across females and even within 
females across time both in this species and others. 
 
 22
Figure 1.1: Relationship between female size and (a) strength of preference or (b) 
preference score toward large males.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 23
Figure 1.2: Correlations between female standard length and female behavioral 
displays.   
There is a significant correlation between female size and proportion of time spent with 
large male (A). No significant correlations with female size and (B) total association 
time, (C) total number of glides or (D) total number of transits. 
 
 
 
 
 
 
 
 
 
 
 
 
 24
Figure 1.3: Significant correlation (n = 122, r = 0.68, p = 7.77 * 10-18) between 
preference measures toward large males (strength of preference and 
preference score). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 25
Figure 1.4: Time spent with large and small size class males for (A) sexually 
experienced and (B) virgin females.  
***, p < 0.001 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 26
Figure 1.5: Box and whisker plots of the frequency of male behaviors.  
Blue and red boxes represent large and small males, respectively. Line represents the 
median and whiskers represent maximum and minimum values. ***, p < 0.001; *, p < 
0.05 
 
 
 
 
 
 
 
 
 
 27
Figure 1.6: Significant main and interaction effects between female strength of 
preference and male behaviors.  
Male behavioral display are A) Circle Swim, B) Solo Swim, C) Up and Down, and D) 
Total Activity. Diamonds represent displays by large males and X represents displays by 
small males. 
 
 
 
 
 
 
 
 
 
 
 
 28
Table 1.1: Definitions of stereotyped male behaviors 
Male Behavior Description 
Backward 
Float 
The male swims backward without changing orientation for a half a 
body length. 
Bobbing 
The male moves his head quickly in a side to side fashion with his snout 
against/near glass. The male head movement is slight, but in rapid 
succession.    One bobbing event is a continuous event. Bobbing events 
can be as short as 1-2 seconds, but can range as long as 5-10 seconds. 
Circle Swim 
The male swims away from the female for about 2 body lengths.  The 
return path is towards the female. Description follows that of glides in 
females (Cummings & Mollaghan 2006). 
Elongated 
Male Wave 
The male swims towards the glass moving his head from left to right at 
distances of about a quarter of its standard length.  The whole length of 
the male undulates with the initial head movement.  Two full cycles of 
the left and right motion usually constitutes the shortest elongated male 
wave. 
Inactivity 
Pause 
A time period of at least 5 seconds where the male does not have any 
propulsive movement. 
Lateral 
Display 
The male turns sideways towards the female for a time of at least one 
second and turns back to face the female.  During this time, male can be 
stationary or slowly drifting forward.  This display can be in a series 
that show off the same side repeatedly or in an alternating fashion.  The 
distance of this display also can vary by occurring at close distances 
(within body length) or at a moderate distance (2-3 body lengths). 
Description follows that in [Cummings and Gelineau-Kattner 2009]. 
Male Foraging A sequence of 3 or more feeding attempts at rocks or other objects. 
Parallel 
Swimming 
Male is swimming in parallel with the female within one body length as 
previously described [Cummings and Gelineau-Kattner 2009] 
Sexual Display Displays by males such as quivers and “C” shape body bends as previously described [Ryan and Causey 1989] 
Sigmoid 
Display 
The male quickly twists his body into “S” shape and brings his 
gonopodium towards his head as previously described [Ryan and 
Causey 1989] 
 
Solo 
Swimming 
A movement by the male that is not orientated towards the female and 
does not correlate with any female movement (Cummings and 
Gelineau-Kattner 2009). 
Sporadic 
Movement 
Male movement is characterized by burst of fast swimming in a random 
direction with multiple bouts possible. 
Up and Down The male swims up and down towards the female for at least two full cycles, modified from Cummings and Gelineau-Kattner (2009). 
 29
Table 1.2: Significant effects of male behaviors and preference measures. 
Strength of preference values are not in parentheses and preference Score values are in 
parentheses. 
 
 B (mean ± SE) χ2 p-value 
Circle Swim -234.968 ± 61.49 14.601 1.0 * 10-4 
Up and Down 56.062 ± 14.113 
(0.037 ± 0.0157) 
15.78 
(5.71) 
7.1 * 10-5 
(0.016) 
Solo Swim 259.096 ± 55.342 21.918 2.8 * 10-6 
Large Male * Circle 
Swim 
250.119 ± 62.144 16.199 5.7 *10-5 
Large Male * 
Bobbing 
-32.884 ± 16.219 4.111 0.042 
Large Male * Solo 
Swimming 
-199.57 ± 72.806 7.514 0.006 
(Large Male * 
Elongated Male 
Wave) 
(0.121 ± 0.0552) (4.83) (0.028) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 30
Chapter 2: Neural correlates of female mate preference behavior: a 
candidate gene approach 
  
ABSTRACT 
Female mate choice behavior is a critical component of sexual selection in many taxa. 
Previous research has identified brain regions involved in female mate recognition and 
assessment by measuring immediate early gene expression as a proxy for cellular 
activity. In this study we use a more context specific marker, a candidate gene associated 
with female mate preference (neuroserpin) in addition to an immediate early gene (egr-1) 
to identify brain regions associated with female preference in Xiphophorus nigrensis. Our 
study demonstrates context specific differences and significant correlations between gene 
expression and mate preference behavior within two telencephalic areas (Dm and Dl) not 
previously linked to sexual response or mate assessment, as well as a region traditionally 
associated with sexual behavior (POA). Network analysis of neuroserpin expression 
patterns show that these candidate preference-associated regions are more connected than 
other regions and that connectivity varies with the male social environment. Our results 
suggest that mate preference behavior may be coordinated not just through brain regions 
associated with sexual response but also through forebrain areas that may integrate 
primary sensory processing and reward.  
INTRODUCTION 
Broadly defined categories of social behavior such as aggression and reproduction 
are common across many taxa. The ability to identify potential unifying proximate 
 31
mechanisms in the brain, however, has been relatively limited until recently. Behavioral 
genomics has provided tremendous insights over the past decade into how behaviors 
across taxa share conserved genetic mechanisms [Fitzpatrick et al., 2005; Robinson et al., 
2008; Robinson et al., 2005; Wong and Hofmann, 2010].  Recent advances have allowed 
researchers to use high throughput molecular techniques to identify candidate genes 
(genes associated with a particular behavior) in behaviors ranging from foraging 
[Whitfield et al., 2003], aggression [Mukai et al., 2009; Renn et al., 2008] to mating 
[Lawniczak and Begun, 2004].  However, using the candidate gene approach [Fitzpatrick 
et al., 2005] to identify candidate brain regions associated with behaviors is relatively 
rare [Stockinger et al., 2005].  Here we employ this approach to determine whether the 
brain regions associated with female mate preference behavior are confined to a 
conserved circuitry regulating sexual behavior, or if additional brain regions are included 
in a preference-associated neural network.  
In the absence of candidate genes, researchers have used general cellular activity 
markers such as immediate early genes (IEGs) as a proxy for neural activity [Clayton, 
2000] to identify important brain regions associated with specific behaviors.  
Examinations of candidate gene expression patterns in the brain, however, provide the 
opportunity to identify underlying regions with greater specificity. In prairie voles, for 
example, peptides known to regulate social behavior (arginine vasopressin, oxytocin and 
their respective receptors) have identified a greater set of critical brain regions involved 
with pair bond formation than IEG investigations [Curtis and Wang, 2003; Young et al., 
2011].  Furthermore, in fruit flies the localization and subsequent manipulation of fru, a 
 32
gene associated with male courtship behavior, led to the identification of a neural circuit 
for male courtship behavior [Demir and Dickson, 2005; Stockinger et al., 2005].   
While previous research on female mate assessment in vertebrates have utilized 
IEGs [Desjardins et al., 2010; Gentner et al., 2001; Hoke et al., 2004; Hoke et al., 2005; 
Maney et al., 2003; Sockman et al., 2002; 2005; Woolley and Doupe, 2008], here we use 
a recently identified gene (neuroserpin) associated with female mate preference behavior 
in a northern swordtail, Xiphophorus nigrensis [Cummings et al., 2008]. Female 
swordtails exhibit a range of preference behavior strongly correlated with whole brain 
expression levels of neuroserpin in a context- and behavior- specific manner [Cummings 
et al., 2008].    Here we use in situ hybridization of this candidate gene (a context specific 
marker), neuroserpin, along with an IEG (cellular activity marker), egr-1, to more 
precisely identify brain regions associated with mate preference in X. nigrensis females. 
Neuroserpin is an extracellular serine protease inhibitor that has been characterized for its 
role in modulating synaptic plasticity and remodeling, neuroprotection and exploratory 
behavior in rodents [Madani et al., 2003; Miranda and Lomas, 2006]. Neuroserpin 
expression in the context of mate choice may facilitate exploratory behavior in females, 
leading to a higher sampling of males. 
From mammals to teleosts, there is a conserved network of brain regions 
implicated in social behavior (e.g. sexual behavior, territoriality, maternal care) known as 
the social behavior network (SBN, [Goodson, 2005; Newman, 1999]). Specific nodes of 
the SBN have been implicated in female mate assessment and sexual behavior [Blaustein 
and Erskine, 2002; Desjardins et al., 2010; Goodson, 2005; Kendrick et al., 1995; 
 33
Newman, 1999]. We quantify neuroserpin and egr-1 expression in 10 brain regions 
including five linked to sexual behavior (e.g. SBN nodes) and five linked to other social 
processes (e.g. reward, memory, learning, and multisensory processing). By exploring 
expression patterns both within as well as outside the SBN we can begin to characterize 
whether female mate preference is regulated by the same circuitry as female 
receptivity/copulation or whether preference invokes a unique network. To this end, we 
work with a study taxon that allows us to evaluate the neural processes involved in mate 
choice prior to sexual contact, X. nigrensis (Poecillidae). 
 Poeciliid fishes are a useful taxa to study female mate choice due in part to the 
variety of sexually dimorphic traits, wide range of individual variation, and the relative 
ease of studying female behavior in the lab and field (reviewed in [Houde, 1997; Ryan 
and Rosenthal, 2001]). The ability to measure behavioral proxies for female preference 
(e.g. association time bias [Basolo, 1990; Cummings and Mollaghan, 2006; Ryan and 
Causey, 1989; Ryan and Wagner, 1987], preference score [Cummings et al., 2008], and 
receptivity (glide displays [Cummings and Mollaghan, 2006]) in the laboratory without 
the confounds of physical contact between the sexes is advantageous when exploring the 
neural and molecular mechanisms of mate preference. In our focal poeciliid, X. nigrensis, 
males mature into genetically influenced size classes with larger size classes that court 
females and the smallest class exhibiting force copulations [Ryan and Causey, 1989]. 
Females prefer the larger sized males in the laboratory [Cummings and Mollaghan, 2006; 
Ryan and Causey, 1989], which is consistent with the observed mating advantage of 
larger males in the wild [Ryan et al., 1990; Ryan et al., 1992], and female preference 
 34
behaviors are not dependent on reproductive cycle status [Ramsey et al., 2011]. This 
makes them an excellent taxon to work with for studying preference independent of 
reproductive state.  
In this study we use a candidate gene approach in an established system for 
studying female mate choice to identify specific brain regions associated with female 
mate preference. We quantified the expression of a gene associated with mate preference 
(neuroserpin, Experiment 1) and an immediate early gene (egr-1, Experiment 2) in 
females exposed to mate preference or non-mate preference social contexts. With this 
approach, we test the hypothesis that the neural circuitry associated with female mate 
preference behavior is constrained to brain regions governing sexual response (nodes of 
the SBN) as well as an alternative hypothesis that female mate preference behavior is 
associated with brain regions distinct from those associated with sex.  
 
MATERIAL AND METHODS 
Behavioral Paradigm 
All experiments were conducted with sexually mature wild caught or progeny of wild 
caught female X. nigrensis maintained at UT’s Brackenridge Field Laboratories in 
Austin, TX. We followed established behavioral measurements, dichotomous choice 
paradigm, and natural lighting conditions to assess preference in this species [Cummings 
et al., 2008]. Immediately prior to the behavior trials, we measured the estradiol levels of 
each female through a non-invasive waterborne assay (see below). Females were 
acclimated for 5 min. in the center (neutral region) of the experimental tank and allowed 
 35
to interact with either end stimuli (behind plexiglass partitions) for 30 min., with stimuli 
switched after 15 minutes to avoid side bias. Females in Experiment 1 (neuroserpin 
quantification) were subjected to one of five conditions: two size-matched large males 
(LL, n = 10) with one male behind a UV pass barrier and the other behind a UV blocking 
filter, one large and small male (LS, n = 13), two size-matched small males (SS, n = 7), 
two size-matched females (FF, n = 12), or collected from their home tank (HT, n = 6), a 
treatment serving as an asocial control group (females are housed in isolation).  Home 
tank females underwent estradiol measurements as the other groups but were then 
returned to their home tanks for 30 min. prior to sacrifice. We selected our three male-
exposure groups to represent a gradient in mate choice complexity with the LL treatment 
group representing a relatively complex mate preference environment (two preferred 
phenotypes varying in UV ornamentation), and the LS and SS treatment groups 
representing a simple and minimum mate preference environment, respectively. Females 
in Experiment 2 (egr-1 quantification) were subjected to either LS (n = 10), FF (n = 10) 
social exposure conditions, or an asocial control (AA, n = 10) wherein the focal female 
was placed in the experimental tank without any stimuli at either end.  All size-matched 
stimuli differed by no more than 1 mm standard length.   
We recorded the number of female glides (a proxy for receptivity that can precede 
copulatory events [Cummings and Mollaghan, 2006; Liley, 1965]), transits to the center 
(proxy for general locomotor activity), and association bias. Association bias is defined 
as proportion of time spent with stimulus a where time spent with stimulus a > stimulus 
b. We then used these measures to calculate a composite preference score that 
 36
encompasses both time and behavior (preference score  (PS) = association bias + log[(1 + 
receptivity displays towards the biased stimulus)/total transits]) as in [Cummings et al., 
2008]. As the PS involves a log transformation of our behavioral measures, more positive 
PS indicates the female showed both a relatively higher bias in association time and 
glides toward one stimulus (normalized by general locomotor activity); whereas more 
negative PS indicates the female generally showed relatively little bias in association time 
and/or behavior.  
Females were isolated at least two weeks before behavioral testing to ensure 
sexual motivation. Each female was pre-tested twice with LS stimuli prior to group 
assignment to ensure similar baseline preference responses across experimental groups.  
Females assigned to five experimental groups in experiment 1 (LL, LS, SS, FF, HT) or to 
three experimental groups in experiment 2 (LS, FF, and AA) showed no significant 
differences in pre-test preference trials (neuroserpin, ANOVA, PS: p = 0.992; egr-1 
ANOVA, PS: p = 0.122).  We identified “high” performing females for each behavior of 
interest (preference score, transits, glides, or association bias) as those exhibiting greater 
than the median value and compared their gene expression in each brain region with 
females identified as “low” performing (exhibiting less than the median value).  For 
context specific comparisons, we examined the relationship between gene expression and 
behavior in each region for females exposed to males (LL, LS, and SS) relative to 
female-exposed females (FF).  We subsequently examined the unique covariation 
patterns between brain regions for each male-exposed environment (LL, LS or SS). 
 
 37
Estradiol Measurements  
We quantified a proxy for circulating estradiol levels for all females through a non-
invasive waterborne assay following an established protocol for teleosts [Earley et al., 
2006; Kidd et al., 2010; Scott et al., 2008] and validated in our focal species [Ramsey et 
al., 2011]. Briefly, females were placed in a 250 mL glass beaker containing 150 mL of 
reservoir water (treated tap water used for home and experimental tank) for one hour 
prior to behavior trials. Estradiol was extracted from the water using C18 Solid Phase 
Extraction columns (Sep-Pak Plus C18 cartridge 55–105 lm; Waters Corporation, 
Milford, MA) and measured on a Correlate-EIA 17β-estradiol Enzyme Immunoassay Kit 
(Assay Designs) according to manufacturer’s protocol. Hormone samples were run on 
three 96-well EIA assay plates: inter-assay CV was 6.5% and intra-assay CV was 1.9%.  
 
Tissue Processing 
Female brains were cryosectioned at 16 µm thickness onto four serial series. For each 
experiment, all series were simultaneously post-fixed in cold 4% paraformaldahyde/PBS 
solution, washed in PBS and acetylated in 0.25% acetic anhydride/triethanolamine. 
Subsequently, slides were washed in 2X standard saline citrate, dehydrated in increasing 
ethanol series and stored at -80 C. In Experiment 1, one series was used to localize and 
quantify neuroserpin expression using a digoxigenin (DIG) labeled probe (see below) , 
and an adjacent series was used to validate the use of DIG as a semi-quantitative method 
by using S35 labeled neuroserpin probe (see below). 
 
 38
Probe Synthesis 
Digoxigenin (DIG) labeled neuroserpin and egr-1 riboprobes were generated using a 1:3 
ratio of UTP and DIG-UTP (Roche) following a modified manufacturer’s protocol 
(Megascript T7/T3, Ambion). A 379 base pair neuroserpin DIG probe template was 
subcloned (Gene Accession #DQ839542.1) by using primer pair 5’- 
TTATCGCTCTCTGTGGTCTGTGCT -3’and 5’-
TGGGCAAATCGGATCACATAGTGG -3’. The 337 base pair egr-1 probe template 
was subcloned by using primer pair 5’-CATCCTCCTCATCAACCATCTCG-3’and 5’-
TGACTCTGGAAGGGCTTTTGG-3’on the X. nigrensis egr-1 transcript (Gene 
Accession # DQ835282). After probe synthesis, we removed unincorporated nucleotides 
via column filtration according to manufacturer’s protocol (Megaclear, Ambion) and 
checked probe quality through gel electrophoresis. 
Radioactive (S35-UTP, 800 Ci/mmol, Perkin Elmer) labeled neuroserpin 
riboprobes were generated following manufacturer’s protocol (Maxiscript T7/T3, 
Ambion). A 563 base pair probe was subcloned by using primer pair 5’- 
CAGGCTGAAGTGGAAACTGTGG -3’and 5’- GCATCATCGGTGAAGATTTTGG -
3’on the X. nigrensis neuroserpin transcript (Gene Accession # HM107107). A pilot 
study showed that this probe template generated a higher signal to noise ratio with S35 
UTP than probe template used for the DIG in situ hybridization. We used Nucaway spin 
columns (Ambion) to purify the probe according to manufacturer’s protocol. We verified 
probe integrity by gel electrophoresis and quantified using a scintillation counter.  
 
 39
in situ hybridization 
We used a modified digoxigenin in situ hybridization protocol [Shoemaker et al., 2007]. 
Briefly, slides were prehybridized with solution containing 50% formamide, 5X SSC, 5X 
Denhardt’s solution, 250 ug/ml yeast tRNA, and 500 ug/ml herring sperm DNA for 6 
hours at 60°C in a hybridization chamber containing chamber buffer solution (50% 
formamide, 2X SSC). We then hybridized the slides overnight at 65°C with fresh 
prehybridization solution containing 0.25 ng of antisense riboprobe per slide in 
hybridization chambers containing buffer. Following hybridization we conducted washes 
at low stringency (two washes in 2X SSC at room temperature), then RNase A treated the 
slides (0.5M NaCl, 10 mM Tris pH 8.0, 2.25 mM EDTA, 0.2 µg/ml RNase A), followed 
by increasingly stringent washes (2X, 1X, 0.5X, 0.25X SSC) and then a final wash in 
Buffer B1 (100 mM Tris pH 7.5, 150 mM NaCl). Sections were then incubated overnight 
at 4°C with Anti-Digoxigenin AP antibody (Roche). After antibody incubation we 
washed sections twice in Buffer B1 and then blocked endogenous alkaline phosphatase 
activity with a 30 min wash in the dark in Buffer B3 (100mM Tris pH 9.5, 100 mM 
NaCl, 50 mM MgCl2, 5 mM levamisole). We carried out colorimetric detection using 
NBT/BCIP stock solution (Roche). We stopped the colorimetric reaction (75 min) by 
rinsing sections three times in ultrapure type 1 water, then progressively dehydrated 
sections in ethanol washes (25%, 50%, 70%, 95%), and finally coverslipped the slides 
with Permount adhesive (Fisher). Within each experiment all individuals were processed 
simultaneously to avoid any potential colorimetric development differences across 
individuals due to batch effects. We conducted the following in situ hybridization 
 40
controls to test for binding specificity for egr-1: DIG egr-1 sense riboprobe, ½ 
concentration of antisense DIG riboprobe, unlabeled antisense riboprobe, 1:25 cold 
competitor antisense riboprobe, no probe added and a RNase treated series prior to 
hybridization with experimental concentrations of DIG antisense. All controls showed 
very little or reduced expression of egr-1 (data not shown). For neuroserpin, DIG labeled 
neuroserpin sense riboprobes showed no expression (data not shown). 
 To validate our use of DIG-labeled probes as a semi-quantitative method, we also  
conducted an in situ hybridization using S35 neuroserpin riboprobes on an adjacent series 
to the DIG-labeled series of Experiment 1, by modifying a previous protocol [Hoke et al., 
2004]. Briefly, for all slides we applied 250 ul of hybridization solution (50% formamide, 
10% Dextran sulfate, 300 mM sodium chloride, 8 mM Tris pH 8.0, 0.8 mM EDTA pH 
8.0, 1X Denhardt’s solution, 8 mM dithiothreitol (DTT), 250 ug/mL yeast tRNA, and 50 
ug/mL Herring Sperm DNA) to each slide with 4500 CPM of probe/µl of solution, 
coverslipped and hybridized 16 hrs at 65°C in hybridization chambers with buffer (50% 
formamide, 2X SSC). After hybridization, we washed slides to remove nonspecific probe 
binding as follows: 1 hr in 4X SSC at 65 C, 1 hr in 50% formamide and 2X SSC, 15 min 
in RNAse buffer (0.5M NaCl, 10mM Tris pH 8.0, 0.1125 µm EDTA) 30 min RNAse 
treatment (20 µg/ml) at 37 C, 15 min in RNAse buffer, 15 min in 2X SSC and finally 15 
min in 0.1X SSC. All washes unless otherwise stated occurred at room temperature and 
contained 1 mM DTT except for the RNAse treatment. After washes, slides were 
dehydrated in increasing ethanol series (25%, 50%, 70%, 95%, 100%) containing 300 
mM ammonium acetate and cleared in xylenes. To visualize the probes, all slides were 
 41
dipped in 37ºC Kodak NTB2 emulsion (Kodak) and dried at 60ºC for 2 hours. Slides 
were then stored in light proof boxes at 4ºC for 7 days. We developed slides using D19 
developer (Kodak) and Kodak Fixer (Kodak). Afterwards, we stained the tissue in cresyl 
violet and coverslipped with Permount (Fisher).Sense S35-labeled probes showed nearly 
no signal above background (data not shown). 
 
Gene expression quantification 
Using a X. helleri brain atlas for reference and terminology [Anken and Rahmann, 1994], 
we identified and quantified digoxigenin-labeled riboprobe expression in 10 brain 
regions. These brain regions include putative tetrapod homologs [Bruce and Bradford, 
2009; Goodson, 2005; Goodson and Kabelik, 2009], for nodes in the social behavior 
network (SBN[Goodson, 2005; Newman, 1999]), reward system (the mesolimbic 
dopaminergic reward pathway [Wise, 2002], for teleost homologs, see [O’Connell and 
Hofmann accepted]), and other regions. The 10 regions are (abbreviation, putative 
tetrapod homolog, pathway classification): cerebellum (Cb, cerebellum, neither SBN nor 
reward system), area dorsolateralis telencephali (Dl, pallial hippocampus, reward 
system), area dorsomedialis telencephali (Dm, basolateral amygdala, reward system), 
central gray (GC, periaqueductal gray, SBN), hypothalamus ventralis (HV, ventral 
hypothalamus, neither SBN nor reward system), pituitary (Pit, pituitary, neither SBN nor 
reward system), nucleus preopticus (POA, preoptic nucleus, SBN), nucleus tuberis 
anterioris (TA, ventromedial hypothalamus, SBN), ventralis supracommissuralis 
telencephali (Vs,medial amygdala, SBN), and area ventroventralis telencephali (Vv, 
 42
lateral septum, SBN and reward system). After tissue processing final sample sizes for 
quantifying neuroserpin were: LL, n = 10; LS, n = 10; SS, n = 5; FF, n = 9; HT, n = 5. 
For quantifying egr-1 (experiment 2) final sample sizes were: LS, n = 6; FF, n = 7; AA, n 
= 10.   
 
Digoxigenin quantification 
We quantified gene expression by measuring the optical density (OD) of the digoxigenin 
labeled probes, which has been established as a semi-quantitative measure of gene 
expression in other systems [Guiot and Rahier, 1997; Harvey-Girard et al., 2010; Larsson 
et al., 1991; Machaalani and Waters, 2002; O’Connell et al., 2011; Robbins et al., 1991; 
Vecchiola et al., 1999; Zhang et al., 1995; Zhao et al., 2008]. For each slide, we 
normalized the mean intensity of all measures to the background (mean intensity of slide 
not containing tissue), which produced a value for the fractional transmittance of the 
brain region in each section. Fractional transmittance was mathematically converted to 
optical density by the equation OD = 2-log(Fractional Transmittance), which was derived 
specifically for the imaging setup (Nikon Eclipse 80i) in our laboratory using neutral 
density filters 0, 8 and 32.  Using NIS Elements image analysis software (Nikon), we 
measured the OD of neuroserpin and egr-1 expression across individuals from a 
standardized portion of each brain region (ranging from 1737 - 29152 μm2 depending on 
size of the brain region of interest). 
 Brain section images were captured at 4X using a Nikon 12-bit 2-megapixel 
monochrome camera (DS-2MBWc). For each brain region, we used Nikon NIS Elements 
 43
2.3 to measure a standardized rectangular box (area of box given after each brain region, 
see below) within the borders of each brain region and measured the mean intensity of 
neuroserpin and egr-1 expression within the box. Unless otherwise stated, the measuring 
box was always placed in the middle of the brain region on the dorsal-ventral plane. We 
measured the mean intensity in both hemispheres if available and then averaged the 
values for the section. For all brain regions we then proceeded to measure the mean 
intensity for all sections of the individual in which we could identify the desired brain 
region and then averaged for that individual. Depending on the size of the brain region, 
the number of sections averaged per individual ranged from two to eight consecutive 
sections. Consecutive sections spanned 48 μm apart. Specifically, we measured Dl 
(19906 μm2) & Dm (19906 μm2) for sections from the rostral most section until the 
disappearance of the anterior commissure; Vv (5877 μm2) & Vs (5877 μm2) for sections 
only containing the telencephalic ventricle; POA (2843 μm2) for sections with anterior 
commissure present until the appearance of the optic tectum; Pit (12832 μm2) for all 
sections observed;  HV (1737 μm2) from the pituitary until the lateral extension of the 
lateral hypothalamus; TA (5339 μm2) for 3 sections preceding the lateral extension of the 
dorsal hypothalamus; Cb (29152 μm2) and CG (7298 μm2) for all sections observed (see 
main text for brain region abbreviations). All image capture and data collection was done 
by a single person blind to the treatments. 
 
 44
DIG validation with S-35 riboprobe 
For one brain region, Dm, we manually counted the number of S35-labeled cells 
expressing neuroserpin (containing at least one silver grain) to compare with the optical 
density measures of DIG-labeled neuroserpin expression. To quantify the number of 
neuroserpin positive cells for each individual, we averaged the number of cells counted 
from three consecutive sections (each section spans 48 μm apart). For each section we 
collected images of two non-overlapping fields of Dm at 100X (each field measured 
12124 µm2) modifying a previously established protocol [Burmeister et al., 2005]. For 
each unique field, two images were taken: one under brightfield where we focused on cell 
resolution and the second image taken under darkfield where we focused on silver grain 
resolution. The counting image was created by superimposing the two original images 
using Photoshop CS4. Dm images for all individuals were counted by three observers 
blind to the treatment using Photoshop CS4. Total number of cells did not differ by 
treatment group (F=0.678, p = 0.572).  
Statistics 
All statistics were performed in SPSS (ver. 18) and the network statistics were conducted 
using Ucinet [Borgatti et al., 2002]. We used a t-test to examine group-wide behavioral 
and gene expression differences between individuals expressing high versus low 
behaviors in each treatment group and a Benjamini-Hochberg correction [Benjamini et 
al., 2001] for multiple hypothesis testing. To assess relationships between individual 
variation of preference behavior, gene expression, or estradiol levels we used Pearson’s 
 45
correlation when the data was normal and Spearman correlation when data was non-
normal (glides and transits). 
 We characterize neuroserpin network expression patterns by examining pairwise 
correlations of neuroserpin expression between regions in specific social contexts. To 
analyze network patterns, we converted all Benjamini-Hochberg corrected significant 
correlations of neuroserpin expression between regions into binary values in an 
association matrix (1 = significant correlation, 0 = non-significant).  Degree centralities 
for each node in each network were calculated in Ucinet and then compared to other 
nodes in the same network using a Wilcoxon Rank Sum test in SPSS. For each mate 
preference group (LL, LS, SS), we define unique correlations as those that are not also 
significant in the FF or HT groups. Due to small sample size, we did not analyze egr-1 
network expression patterns. 
 
RESULTS 
As expected when looking at association time (LS, tn=10 = 2.1, p = 1.1 * 10-7 for 
neuroserpin experiment and tn=6 = 2.17, p = 0.009 for egr-1 experiment, Figure 2.1A & 
B), or preference score calculated for each stimulus type (LS, mean ± SE, PSlarge male = -
0.05 ± 0.2, mean PSsmall male = -0.79 ± 0.14, tn=10  = 2.91,  p = 0.009 for neuroserpin 
experiment; mean PSlarge male =  0.19 ± 0.23, mean PSsmall male = -0.48 ± 0.21 and tn=6 = 
2.12, p = 0.05 for egr-1 experiment), females in our study displayed a clear preference 
towards the large male over the small male in both experiments. There was no significant 
difference in time spent with a particular stimulus phenotype in the other test groups: two 
 46
females, (FF, tn=9 = 0.84, p = 0.41 for neuroserpin and tn=8 = -1.28, p = 0.22 for egr-1); 
two small males, (SS, tn=5 = 0.12, p = 0.9 for neuroserpin), two large males (LL, tn=10 = 
0.75, p = 0.45 for neuroserpin, Figure 2.1A & B). In experiment 2, females exposed to an 
empty stimulus environment (AA) exhibited a side bias (tn=10 = -2.03, p = 0.05 for egr-1, 
Figure 2.1B).  
The S35-labeled riboprobes provided validation of the DIG-labeled approach as 
evidenced by (i) a significant positive correlation between the two quantification methods 
(r = 0.351, p = 0.008, Figure 2.2), (ii) significant results with both approaches for group 
level comparisons in preference score (both methods showed higher expression of 
neuroserpin in Dm of male-exposed females exhibiting high preference over low 
preference females), (iii) consistent non-significant results with both approaches for other 
behavior comparisons (glides, transits) and for females exposed to other females, and (iv) 
significant correlations between neuroserpin expression in Dm and preference score for 
male-exposed females in both approaches (see below for ii – iv results). 
Using S35-labeled neuroserpin riboprobes, we found that in Dm, females exposed 
to males and expressing a high preference (> median) had a significantly greater number 
of neuroserpin positive cells over females displaying low preference (< median) (mean ± 
SE: high preference score = 309.56 ± 22.85, low preference score = 213.89 ± 18.04, t-
Stat = 2.079, p = 0.003, Table 2.1). DIG-labeled neuroserpin riboprobe OD values 
provided consistent results as the S35 riboprobes in Dm (t-Stat = 3.284, p = 0.003, 
corrected for multiple hypothesis testing, Figure 2.3). Two other brain regions also 
showed significantly higher neuroserpin OD in females showing high preference over 
 47
females exhibiting low preference in the male-exposed groups (Dl, t-Stat = 2.91, p = 
0.008 ; POA, t-Stat = 3.292, p = 0.003, both corrected for multiple hypothesis testing, 
Figure 2.3). We also observed a significant difference in neuroserpin OD in HV (t-Stat = 
2.489, p = 0.021, Figure 2.3) for male-exposed females but this was not significant after a 
multiple hypothesis correction. There were no significant differences in neuroserpin 
expression in male-exposed females for other behaviors (glides or transits) nor significant 
difference in female-exposed females (FF) exhibiting high relative to low preference 
behavior (Table 2.2 & 2.3). Consistent with OD measures, within Dm we also did not see 
any significant differences in the number of neuroserpin positive cells using S35 labeled 
neuroserpin probes in any other context or behavior (Table 2.1).   
Of the four brain regions showing within-group differences in neuroserpin OD, 
three of these exhibited significant positive correlations between female preference score 
and neuroserpin expression in male-exposed females only: Dm (r n=25 = 0.522, p = 
0.007), Dl (r n=25 = 0.501, p = 0.011), and POA (r n=25 = 0.479, p = 0.015) but not HV (r 
n=25 = 0.368, p = 0.07) (Figure 2.4). We also observed a consistent correlation with 
preference score when quantifying S35 neuroserpin expression in Dm (r n=25 = 0.405, p = 
0.049, Figure 2.5). Similar to the neuroserpin results, egr-1 expression correlated with 
preference score in male-exposed females in Dm (rn=6 = 0.829, p = 0.041) and Dl (r n=6 = 
0.781, p = 0.067) while showing no patterns of correlation in FF or AA (Table 2.4), and 
showed no significant correlations with transits or glides in Dm, Dl, or POA in mate 
preference contexts (Table 2.5). Furthermore there were no significant correlations 
between circulating estradiol levels and preference score, glides, transits, or gene 
 48
expression in any brain region for the male exposed group in experiment 1 (Table 2.6). In 
experiment 2, a correlation between circulating estradiol levels and egr-1 expression was 
found in the TA for LS (r = 0.805, p = 0.028), however, this was not significant after a 
multiple hypothesis correction (Table 2.6). 
To explore neuroserpin expression network patterns across the brain regions 
measured in each context, we only consider significant correlations after correcting for 
multiple hypothesis testing. Within each of the male preference groups (SS, LS, or LL) 
the candidate regions associated with mate preference (Dm, Dl, and POA) have a 
significantly higher degree centrality than other regions (Table 2.7). Focusing on the 
unique neuroserpin correlations within each male preference group (LL, LS, SS) we find 
that some connections are constant across all three groups (e.g. Dm and POA) while 
others appear only in specific male environments (e.g. Dm and Vv in the presence of 
large males (LL, LS) but not in small male only conditions (SS), Figure 2.6). 
Furthermore, there is a significant higher density (i.e. total number of significant 
correlations between regions) for females exposed to LL relative to LS (t = 3.23, p = 
0.0026) and SS (t = 3.52, p = 0.0028) male groups. 
 
DISCUSSION 
Female swordtails showed brain region-specific patterns of neuroserpin 
expression that were both context and preference specific. Females with high preference 
scores had significantly higher neuroserpin expression relative to low preference score 
females in Dm, Dl, POA and a trend in HV. We identified Dm, Dl, and POA as important 
 49
areas in processing female mate assessment information as there are also significant 
positive correlations between individual variation in preference behavior and 
neuroserpin. We obtained complementary results with egr-1 for Dm and Dl. Our results 
suggest that female mate preference behavior is modulated by brain regions associated 
with the Social Behavior Network (e.g. POA) as well as brain regions outside the SBN 
that are distinct from those governing the sexual response.  The preoptic area (POA), 
hypothalamic region, and uni-modal sensory processing centers (e.g. acoustic centers in 
the avian and amphibian brain) have been previously linked to female mate assessment 
[Desjardins et al., 2010; Gentner et al., 2001; Hoke et al., 2004; Hoke et al., 2005; Maney 
et al., 2003; Sockman et al., 2002; 2005; Woolley and Doupe, 2008]. However, this is the 
first study to show a link between female mate preference behavior and non-SBN regions 
of Dm and Dl. 
The association between the telencephalic brain regions of Dm and Dl with 
female preference behavior implies a possible link between sensory processing centers 
and other regions mediating sexual response (e.g. SBN nodes). Given that in teleosts Dm 
and Dl receive multimodal input relayed from the preglomerular complex and project to a 
variety of other fore- and mid-brain regions including the POA [Northcutt, 2006; 2008], 
these telencephalic brain regions may be prime candidates in mediating sensory 
integration and discrimination processes that are then directly relayed to the POA or 
indirectly to the HV to mediate receptivity/copulation behavior. The specific functions of 
the teleost Dm and Dl are still largely unknown, however, lesion studies outside of mate 
 50
choice contexts have shown that Dm and Dl are involved in analogous measures of 
emotional and spatial learning in fish, respectively [Broglio et al., 2005].  
Learning and memory may play a role in mate choice in poeciliids, as evidenced 
in studies showing mate choice copying [Schlupp et al., 1994] and age-dependent female 
preference functions [Morris et al., 2006], however this hypothesis has not yet been tested 
in X. nigrensis. Long-term preference data collected in our lab with X. nigrensis suggests 
that females do change preference behavior as a function of size, a proxy for age [Wong 
et al., in review], thereby lending some support for the role of learning in preference 
response.  Future experiments that compare the response of these candidate brain regions 
between naïve and experienced females in mate choice contexts should be informative. 
The possibility that our results stemmed from circuitry regulating a general social 
preference response rather than a mate choice specific response is unlikely given that we 
did not find any significant relationships between preference behavior and either 
neuroserpin or egr-1 expression in other treatment groups (FF or AA, Table 2.4). 
Furthermore, specific behavioral components (glides or transits) of the preference score 
cannot explain our observations, as there are no significant correlations between these 
behaviors and gene expression in Dm, Dl or the POA (Table 2.5). Rather, the significant 
correlations between gene expression (neuroserpin and egr-1) and preference behavior 
observed only within male-exposed groups suggest that Dm, Dl and the POA are 
candidate regions associated with processing female mate preference information.  
Interestingly, both neuroserpin and egr-1 expression patterns in this study are 
independent of circulating estradiol in our candidate preference-associated brain regions 
 51
in any of the tested contexts (Table 2.6). However, the role of localized estradiol 
production in female preference behavior and candidate preference-associated genes 
cannot be excluded. 
Evidence suggests that Dm and Dl are homologs of the tetrapod pallial 
(basolateral) amygdala and hippocampus, respectively [Bruce and Bradford, 2009; 
Northcutt, 2008]. As these tetrapod homologs are part of the mesolimbic reward pathway 
in mammals [Wise, 2002], female mate preference behavior may also influence or be 
influenced by this pathway. The amygdala and hippocampus in rodents have been 
implicated in modulating motivation and reward in addition to their established roles in 
spatial and emotional memory [Ambroggi et al., 2008; Tracy et al., 2001; Tye and Janak, 
2007]. The reward circuitry in teleosts has recently been described and includes Dm and 
Dl [O’Connell and Hofmann accepted]. There are neuroanatomical connections between 
Dm and the posterior tuberculum (putative ventral tegmental area homolog [O’Connell 
and Hofmann accepted]), which is a major source of dopaminergic neurons projecting to 
the telencephalon [Northcutt, 2006; Rink and Wullimann, 2001] and between Dl and the 
putative nucleus accumbens homolog [O’Connell and Hofmann accepted]. Furthermore, 
in another teleost (Astatotilapia burtoni), Dm, Dl, POA, and HV all express mRNA for 
dopamine receptors [O'Connell et al., 2011]. While copulation and post-copulatory pair-
bond formations involve the activation of the reward circuitry in mammals [Young et al., 
2011], it should be emphasized that copulation was not allowed in our design. This 
suggests an intriguing possibility that reward centers may be primed prior to sexual 
reward in a mate choice context. Given the putative homology and functional 
 52
conservation in Dm, Dl, POA, and HV between teleost and rodents, we hypothesize that 
these brain regions could be modulating or facilitating motivation in female mate 
preference or arousal behavior, possibly via a homologous mesolimbic reward circuitry in 
teleosts.  
Variation in behavior can also stem from unique changes in gene expression 
patterns across multiple regions [Goodson and Kabelik, 2009]. We characterized 
neuroserpin network expression patterns by looking at pair-wise correlations of 
neuroserpin expression between regions in specific social contexts. Candidate regions 
associated with mate preference (Dm, Dl, and POA) had a significantly higher degree 
centrality than other regions (Table 2.7) in each of the male exposed groups (SS, LS, or 
LL). Although at different levels of biological organization, studies examining protein 
interaction networks have found that proteins with a high degree centrality are more 
essential to the network [Hahn and Kern, 2005]. This suggests that these regions are 
important in the neuroserpin brain expression network under mate preference conditions.  
The three male stimuli environments (LL, LS, and SS) elicited different patterns 
of neuroserpin expression across regions. Specifically, the density of the network (i.e. 
total number of significant correlations between regions) changed according to the social 
context with increasing number of connections with increasing complexity of the mate 
choice conditions (Figure 2.6).  For example, in the minimal choice environment (lacking 
a large male phenotype, SS), we observed only a single significant correlation (Dm and 
POA, significant across all male-exposed conditions).  In the simple mate choice 
condition (a single large male phenotype, LS), the number of significant correlations 
 53
doubled.  Finally, in the most complex condition (two large males, LL), we observed 8 
significant correlations between regions including all three of the preference-specific 
brain regions.   Hence, variation in neuroserpin expression networks across male stimuli 
groups reflects a neural response that appears to scale with the complexity of the social 
environment. 
In this study, we begin to identify the circuitry of mate choice by using both a 
context specific marker (a candidate preference gene) as well as a traditional approach 
(IEG). This is the first study to identify multisensory processing, spatial learning, and 
putative reward regions (Dm, Dl) in conjunction with reproductive regions (POA, HV) as 
putative nodes in a female mate preference circuit. As our study evaluates females in the 
act of choosing (e.g. presented with two stimuli simultaneously), Dm and Dl may 
facilitate discernment of stimuli by integrating multi-sensory processing prior to enacting 
a sexual response. Network analysis show that Dm, Dl, POA may be “hubs” in the mate 
preference neuroserpin expression network and that correlated patterns of neuroserpin 
expression between regions increase with increasing complexity of the mate choice 
environment. Building on these putative preference-associated brain regions, future 
studies should examine how perturbations (e.g. lesion studies or pharmacological 
manipulations) directly affect female preference behavior. 
 
 
 
 54
Figure 2.1: Preference measures by stimuli and social group.  
Association time with each stimulus is plotted for (A) experiment 1 (neuroserpin) and 
(B) experiment 2 (egr-1). Circles show female response toward each stimulus with gray 
circles representing responses towards larger (LS, FF) or UV+ (LL) phenotype, or right 
side of experimental tank (AA).  Female Preference Scores for (C) experiment 1 
(neuroserpin) and (D) experiment 2 (egr-1) are shown in black circles. Red horizontal 
line is the mean with whiskers representing the standard error. ***, p < 0.001; **, p < 
0.01; *, p = 0.05 
 
 
 
 
 
 
 
 55
Figure 2.2: Correlation between neuroserpin expression in situ hybridization (ISH) 
quantification methods on adjacent series.  
There is a significant positive correlation (r = 0.351, p = 0.008) between optical density 
measured from digoxigenin ISH and number of neuroserpin positive cells measured from 
S35 ISH. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 56
Figure 2.3: Neuroserpin expression in females exposed to a mate preference context 
(LL, LS, and SS).  
Differences in neuroserpin expression between groups of high preference score (> 
median) females (n=12, blue) and low preference score (< median) females (n=12, red) 
for male-exposed females. **, p < 0.01; *, p < 0.05 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 57
Figure 2.4: Individual variation of preference score and neuroserpin expression.  
Significant correlations between individual variation in preference score and neuroserpin 
expression in A) Dm, B) Dl and C) POA. Triangles, diamonds, and squares represent LL, 
LS, and SS exposed females, respectively. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 58
Figure 2.5: Correlation between neuroserpin expression in Dm and preference score 
using S35 labeled riboprobes.  
Number of neuroserpin positive cells from S35 labeled riboprobes also show a significant 
correlation with preference score (r = 0.405, p = 0.049). Triangles, diamonds, and squares 
represent LL, LS, and SS exposed females, respectively. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 59
Figure 2.6: Neuroserpin expression network by context.  
Unique significant positive pair-wise correlations relative to FF and HT females in 
neuroserpin expression between brain regions (lines) in A) LL, B) LS, and C) SS 
exposed females. Brain regions in bold are those associated with mate preference 
identified in this study. 
 
 
 
 
 
 60
Table 2.1: Comparisons between in situ hybridization (ISH) quantification methods. 
Neuroserpin quantification (mean ± SE)  as related to “high” (> median) and “low” (< 
median) behavior in Dm. n.s., not significant 
 
 
 
 61
Table 2.2: Neuroserpin optical density (mean ± SE) comparisons between “high” (> 
median) and “low” (< median) preference score.  
 
** indicates significance after correcting for multiple hypotheses; * indicates significance 
that does not survive multiple hypothesis testing; n.s., not significant 
 
 
 
 62
Table 2.3: Neuroserpin optical density (mean ± SE) comparisons between “high” (> 
median) and “low” (< median) behaviors.  
 
n.s., not significant 
 
 
 
 63
Table 2.4: Correlations between preference score and gene expression in Dm, Dl, POA 
in non-sexual contexts. 
 
 
 
Experiment 1 
(neuroserpin) 
Experiment 2  
(egr-1) 
FF FF AA 
Correlation 
coefficient p-value 
Correlation 
coefficient p-value 
Correlation 
coefficient p-value 
Preference Score 
Dm 0.239 0.536 0.301 0.511 -0.239 0.505 
Dl 0.183 0.636 0.329 0.471 -0.265 0.458 
POA -0.308 0.418 -0.191 0.711 -0.43 0.247 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 64
Table 2.5: Correlations between glides, transits and gene expression in Dm, Dl, POA in 
male exposed environments. 
 
 
 Experiment 1 
LL, LS and SS 
(neuroserpin) 
Experiment 2  
LS 
(egr-1) 
Correlation 
coefficient p-value 
Correlation 
coefficient p-value 
Glides 
Dm 0.204 0.327 0.714 0.11 
Dl 0.244 0.239 0.771 0.072 
POA 0.164 0.432 -0.257 0.622 
Transits 
Dm -0.306 0.136 0.257 0.622 
Dl -0.258 0.213 0.314 0.544 
POA -0.324 0.113 -0.257 0.622 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 65
Table 2.6: Correlation between a proxy for circulating estradiol levels and preference 
score, glides, transits, and gene expression in different brain regions in male 
exposed environments. 
 
 
 
Experiment 1 
LL, LS and SS 
 (neuroserpin) 
Experiment 2  
LS  
(egr-1) 
Correlation 
coefficient p-value 
Correlation 
coefficient p-value 
Preference Score -0.296 0.149 -0.024 0.957 
Glides 0.401 0.047 -0.071 0.879 
Transits 0.37 0.06 -0.07 0.879 
Dm -0.106 0.611 -0.709 0.114 
Dl -0.133 0.525 -0.649 0.162 
Cb -0.139 0.505 0.543 0.207 
GC -0.166 0.427 0.02 0.971 
Pit -0.327 0.117 0.646 0.116 
POA -0.077 0.712 0.184 0.725 
TA -0.129 0.545 0.805 0.028 
HV -0.057 0.785 0.714 0.071 
Vs -0.173 0.406 0.329 0.523 
Vv -0.109 0.611 -0.391 0.443 
 
 
 
 
 
 
 
 
 
 
 
 66
Table 2.7: Comparison of degree centrality between candidate nuclei (Dm, Dl, POA) 
and other nuclei in each treatment group for females used to localize 
neuroserpin. 
 
 
Average degree 
centrality of 
candidate nuclei  
± standard error 
Average degree 
centrality of 
other nuclei ± 
standard error 
Wilcoxon rank 
sum Z-score p-value 
LL 0.444 ± 0.064 0.158 ± 0.033 2.44 0.0147 
LS 0.259 ± 0.037 0.111 ± 0.024 2.342 0.0192 
SS 0.222  0.031 ± 0.02 2.598 0.0094 
FF 0.185 ± 0.037 0.142 ± 0.046 0.598 0.5501 
HT 0.444 ± 0.064 0.254 ± 0.961 0.932 0.3515 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 67
Chapter 3: Characterizing the roles of tyrosine hydroxylase and 
neuroligin-3 in female mate preference behavior in a teleost 
 
  
ABSTRACT 
Choosing whom to mate with is one of the most important processes in a female’s 
life. Studies are just beginning to identify the neural and molecular mechanisms 
underlying variation in female mate preference. These initial studies have begun to 
identify putative brain regions involved in mate preference, yet how dopamine pathways 
influences this critical behavior is still largely unknown. In this study, we explore the 
hypothesis that female mate preference in Xiphophorus nigrensis is modulated by natural 
changes in rapid (within 30 minutes) dopamine production (via tyrosine hydroxylase, TH) 
that occur in anticipation of copulation. We also characterize the role of neuroligin-3 (a 
gene associated with female mate preference) in the brain, which supports a hypothesized 
network of brain regions associated with female mate preference determined using a 
different gene, neuroserpin. Neuroligin-3 expression in a mate preference context show 
significant correlations with female preference in two telencephalic areas (Dm, Dl), a 
hypothalamic nucleus (HV), and two regions associated with sexual and social behavior 
(POA, Vv). We do not observe any context or behavioral specific changes in dopamine 
production concomitant with female preference in any of the brain regions examined. 
Analyzing TH and neuroligin-3 expression patterns across different brain regions show 
that connectivity varies with male social environment only for neuroligin-3: density of 
 68
the neuroligin-3 network was positively associated with mate choice contexts that require 
more evaluation by the females. This study provides additional support for a set of brain 
regions associated with female mate preference using neuroligin-3 but we did not detect 
any relationship between dopamine production and mate preference with 30 minutes of 
stimulus presentation in X. nigrensis. 
INTRODUCTION 
The role of dopamine in reward-seeking and sexual behavioral displays is widely 
documented across species [Barron et al. 2010; Hull et al. 1999; Melis and Argiolas 
1995; Paredes 2009; Paredes and Agmo 2004; Schultz 2007a].  Decades of research have 
uncovered the activation of dopamine pathways as critically important in male sexual 
arousal ([Ball and Balthazart 2010; Everitt 1990; Melis and Argiolas 1995]), male 
courtship behavior [Goodson et al. 2009; Heimovics et al. 2009; Heimovics and Riters 
2008], female receptivity [Hull et al. 1999; Melis and Argiolas 1995], as well as male and 
female copulation [Everitt 1990; Hull et al. 1999; Melis and Argiolas 1995].  The 
dopaminergic reward circuitry is also important to post-copulatory behavioral patterns 
such as pair bonding, and variation in peptide receptors in this circuitry appears to be a 
causative agent regulating differences between monogamous and promiscuous mammals 
[Curtis et al. 2006; Young et al. 2011].  While the neurochemical mechanisms of 
copulatory behavior and post-copulation partner preference behavior are well understood 
in the brains of both sexes [Blaustein and Erskine 2002; Hull et al. 2002; Melis and 
Argiolas 1995; Paredes and Agmo 2004; Pfaff et al. 1994; Young et al. 2011], what is far 
 69
less studied is how dopamine pathways (e.g. reward circuitry) is involved in female pre-
copulatory behavioral patterns such as mate choice.  
Sexual behavior can be classically and operationally categorized into appetitive 
(preceding copulation) and consummatory (copulation) behavioral patterns (reviewed in 
[Pfaus 1996]). Appetitive behavior in the context of reproduction can be thought of as 
those behavioral patterns that often initiate encounters with the opposite sex and 
subsequent evaluation of the quality of mates (e.g. mate preference). Since dopaminergic 
neurons show electrophysiological changes in neural activity when an organism is 
anticipating a reward [Schultz 1998; Schultz et al. 1997], appetitive sexual behavioral 
patterns might be predicted to invoke dopamine pathways (e.g. dopaminergic reward 
system). Studies investigating the role of dopamine in female sexual behavioral patterns, 
however, have primarily focused on brain regions involved in consummatory and 
copulatory-related behavioral patterns such as receptivity and self-pacing in rodents 
[Ellingsen and Agmo 2004; Graham and Pfaus 2010; Martinez and Paredes 2001; Melis 
and Argiolas 1995].   
Studies that have examined the role of dopamine pathways in female appetitive 
behavioral patterns (e.g. mate preference) have received far less attention. Dopamine 
manipulation experiments have shown that dopamine can play a modulatory role in 
female preference response in European Starlings [Pawlisch and Riters 2010; Riters et al. 
2007] (Riters et al 2007; Sockman & Salvante 2007; Pawlisch & Riters 2010). The bulk 
of the studies have identified expression patterns of enzymatic precursors (e.g. tyrosine 
hydroxylase, a rate limiting step in catecholamine synthesis [Levitt et al. 1965]) or 
 70
metabolites of dopamine and focused their analysis to brain regions involved in acoustic 
perception (e.g. NCM) or known nodes involved in female receptivity (e.g. VMH) 
[Pawlisch and Riters 2010; Riters et al. 2007; Sockman and Salvante 2008; Svec et al. 
2009].  Here we explore the hypothesis that female mate preference (a proxy for female 
sexual motivation, see below) is modulated by dopamine producing areas in the brain in 
anticipation of copulation.  To this end, we will compare dopamine production (via 
tyrosine hydroxylase mRNA) across brain regions containing dopaminergic neurons 
[Filippi et al. 2010] across social contexts. In addition we examine the expression of a 
gene associated with mate preference (neuroligin-3) across the putative mate preference 
network in the teleost brain during mate choice events.  
Female mate preference is a type of social behavior that involves interacting with 
at least one male. During mate preference interactions, many different genes (e.g. 
neuroserpin, neuroligin-3) are differentially expressed in a female’s brain [Cummings et 
al. 2008]. As females approach and evaluate males, there can be changes in neural 
plasticity through a variety of cellular mechanisms. One example involve regulation of 
neuroligins (cell adhesion proteins), which are critical in synaptogensis and reinforcing 
synaptic connections [Chih et al. 2005; Dean and Dresbach 2006; Levinson et al. 2010; 
Lise and El-Husseini 2006; Tabuchi et al. 2007; Varoqueaux et al. 2006]. Altering the 
role of neuroligins in forming synapses produces notable changes in social behavior in 
rodents and humans [Jamain et al. 2003; Laumonnier et al. 2004; Radyushkin et al. 2009; 
Sudhof 2008; Tabuchi et al. 2007]. For example mutations in neuroligin-3 result in mice 
showing decreased social behavior that is likely associated with an increase in the 
 71
number of inhibitory synapse formation [Tabuchi et al. 2007]. Similarly, for some cases 
of autism in humans, studies have discovered that the subjects had mutations in 
neuroligin genes (including neuroligin-3) [Dean and Dresbach 2006; Garber 2007; 
Jamain et al. 2003; Laumonnier et al. 2004].  Neuroligins’ precise role and localization in 
female mate preference is not currently known. 
Recent research has begun to identify the genomic responses [Cummings et al. 
2008] and brain regions associated [Desjardins et al. 2010; Gentner et al. 2001; Hoke et 
al. 2004; Hoke et al. 2005; Sockman et al. 2002, 2005; Woolley and Doupe 2008] with 
female mate preference behavior—where females discriminate between stimuli prior to 
sex. These studies have observed correlated patterns between preference behavior and 
immediate early gene or genes associated with mate preference in primary sensory 
processing brain regions [Gentner et al. 2001; Hoke et al. 2004; Sockman et al. 2002; 
Woolley and Doupe 2008] as well as multisensory processing areas associated with 
reward [Wong et al. in review] and social behavior [Desjardins et al. 2010].  Here we 
plan to extend the investigation of brain regions involved in mate choice events to regions 
expressing tyrosine hydroxylase to determine if mate preference behavior increases 
dopamine synthesis related to the strength of preference. In addition, we also examine 
whether neuroligin-3, a gene causally linked to social behavior, shows expression 
patterns related to mate preference behavior within brain regions previously implicated in 
mate preference or social behavior. We conduct this investigation with the northern 
swordtail, Xiphophorus nigrensis, as it is ideal for exploring appetitive sexual behavioral 
patterns independent of consummatory sexual behavioral patterns (preference can be 
 72
measured in the lab in a non-contact environment [Ryan and Wagner 1987]). 
Furthermore, we can use this species to explore dopamine pathways (e.g. mesolimbic 
dopaminergic reward circuit) in a system where preference and reproductive status can be 
decoupled (females store sperm in this species and female preference is not predicted by 
her reproductive cycle [Ramsey et al. 2011]).  
Swordtail fish belong to a larger family of internal fertilizing and live-bearing 
fishes, Poecillidae, which are a useful system for studying the evolution of female mate 
choice [Houde 1997; Ryan and Rosenthal 2001]. In X. nigrensis, females prefer the 
genetically influenced large males over the small size males that use forced copulation 
tactics [Ryan et al. 1990]. We have identified candidate genes associated with female 
mate preference (e.g. neuroserpin, neuroligin-3) in the whole brain through microarray 
and quantitative real time PCR analyses that show predictive patterns of expression with 
variation in female preference behavior across 30 minute behavioral trials [Cummings et 
al. 2008]. Subsequent experiments localizing the expression of neuroserpin have 
identified possible brain regions involved in processing mate assessment information 
[Wong et al. in review]. Given these characteristics, X. nigrensis may represent a 
promising system to evaluate mechanisms underlying female mate assessment, including 
the role of dopamine in parts of the mesolimbic reward system of teleosts. 
In the present study we use quantitative in situ hybridization to examine tyrosine 
hydroxylase (TH) and neuroligin-3 expression in adjacent series throughout the brains of 
female X. nigrensis. Since no study has examined the role of TH expression in relation to 
female mate preference contexts in teleosts, we first describe the distribution of TH in the 
 73
swordtail brain. We then relate differences in TH expression to mate preference behavior 
in regions producing dopamine (and not other catecholamines) that are associated with 
sexual behavior, (e.g. some putative homologs to the mesolimbic reward system 
[O'Connell and Hofmann accepted]). As neuroligin-3 expression has been previously 
linked to mate preference behavior in X. nigrensis [Cummings et al. 2008], we use this 
context specific marker to assess its distribution in the putative mate choice circuitry 
[Wong et al. in review] in addition to providing an independent and complementary 
validation of the mate choice network. We predict that neuroligin-3 expression will be 
context-dependent and significantly correlated with female preference behavior in 
putative mate choice network in male exposure environments. Similarly, if a dopamine 
pathway is associated with female mate preference behavior, we predict changes in TH 
expression concomitant with mate preference behavior in dopamine producing regions. 
MATERIAL AND METHODS 
Behavioral Paradigm 
Our study subjects consisted of 48 sexually mature female X. nigrensis that were 
either wild caught or progeny of wild caught individuals maintained in semi-natural 
conditions at UT’s Brackenridge Field Laboratories in Austin, TX. We measured each 
female’s mating preference in a two-way choice experimental design using established 
behavioral measures and natural lighting conditions for this species [Cummings et al. 
2003; Cummings et al. 2008]. The experimental tank was subdivided into three sections. 
Stimuli fish were placed behind UV transparent barriers on each end of the experimental 
tank unless otherwise noted. The center section of the experimental tank was further 
 74
subdivided into three 24 cm regions comprised of two association zones adjacent to the 
barriers and the centermost region termed the neutral zone. The neutral zone contained a 
plastic aquarium plant for refuge. Immediately prior to the behavior trials, we measured a 
proxy for circulating estradiol levels for each female through a non-invasive waterborne 
assay to assess estradiol influences on gene expression (see below).  Females were placed 
into the neutral zone and allowed to acclimate for five minutes behind an opaque barrier. 
Afterwards, the opaque barrier was removed and another one minute passed for further 
acclimatization before we recorded behavioral activity for 30 minutes. During this time 
females were allowed to interact with either stimulus by swimming into the adjacent 
association zones or remain in the neutral zone. Halfway through the trial we switched 
the sides of the stimuli in order to avoid confounding side bias with preference. Females 
were exposed to one of five conditions: two size-matched large males (LL, n = 10) with 
one male behind a UV pass barrier and the other behind a UV blocking filter (UV filters 
were switched half-way between the trial and not the males), one large and small male 
(LS, n = 13), two size-matched small males (SS, n = 7), two size-matched females (FF, n 
= 12), or collected from their home tank (HT, n = 6), a treatment serving as an asocial 
control group (females are housed in isolation).  Home tank females underwent estradiol 
measurements as the other groups but were then returned to their home tanks for 30 
minutes prior to sacrifice. We selected our three male-exposure groups to represent a 
presumed gradient in mate choice complexity with the LL treatment group representing a 
relatively complex mate preference environment (two preferred phenotypes varying in 
presence or absence of UV ornamentation and presumably having the highest incentive 
 75
value). The LS treatment group represented a simple mate preference environment with a 
presumably intermediate level of incentive value whereas the SS treatment group 
presumably representing a minimum mate preference environment and lowest incentive 
value (two males that employ forced copulation strategies only). All size-matched stimuli 
differed by no more than 1 mm standard length (length between tip of snout to caudal 
most part of the caudal peduncle).  For each female we measured the number of glides (a 
proxy for receptivity, which can precede copulatory events [Cummings and Mollaghan 
2006; Liley 1965]), transits to the center (a proxy for general locomotor activity), and 
association bias (measure of preference). Association bias is defined as the amount of 
time spent with stimulus a (where time spent with stimulus a > stimulus b) divided by the 
total amount of time spent in the association zones of both stimuli (this behavioral 
measure can be interpreted as proxy for sexual motivation [Blaustein and Erskine 2002]).   
Estradiol Measurements  
To assess the relationship between estradiol levels and tyrosine hydroxylase and 
neuroligin-3 expression, we quantified a proxy for circulating estradiol levels. We 
measured estradiol levels for all females through a non-invasive waterborne assay 
following an established protocol for teleosts [Earley et al. 2006; Kidd et al. 2010; Scott 
et al. 2008] and validated in our focal species [Ramsey et al. 2011]. Briefly, females were 
placed in 150 mL of reservoir water (treated tap water used for home and experimental 
tank) for one hour prior to behavior trials. Estradiol was extracted from the water using 
C18 Solid Phase Extraction columns (Sep-Pak Plus C18 cartridge 55–105 lm; Waters 
Corporation, Milford, MA) and measured on a Correlate-EIA 17β-estradiol Enzyme 
 76
Immunoassay Kit (Assay Designs) according to manufacturer’s protocol. Hormone 
samples were run on three 96-well EIA assay plates: inter-assay CV was 6.5% and intra-
assay CV was 1.9%.  
Cloning tyrosine hydroxylase 
While there are two tyrosine hydroxylase genes in teleosts (TH1, TH2 
[Yamamoto et al. 2010]), we cloned TH1 (now referred to here as TH) because it is more 
widely expressed across the brain [Filippi et al. 2010; Yamamoto et al. 2010]. To create a 
tyrosine hydroxylase (TH) probe for in situ hybridization (see below), we first cloned a 
fragment of the gene in X. nigrensis using methodology described elsewhere [Cummings 
et al. 2008]. Briefly, after isolating total RNA from whole brain homogenates we 
synthesized cDNA according to manufacturer’s protocol (SuperScript First-Strand 
Synthesis, Invitrogen). Using Codehop (http://blocks.fhcrc.org/codehop.html), we 
generated the degenerate primers 5’-TGGATCTTCAGGGGGTTGTCNARNACYTC-3’ 
and 5’-GGCAGTCCCTGATCGAGGAYGCNMGNAA-3’ to amplify a 1268 base-pair 
TH fragment. Reaction conditions were one denaturing cycle (94C for two minutes) 
followed by 30 amplification cycles (94C denaturing for 30 seconds, 61C annealing for 
90 seconds, and 72C elongation for one minute) and a final 10 minute elongation cycle 
(72C). After amplification, we ligated the fragment into pCR4-TOPO vector and 
transformed into One Shot chemically competent cells according to manufacturer’s 
protocol (TOPO TA Cloning kit for Sequencing with One Shot TOP10 Chemically 
 77
Competent E. Coli, Invitrogen). We subsequently verified sequence identity (University 
of Texas ICMB Core Facilities, Gene Accession #: HM107109.1).  
Tissue Processing 
We cryosectioned female brains at 16 µm thickness onto four serial series. All 
series were post-fixed in cold 4% paraformaldahyde/PBS solution, washed in PBS and 
acetylated in 0.25% acetic anhydride/triethanolamine. Subsequently, slides were washed 
in 2X standard saline citrate, dehydrated in increasing ethanol series and stored at -80 C. 
All slides in a series were processed simultaneously. Using a digoxigenin (DIG) labeled 
riboprobe (see below), one series was used to localize and quantify neuroligin-3 
expression and an adjacent series was used to localize and quantify tyrosine hydroxylase 
expression. The remaining series were used to quantify neuroserpin (with one series used 
for DIG labeled neuroserpin and the other for S35 labeled neuroserpin for validation 
[Wong et al. in review]). 
Probe Synthesis 
DIG labeled neuroligin-3 and TH riboprobes were generated using a 1:3 ratio of 
UTP and DIG-UTP (Roche) following a modified manufacturer’s protocol (Megascript 
T7/T3, Ambion). A 393 base-pair TH DIG probe template was subcloned (Gene 
Accession # HM107109.1) by using primer pair 5’-
TTTGAGGAGGAGGACGGAAAAG-3’and 5’-TCTTCTCTGTCTGTAGGCAGGGTC-
3’. The 345 base pair neuroligin-3 probe template was subcloned by using primer pair 5’-
CCAGATGACATCCCTCTGATGACC-3’and 5’-
GTGCTGTATGGACTCATGTTGGAG-3’on the X. nigrensis neuroligin-3 transcript 
 78
(Gene Accession # DQ835282). After probe synthesis, we removed unincorporated 
nucleotides via column filtration according to manufacturer’s protocol (Megaclear, 
Ambion) and checked probe quality through gel electrophoresis. 
in situ hybridization 
We used an established digoxigenin in situ hybridization protocol [Wong et al. in 
review]. Briefly, we first prehybridized the slides for 6 hours at 60°C in a hybridization 
chamber. Slides were then hybridized overnight (16 hours) at 65°C with fresh 
prehybridization solution containing 0.25 ng of antisense riboprobe per slide in 
hybridization chambers. Following hybridization slides were washed to remove 
nonspecific probe binding and then incubated overnight at 4°C with Anti-Digoxigenin 
AP antibody (Roche). After antibody incubation we blocked endogenous alkaline 
phosphatase activity and used NBT/BCIP stock solution (Roche) for colorimetric 
detection. After stopping colorimetric reactions (two hours for neuroligin-3 and 20 hours 
for TH), we progressively dehydrated sections in ethanol washes and finally coverslipped 
the slides with Permount adhesive (Fisher). Within each series all individuals were 
processed simultaneously to avoid any potential colorimetric development differences 
across individuals due to batch effects.  Please see [Wong et al. in review] for technical 
details. DIG-labeled neuroligin-3 and tyrosine hydroxylase sense riboprobes showed no 
expression. 
Gene expression quantification 
Using a X. helleri brain atlas for reference and terminology [Anken and Rahmann 
1994], we identified and quantified digoxigenin-labeled neuroligin-3 riboprobe 
 79
expression in nine brain regions. The nine nuclei are (teleost nuclei: abbreviation, 
putative tetrapod homolog [Bruce and Bradford 2009; Goodson 2005; Northcutt 2008; 
O'Connell and Hofmann accepted; Rink and Wullimann 2001]): cerebellum (Cb, 
cerebellum), area dorsolateralis telencephali (Dl, pallial hippocampus), area 
dorsomedialis telencephali (Dm, basolateral amygdala), central gray (GC, periaqueductal 
gray), hypothalamus ventralis (HV, ventral hypothalamus), nucleus preopticus (POA, 
preoptic nucleus), nucleus tuberis anterioris (TA, ventromedial hypothalamus), ventralis 
supracommissuralis telencephali (Vs, medial amygdala), and area ventroventralis 
telencephali (Vv, lateral septum). For tyrosine hydroxylase, we identified the expression 
in eight brain regions previously demonstrated to contain TH in other teleosts [Filippi et 
al. 2010; O'Connell et al. 2011a; Parafati et al. 2009; Rink and Wullimann 2001, 2002; 
Yamamoto et al. 2010, 2011]: olfactory bulb (OB, olfactory bulb), area ventroventralis 
telencephalic (Vv, partial homologue nucleus accumbens), area centroventralis 
telencephalic (Vc, striatum), nucleus preopticus (POA, preoptic nucleus), nucleus 
pretectalis periventricularis pars ventralis (PPv, pretectal nucleus), nucleus 
periventricularis tuberculum posterioris (TPp, partial homologue to ventral tegmental 
area), nucleus tuberis posterioris (PTN), and isthmus nucleus (is, locus coeruleus). The 
isthmus nucleus contains a dense population of noradrenergic neurons and is only 
included for completeness [Filippi et al. 2010; Kaslin and Panula 2001; Yamamoto et al. 
1977]. Since we were interested in assessing the role of dopamine and mate assessment, 
we only analyzed TH expression in dopaminergic brain regions previously implicated in 
sexual behavior or are part of the putative teleost mesolimbic reward system (OB, Vv, 
 80
Vc, POA, TPp) [Dufour et al. 2010; O'Connell et al. 2011a; O'Connell and Hofmann 
accepted; Rink and Wullimann 2001; Yamamoto et al. 2011]. While TH is the rate 
limiting step of catecholamine synthesis [Levitt et al. 1965], all brain regions examined 
do not express a marker for noradrenaline (except the isthmus nucleus) but do contain 
dopamine immunoreactive cell bodies in another teleost [Filippi et al. 2010; Yamamoto 
et al. 2011]. Hence it is likely TH expression in all areas measured leads to dopamine 
synthesis and not noradrenaline. After tissue processing final sample sizes for were: LL, 
n = 10; LS, n = 10; SS, n = 5; FF, n = 9; HT, n = 5. 
Digoxigenin quantification 
For both experiments we quantified gene expression by measuring the optical 
density (OD) of the digoxigenin labeled probes, which has been established as a semi-
quantitative measure of gene expression in other systems [Guiot and Rahier 1997; 
Harvey-Girard et al. 2010; Larsson et al. 1991; O'Connell et al. 2011b; Zhang et al. 1995] 
and validated in the focal species [Wong et al. in review]. For each slide, we normalized 
the mean intensity of all measures to the background (mean intensity of slide not 
containing tissue), producing a value for the fractional transmittance of the brain region 
in each section. Fractional transmittance was mathematically converted to optical density 
by the equation OD = 2-log(Fractional Transmittance), which was derived specifically for 
the imaging setup (Nikon Eclipse 80i) in our laboratory using neutral density filters 0, 8 
and 32.  Using NIS Elements image analysis software (Nikon), we measured the OD of 
neuroligin-3 and tyrosine hydroxylase expression across individuals from a standardized 
 81
portion of each brain region (ranging from 1737 - 29152 μm2 depending on size of the 
brain region of interest). 
 Brain section images were captured at 4X with a Nikon 12-bit 2-megapixel 
monochrome camera (DS-2MBWc). For each brain region, we used Nikon NIS Elements 
2.3 to measure a standardized rectangular box (area of box given after each brain region, 
see below) within the borders of each brain region and measured the mean intensity of 
neuroligin-3 or TH expression within the box. Unless otherwise stated, the measuring 
box was always placed in the middle of the brain region on the dorsal-ventral plane. We 
measured the mean intensity in both hemispheres if available and then averaged the 
values for the section. For all brain regions we then proceeded to measure the mean 
intensity for all sections of the individual in which we could identify the desired brain 
region and then averaged for that individual. Depending on the size of the brain region, 
the number of sections averaged per individual ranged from two to eight consecutive 
sections. Consecutive sections spanned 48 μm apart. For neuroligin-3, specifically, we 
measured Dl (19906 μm2) & Dm (19906 μm2) for sections from the rostral most section 
until the disappearance of the anterior commissure; Vv (5877 μm2) & Vs (5877 μm2) for 
sections only containing the telencephalic ventricle; POA (2843 μm2) for sections with 
anterior commissure present until the appearance of the optic tectum; HV (1737 μm2) 
from the pituitary until the lateral extension of the lateral hypothalamus; TA (5339 μm2) 
for 3 sections preceding the lateral extension of the dorsal hypothalamus; Cb (29152 μm2) 
and CG (7298 μm2) for all sections observed (see main text for brain region 
abbreviations). For tyrosine hydroxylase, we measured OB (17983 μm2) for all sections 
 82
observed; Vv (4599 μm2) for all sections only containing the telencephalic ventricle and 
one section preceding the appearance of the ventricle; Vc (4599 μm2) from the 
appearance of the telencephalic ventricle until the disappearance of the anterior 
commissure; POA (4134 μm2, box placed in most medial and ventral part of the region) 
for sections with anterior commissure present until the appearance of the optic tectum; 
PPv (2716 μm2) and TPp (2716 μm2) for all sections containing the posterior 
commissure, PTN (6564 μm2) for all sections observed until the disappearance of the 
torus longitudinalis; is (1486 μm2) for all sections observed. All image capture and data 
collection was performed by a single person (RYW) blind to the treatments. 
Statistics 
All statistics were performed in SPSS (ver. 18) and the network statistics were 
conducted using Ucinet [Borgatti et al. 2002]. To examine group-wide behavioral and 
gene expression differences between individuals expressing high (> median) versus low 
(< median) behavioral patterns within each treatment group, we used a t-test. Due to a 
small sample size, we did not perform any statistical tests involving high and low 
behavioral patterns in the SS group. To account for multiple hypotheses testing, we used 
a Benjamini-Hochberg correction [Benjamini et al. 2001]. To assess relationships 
between individual variation of preference behavior, gene expression, or estradiol levels 
we used Pearson’s correlation when the data was normal and Spearman correlation when 
data was non-normal (glides and transits). 
 We characterize neuroligin-3 and TH network expression patterns by looking at 
pairwise correlations of gene expression between brain regions in specific social contexts. 
 83
To analyze network patterns, we converted all Benjamini-Hochberg corrected significant 
correlations of neuroligin-3 and TH expression between brain regions into binary values 
in an association matrix. Significant relationships had a value of 1 while non-significant 
values were designated 0. We compared densities of the networks across conditions using 
Ucinet. For each mate preference group (LL, LS, SS), we define unique correlations as 
those that are not also significant in the FF or HT groups.  
RESULTS 
Behavior: As reported elsewhere [Wong et al. in review], females exposed to the LS 
treatment condition showed a clear preference for the large male by spending a 
significantly greater amount of time in the association zone next to the large male relative 
to the small male (t = 2.1, p = 1.1 * 10-7). Females showed no preference for a stimulus in 
the other test groups: two females, (FF, t = 0.84, p = 0.4) two small males, (SS, t = 0.12, p 
= 0.9), two large males (LL, t = 0.75, p = 0.45). There were no significant differences in 
the number of glides or transits displayed across treatment groups (glides: χ2 = 1.36, p = 
0.71; transits: χ2 = 4.69, p = 0.19).  
Tyrosine Hydroxylase expression: Across all females, TH in the forebrain was expressed 
in the olfactory bulb, the ventral and caudal portion of the ventral telencephalon and the 
preoptic area (Figure 3.1). There were no significant differences in expression of TH OD 
amongst any forebrain regions examined (Figure 3.2).  Within the midbrain, we observed 
TH expression in the PPv, TPp, and PTN across all females (Figure 3.1). TH OD in the 
PTN was significantly lower than in the PPv and TPp (Figure 3.2). The brain regions 
within the midbrain had a significantly higher TH expression than those in the forebrain 
 84
(Figure 3.2). Within the hindbrain, we only observed TH expression in the isthmus 
nucleus, which had the highest TH OD for the brain regions measured (Figure 3.2).  
When examining the relationship between TH expression and social contexts 
there were no significant differences in TH optical density in the OB, Vv, Vc, POA or 
TPp across treatment groups (F > 0.1 and p > 0.1 for all brain regions, Figure 3.3, Table 
3.1). Looking within each treatment group, we also did not observe any differences in TH 
OD between high preference (> median) females and low preference females (< median) 
in any context (Figure 3.4, Table 3.1).  Within the female exposed females (FF), we 
found that females displaying lower than the median level of glides had a significantly 
higher TH OD in the Vc and POA than females expressing high glides (Table 3.1). The 
difference in Vc, however, is not significant after a multiple hypothesis correction test. 
We did not observe any significant differences in TH OD in any brain regions between 
high and low glides or transits for the other treatment groups (Table 3.1). Of the two 
brain regions showing within group level differences of TH OD for glides and transits in 
FF females, only the POA showed a significant correlation between glides and TH OD (r 
= - 0.80, p = 0.008). Furthermore there were no significant correlations between the 
proxy for circulating estradiol levels and association bias, glides, transits, or TH OD in 
any brain region and context (Table 3.2). 
Neuroligin-3 expression: Of all the treatment groups tested, only the females exposed to 
the LS environment showed differential expression in neuroligin-3 OD that was matched 
by differences in association bias (Table 3.3). Specifically, neuroligin-3 OD was 
significantly higher in Dl, Dm, POA, Vv, and HV for females showing a high preference 
 85
relative to low preference females after multiple hypothesis correction (Figure 3.5, Table 
3.3). When looking at transits, LS females had significantly higher neuroligin-3 OD in 
Cb and TA for high transit compared to low transit females but these were not significant 
after a multiple hypothesis correction (Table 3.4). We did not observe any significant 
differences in neuroligin-3 OD between high glides or transit and low glides or transit 
females in any other brain region or context after correcting for multiple hypothesis 
testing (Table 3.4 & 3.5). Of the five brain regions that showed within group differences 
in preference for LS females, all of them also showed a significant and positive 
correlation between neuroligin-3 OD and association bias (Dl, r = 0.65, p = 0.04; Dm, r = 
0.68, p = 0.02; HV, r = 0.82, p = 0.003; POA, r = 0.90, p = 0.0005; Vv, r = 0.8, p = 0.004; 
Figure 3.6). Interestingly, when looking at these specific brain regions in other social 
treatments, a significant positive correlation between neuroligin-3 OD and association 
bias was only observed in females exposed to two small males (SS) in Dl (r = 0.88, p = 
0.047), Hv (r = 0.96, p = 0.007) and Vv (r = 0.95, p = 0.01). Estradiol levels measured 
from the water did not correlate with neuroligin-3 expression in any brain region and 
context (Table 3.2). 
Tyrosine hydroxylase and neuroligin-3 expression networks: To examine the neuroligin-3 
and tyrosine hydroxylase network expression patterns across the brain regions in each 
social context, we used significant correlations that were corrected for multiple 
hypothesis testing. Focusing on the unique neuroligin-3 correlations within each male 
preference group (LL, LS, or SS) we did not find any connections that are consistent 
across all three groups. However, all of the significant correlations seen in the LS 
 86
treatment group are also seen in LL-exposed females (Figure 3.7). Furthermore, there is a 
significantly higher density (i.e. total number of significant correlations between brain 
regions) for females exposed to LL relative to LS (t = 2.27, p =0.013) and SS (t =5.65, p 
= 0.0002) groups. Neuroligin-3 network density for females exposed to LS is 
significantly higher than SS (t = 3.9, p = 0.0002). In contrast, TH expression showed very 
little unique coordinated expression pattern across the brain in any context (Figure 3.8). 
Of particular note, after correcting for multiple hypothesis testing, we did not observe any 
significant correlations of TH expression between brain regions in any context with the 
exception of a significant positive correlation of TH expression between Vv and OB in 
the LS environment. 
DISCUSSION 
Dopamine is widely known to influence decision making, incentive valuation, and 
reinforcement learning processes [Curtis et al. 2006; Hull et al. 1999; Melis and Argiolas 
1995; Schultz 1998, 2007a, b; Schultz et al. 1997; Wise 2002]. In this study we do not 
see any significant changes in dopamine production (via tyrosine hydroxylase) across 
dopamine-producing brain regions in mate preference contexts or across social and 
asocial contexts within 30 minute behavioral trials; yet we see significant changes in a 
gene associated with mate preference (neuroligin-3) within this same timeframe. 
Although females displayed a clear preference/sexual motivation for the large male in the 
LS treatment group, high preference and low preference females showed no significant 
difference in TH expression in any of the measured TH expressing brain regions (Figure 
3.4, Table 3.1). Furthermore, TH expression didn’t vary by social context despite 
 87
environments with presumably different incentive value (two large (LL), one large (LS), 
and no large male (SS) environments, Figure 3.3). TH level, as measured by quantitative 
real-time PCR under a similar behavioral paradigm, also does not significantly vary with 
behavior or context in the whole brain (Ramsey & Cummings, unpublished). Taken 
together, we did not find evidence for rapid changes (within 30 minutes) in dopamine 
production in anticipation of potential copulation or to variation in motivation/signal 
valuation of mate preference in this species.  
 Our results with TH expression are consistent with studies in female European 
starlings, which have also reported no changes in dopamine levels (as measured by 
immunocytochemistry or HPLC) in response to different quality male songs in the 
putative mammalian homologs of the striatum, nucleus accumbens, or ventral tegmental 
area [Riters et al. 2007; Svec et al. 2009]. Studies, however, have found significant 
changes as measured by HPLC in the lateral septum, ventromedial hypothalamus [Riters 
et al. 2007] as well as an acoustic processing center, caudomedial nidopallum [Sockman 
and Salvante 2008]. Pharmacological manipulations of dopamine in female European 
starlings also did not produce changes in behavioral proxies for motivation (time spent 
next to or looking at nest box) when hearing attractive calls and less attractive conspecific 
calls [Pawlisch and Riters 2010]. This suggests that the role of dopamine pathways (e.g. 
mesolimbic reward system) may not have a significant influence or be influenced by 
female mate assessment behavior in X. nigrensis and possibly other species. We cannot 
rule out the possibility that since X. nigrensis females have the ability to store enough 
sperm for multiple broods, catecholamine mechanisms underlying female sexual 
 88
motivation and/or incentive valuation may be different in these females. Furthermore, we 
measured mRNA expression level changes within 30 minutes of stimulus presentation 
and dopamine release may have occurred on a different time scale. Of note, we observed 
TH expression in most of the brain regions previously described to contain TH in several 
other teleosts [Filippi et al. 2010; O'Connell et al. 2011a; Parafati et al. 2009; Yamamoto 
et al. 2010, 2011]. Although other studies have observed cell bodies expressing TH in the 
teleost ventral thalamus, suprachiasmatic nucleus and various subregions of the 
hypothalamus [Filippi et al. 2010; O'Connell et al. 2011a; Yamamoto et al. 2010, 2011], 
in our study these regions may not synthesize TH mRNA with just a 30 minute exposure 
to behaviorally relevant stimuli. 
While we found no support for changes in dopamine synthesis co-occurring with 
preference behavior across female swordtails, we cannot rule out the possibility that 
preference was modulated by changes in dopamine release or receptor activity. Many 
studies have shown that specific dopamine receptor agonists and antagonists have 
important roles in sexual behavior for both sexes of rodents, male quails and lizards 
[Balthazart et al. 1997; Graham and Pfaus 2010; Kleitz-Nelson et al. 2010; Kleitz et al. 
2009; Melis and Argiolas 1995; Woolley et al. 2001]. In female rodents, D2 receptors in 
the medial POA appear to influence select appetitive sexual behavioral patterns and the 
ratio of D1/D2 receptor may underlie behavioral transitions to copulation [Graham and 
Pfaus 2010]. Long-term partner preference formation in monogamous voles is also 
mediated by dopamine binding to D2 receptors and maintained by D1 receptors [Curtis et 
al. 2006; Young et al. 2011]. Hence, differences in receptor subtype densities or binding 
 89
efficiency may allow for the dopamine pathways to influence mate preference. In birds, 
receptor densities in specific brain regions can explain individual differences in male 
motivation to sing courtship songs [Heimovics et al. 2009; Heimovics and Riters 2008; 
Riters 2010, 2011]. Within female swordtails it is therefore plausible that there may be 
differential expression of particular dopamine receptor subtypes in candidate brain 
regions where dopamine can bind and influence mate preference. All of the brain regions 
showing differential neurologin-3 expression (Dm, Dl, HV, POA, Vv) express dopamine 
receptors in another teleost [O'Connell et al. 2011a].  Changing the ratio or densities of 
dopamine receptor subtypes may allow for females to respond to varying quality of males 
in the mate preference context while showing no differences in dopamine production. 
Future studies should explore roles of receptor subtypes in mate assessment.  
 Within 30 minutes of stimulus presentation, we saw evidence that the brain 
exhibits context dependent differences in neuroligin-3 expression. In the LS treatment 
group, high preference females had a significantly higher neuroligin-3 expression in Dm, 
Dl, HV and POA relative to females expressing a low preference for a particular male. 
Furthermore, neuroligin-3 expression in all of these regions showed a significant positive 
correlation with preference (Figure 3.6). These results are consistent with previous 
analyses of another gene associated with mate preference, neuroserpin, which showed 
significant covariation with female preference behavior in the same four brain regions 
[Wong et al. in review]. Two of these brain regions (Dm and Dl) integrate multisensory 
information [Northcutt 2008], are part of the putative mesolimbic reward pathway in 
teleosts and may receive signals from dopaminergic neurons that project from TPp or Vc 
 90
[O'Connell and Hofmann accepted; Rink and Wullimann 2001]. Dm and Dl project to 
areas involved in receptivity and/or copulation (POA) or indirectly to HV in the 
hypothalamus. The POA is a node in the Social Behavior Network (SBN), which is a set 
of brain regions implicated in a variety of sexual and social behavior [Goodson 2005; 
Newman 1999]. Neuroligin-3 showed significant differences in expression between high 
and low preference LS-exposed females in another SBN node, Vv. This region, Vv, 
which has been previously linked to sexual behavior in rodents as well as female mate 
assessment in cichlids [Desjardins et al. 2010; Newman 1999], receives projections from 
Dm and Dl [Northcutt 2006]. Hence our current model of the female mate preference 
network suggests that mate assessment information is processed in multisensory 
integration areas (Dm, Dl) and then projected to areas more directly involved in 
receptivity and copulation (HV or nodes in the SBN such as POA and Vv). 
Comparisons of gene expression networks across the entire brain during mate 
preference contexts revealed that tyrosine hydroxylase (Figure 3.8) and neuroligin-3 
(Figure 3.7) have dramatically different responses. TH showed almost no coordinated 
response across brain regions, yet neuroligin-3 expression network complexity increases 
with presumably more complex mate preference environments; there are greater number 
of brain regions with correlated neuroligin-3 expression in the LL and LS environment 
compared to SS. This pattern of differential network complexity mirroring the presumed 
complexity of the mate choice environment is parallel to that observed with neuroserpin 
expression across these females [Wong et al. in review]. The overlapping expression 
patterns we observed from these two genes (neuroligin-3 and neuroserpin) suggest that 
 91
there may be different neural responses depending on mate preference contexts. That is, 
if two males are nearly equally attractive (LL), a female’s brain may require more cross-
talk across the brain than when given a simpler choice (LS) or non-attractive options 
(SS). Differences in connectivity among brain regions depending on mate choice contexts 
are also seen in female tungara frogs [Hoke et al. 2005]. We did not, however, observe 
any coordinated unique TH expression patterns across mate preference contexts. 
Although we did observe that midbrain dopaminergic centers had higher TH expression 
than forebrain areas across all females, this is not surprising as the major ascending 
dopaminergic sites are located in the midbrain [O'Connell et al. 2011a; O'Connell and 
Hofmann accepted; Rink and Wullimann 2001]. The lack of any coordinated patterns of 
TH expression across the brain could also indirectly support the interpretation that 
dopamine synthesis in the dopamine pathways does not play a major part in mate 
preference in this species. 
 Interestingly, there are some major differences between the neuroserpin [Wong et 
al. in review] and neuroligin-3 (Figure 3.7) unique expression networks. First there are 
many more unique connections in the LL and LS contexts in neuroligin-3 expression 
networks relative to neuroserpin. Secondly, neuroligin-3 network density in LS is 
significantly higher than SS, which was not the case for neuroserpin.  Differences in 
social behavior associated with neuroligin-3 and neuroserpin, which can be related to 
differences in cellular functions, may explain why we observed differences in 
coordinated expression patterns across the brain within the same females. Neuroserpin 
expression appears important for rodents to exhibit exploratory behavior and decreased 
 92
neophobia [Madani et al. 2003]. Neuroligin-3 expression does not seem to influence 
exploratory behavior but appears to regulate willingness to be social in rodents [Tabuchi 
et al. 2007] in addition to sensory perception [Biswas et al. 2010; Radyushkin et al. 
2009]. Sensory-deprived honeybees (held in constant darkness) show significant 
differences in whole-brain expression of neuroligins compared to bees held under natural 
conditions [Biswas et al. 2010]. Furthermore, neuroligin-3 knockout mice show 
abnormalities in social memory that may be linked to olfactory deficiency [Radyushkin et 
al. 2009].  As female mate preference requires social interaction and sensory perception, 
it may not be surprising that neuroligin-3 expression showed a much greater coordinated 
expression pattern across the brain relative to neuroserpin.  
The increasing density of the neuroligin-3 expression network concomitant with 
complexity of the mate preference environment may be related to perceptual differences 
in sensory information (e.g. size of male and ornamentation). In the two large male 
environment (LL), females assess both attractive males that phenotypically differed only 
in the presence or absence of UV ornamentation; in the LS environment, females are 
given a simple preference environment with just one attractive option and no attractive 
options in the SS environment. Having increasing sensory complexity (LL > LS > SS) 
may require more coordinated expression of neuroligin-3 across the brain to process this 
information. For neuroserpin, if given two large males, her willingness to explore both 
attractive large males may be one reason why we see the highest network density in the 
LL environment relative to LS and SS [Wong et al. in review]. It is an intriguing 
possibility that neuroserpin and neuroligin-3 are serving complementary roles in 
 93
facilitating females to encounter males and subsequently evaluate his attractiveness. 
Future studies should determine how the amount of co-localization of neuroserpin and 
neuroligin-3 relates to female mate preference behavior. 
SUMMARY 
There are a couple of proposed networks of brain regions (SBN, mesolimbic reward 
circuitry) that can contribute to natural behavioral displays across vertebrate taxa 
[Newman 1999; Wise 2002]. In another study using neuroserpin [Wong et al. in review] 
and confirmed in the current study (neuroligin-3) we identified a set of brain regions 
associated with female mate preference behavior that are part of either the SBN or 
mesolimbic reward circuitry. However, by examining tyrosine hydroxylase expression, 
we did not detect any differences in dopamine production with a 30 minute stimulus 
exposure period. Variation in female mate preference behavior (e.g. sexual motivation) 
does not seem to lead to correlated changes in dopamine synthesis in the dopaminergic 
regions in preparation of copulation. While there are no rapid changes in dopamine 
production across contexts, one potentially fruitful avenue of research would be to 
examine how differences in the densities of individual dopamine receptors or the ratio of 
receptors within Dm, Dl, HV, Vv and POA differ across females displaying different 
levels of mate preference. 
 
 
 
 
 94
Figure 3.1: Distribution of tyrosine hydroxylase across all female X. nigrensis tested. 
Top left panel represents a sagittal view of the brain with each line and letter representing 
sections in the remaining panels. Left hemisphere in each panel show additional select 
brain regions not measured but are there for illustrative and neuroanatomical orientation 
purposes. Black dots on the right hemisphere indicates location of TH expression. AC, 
anterior commissure; Dc, area centralis telencephalic; Hc, hypothalamus caudalis; HD, 
hypothalamus dorsalis; NDIL, nucleus diffuses lobi inferioris; NGp, nucleus 
glomerulosus posterioris; OT, optic tectum; PC, posterior commissure; PGm, nucleus 
preglomerulosus medialis; Tl, torus longitudinalis; VC, valvula cerebilli; Vd, area 
dorsoventralis telencephalic. 
 
 
 95
Figure 3.2: Distribution of tyrosine hydroxylase in brains of female X. nigrensis.  
Overall the TH expression in the brain regions in the mid- and hindbrain are significantly 
higher than the forebrain for all females (n=39). b, p < 0.0001; c, p < 0.0001; d, p < 
0.001. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 96
Figure 3.3: Tyrosine hydroxylase expression in select brain regions across treatment 
groups.  
Blue, red, green, purple, and teal represents LL, LS, SS, FF and HT, respectively. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 97
Figure 3.4: Tyrosine Hydroxylase expression in females exposed to a large vs. small 
male (LS) context. 
Differences in TH expression between groups of high association bias (> median) 
females (n=5, blue) and low association bias (< median) females (n=5, red) for LS 
females. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 98
Figure 3.5: Neuroligin-3 expression in females exposed to a large vs. small male (LS) 
context.  
Differences in neuroligin-3 expression between groups of high association bias (> 
median) females (n=5, blue) and low association bias (< median) females (n=5, red) for 
LS females. *, p < 0.05, **, p < 0.01 
 
 
 
 
 
 
 
 
 
 
 
 
 
 99
Figure 3.6: Individual variation of association bias and neuroligin-3 expression in LS 
exposed females.  
Significant correlations between individual variation in association bias and neuroligin-3 
expression in A) Dl, B) Dm, C) HV, D) POA, and E) Vv. 
 
 
 
 
 
 
 
 
 
 100
Figure 3.7: Neuroligin-3 expression network by context. 
Unique significant positive pairwise correlations relative to FF and HT females in 
neuroligin-3 expression between brain regions (lines) in A) LL, B) LS, and C) SS 
exposed females. Brain regions in bold circles are those associated with mate preference 
identified in this study.  
 
 101
Figure 3.8: Tyrosine hydroxylase expression network by context. 
Unique significant positive pairwise correlations relative to FF and HT females in TH 
expression between brain regions (lines) in A) LL, B) LS, and C) SS exposed females.  
 
 102
Table 3.1: Tyrosine hydroxylase optical density (mean ± SE) comparisons between 
“high” (> median) and “low” (< median) behavioral patterns. 
** indicates significance after correcting for multiple hypotheses; * indicates significance 
that does not survive multiple hypothesis testing; n.s., not significant 
 
Brain 
Region 
LL LS FF 
High Low p-value High Low p-value High Low p-value 
Association Bias 
OB 0.02 ±  
0.003 
0.019 ±  
0.002 
n.s. 
(0.88) 
0.023 ± 
0.003 
0.018 ± 
0.0008 
n.s. 
(0.19) 
0.018 ±  
0.001 
0.022 ±  
0.004 
n.s. 
(0.45) 
Vv 0.01 ± 
0.004 
0.011 ± 
0.004 
n.s. 
(0.89) 
0.019 ± 
0.009 
0.006 ± 
0.004 
n.s. 
(0.27) 
0.004 ±  
0.004 
0.009 ±  
0.005 
n.s. 
(0.46) 
Vc 0.02 ± 
0.002 
0.019 ± 
0.001 
n.s. 
(0.76) 
0.02 ± 
0.003 
0.018 ± 
0.002 
n.s. 
(0.7) 
0.02 ±  
0.002 
0.017 ±  
0.002 
n.s. 
(0.41) 
POA 0.021 ± 
0.001 
0.021 ± 
0.001 
n.s. 
(0.94) 
0.02 ± 
0.002 
0.018 ± 
0.002 
n.s. 
(0.52) 
0.024 ±  
0.002 
0.019 ±  
0.002 
n.s. 
(0.29) 
TPp 0.048 ± 
0.006 
0.052 ± 
0.007 
n.s. 
(0.74) 
0.072 ± 
0.01 
0.054 ± 
0.005 
n.s. 
(0.18) 
0.08 ±  
0.019 
0.05 ±  
0.005 
n.s. 
(0.17) 
Glides 
OB 0.019 ± 
0.002 
0.019 ± 
0.003 
n.s. 
(0.98) 
0.017 ± 
0.001 
0.024 ± 
0.003 
n.s. 
(0.09) 
0.016 ± 
0.001 
0.019 ± 
0.001 
n.s. 
(0.18) 
Vv 0.01 ± 
0.004 
0.01 ± 
0.004 
n.s. 
(0.85) 
0.003 ± 
0.003 
0.023 ± 
0.008 
n.s. 
(0.05) 
0.01 ± 
0.004 
0.005 ± 
0.005 
n.s. 
(0.44) 
Vc 0.019 ± 
0.001 
0.02 ± 
0.002 
n.s. 
(0.68) 
0.185 ± 
0.002 
0.02 ± 
0.003 
n.s. 
(0.61) 
0.014 ± 
0.001 
0.021 ± 
0.002 
* 
(0.049) 
POA 0.019 ± 
0.001 
0.023 ± 
0.001 
n.s. 
(0.05) 
0.019 ± 
0.001 
0.019 ± 
0.002 
n.s. 
(0.85) 
0.015 ± 
0.001 
0.025 ± 
0.001 
** 
(0.007) 
TPp 0.045 ± 
0.005 
0.055 ± 
0.007 
n.s. 
(0.34) 
0.065 ± 
0.012 
0.061 ± 
0.004 
n.s. 
(0.75) 
0.058 ± 
0.01 
0.083 ± 
0.016 
n.s. 
(0.24) 
Transits 
OB 0.02 ± 
0.003 
0.019 ± 
0.0008 
n.s. 
(0.83) 
0.02 ± 
0.002 
0.021 ± 
0.002 
n.s. 
(0.92) 
0.01 ± 
0.001 
0.02 ± 
0.003 
n.s. 
(0.14) 
Vv 0.01 ± 
0.004 
0.01 ± 
0.004 
n.s. 
(0.92) 
0.006 ± 
0.006 
0.019 ± 
0.008 
n.s. 
(0.22) 
0.01 ± 
0.004 
0.005 ± 
0.005 
n.s. 
(0.44) 
Vc 0.02 ± 
0.002 
0.02 ± 
0.001 
n.s. 
(0.92) 
0.018 ± 
0.002 
0.02 ± 
0.003 
n.s. 
(0.76) 
0.014 ± 
0.001 
0.02 ± 
0.001 
n.s. 
(0.05) 
POA 0.021 ± 
0.001 
0.021 ± 
0.001 
n.s. 
(0.77) 
0.018 ± 
0.002 
0.02 ± 
0.002 
n.s. 
(0.59) 
0.015 ± 
0.001 
0.025 ± 
0.001 
n.s. 
(0.08) 
TPp 0.043 ± 
0.004 
0.057 ± 
0.007 
n.s. 
(0.14) 
0.071 ± 
0.01 
0.054 ± 
0.006 
n.s. 
(0.2) 
0.058 ± 
0.01 
0.06 ± 
0.01 
n.s. 
(0.86) 
 
 
 103
Table 3.2: Correlation between a proxy for circulating estradiol levels and association 
bias, glides, transits, and gene (neuroligin-3 and tyrosine hydroxylase) 
expression in different brain regions. 
 
 
 
 104
Table 3.3: Neuroligin-3 optical density (mean ± SE) comparisons between “high” (> 
median) and “low” (< median) association bias.  
** indicates significance after correcting for multiple hypotheses; * indicates significance 
that does not survive multiple hypothesis testing; n.s., not significant 
 
 
Brain 
Region 
LL LS FF 
High Low p-value High Low p-value High Low p-value 
Dm 0.028 ± 
0.009 
0.014 ± 
0.003 
n.s. 
(0.19) 
0.053 ± 
0.012 
0.011 ± 
0.001 
** 
(0.009) 
0.038 ± 
0.014 
0.027 ± 
0.013 
n.s. 
(0.58) 
Dl 0.026 ± 
0.01 
0.012 ± 
0.002 
n.s. 
(0.23) 
0.033 ± 
0.006 
0.012 ± 
0.002 
** 
(0.01) 
0.026 ± 
0.009 
0.024 ± 
0.011 
n.s. 
(0.87) 
Cb 0.081 ± 
0.025 
0.035 ± 
0.014 
n.s. 
(0.16) 
0.087 ± 
0.028 
0.043 ± 
0.014 
n.s. 
(0.21) 
0.06 ± 
0.024 
0.093 ± 
0.036 
n.s. 
(0.48) 
GC 0.008 ± 
0.001 
0.004 ± 
0.0004 
n.s. 
(0.11) 
0.007± 
0.001 
0.004 ± 
0.0008 
n.s. 
(0.19) 
0.005 ± 
0.0008 
0.01 ± 
0.003 
n.s. 
(0.19) 
POA 0.104 ± 
0.056 
0.03 ± 
0.004 
n.s. 
(0.22) 
0.165 ± 
0.036 
0.052 ± 
0.014 
** 
(0.023) 
0.142 ± 
0.065 
0.101 ± 
0.029 
n.s. 
(0.58) 
TA 0.078 ± 
0.036 
0.026 ± 
0.011 
n.s. 
(0.21) 
0.085 ± 
0.014 
0.047 ± 
0.009 
n.s. 
(0.05) 
0.09 ± 
0.038 
0.096 ± 
0.028 
n.s. 
(0.91) 
VH 0.127 ± 
0.06 
0.054 ± 
0.019 
n.s. 
(0.28) 
0.213 ± 
0.03 
0.069 ± 
0.015 
** 
(0.002) 
0.174 ± 
0.065 
0.222 ± 
0.084 
n.s. 
(0.66) 
Vs 0.106 ± 
0.042 
0.04 ± 
0.012 
n.s. 
(0.16) 
0.142 ± 
0.04 
0.055 ± 
0.011 
n.s. 
(0.07) 
0.093 ± 
0.046 
0.061 ± 
0.023 
n.s. 
(0.56) 
Vv 0.106 ± 
0.042 
0.034 ± 
0.014 
n.s. 
(0.14) 
0.114 ± 
0.016 
0.041 ± 
0.01 
** 
(0.005) 
0.1 ± 
0.049 
0.064 ± 
0.028 
n.s. 
(0.55) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 105
Table 3.4: Neuroligin-3 optical density (mean ± SE) comparisons between “high” (> 
median) and “low” (< median) transits. 
** indicates significance after correcting for multiple hypotheses; * indicates significance 
that does not survive multiple hypothesis testing; n.s., not significant 
 
Brain 
Region 
LL LS FF 
High Low p-value High Low p-value High Low p-value 
Dm 0.022 ± 
0.01 
0.02 ± 
0.003 
n.s. 
(0.8) 
0.042 ± 
0.015 
0.022 ± 
0.008 
n.s. 
(0.31) 
0.024 ± 
0.014 
0.028 ± 
0.012 
n.s. 
(0.84) 
Dl 0.024 ± 
0.011 
0.014 ± 
0.002 
n.s. 
(0.45) 
0.03 ± 
0.007 
0.016 ± 
0.004 
n.s. 
(0.14) 
0.019 ± 
0.009 
0.024 ± 
0.011 
n.s. 
(0.74) 
Cb 0.083 ± 
0.027 
0.033 ± 
0.009 
n.s. 
(0.11) 
0.098 ± 
0.025  
0.033 ± 
0.01 
* 
 (0.04) 
0.082 ± 
0.035 
0.059 ± 
0.031 
n.s. 
(0.63) 
GC 0.006 ± 
0.001 
0.006 ± 
0.002 
n.s. 
(0.98) 
0.006 ± 
0.002 
0.004± 
0.0005 
n.s. 
(0.3) 
0.009 ± 
0.003 
0.006 ± 
0.001 
n.s. 
(0.48) 
POA 0.098 ± 
0.058 
0.037 ± 
0.005 
n.s. 
(0.32) 
0.13 ± 
0.039 
0.081 ± 
0.033 
n.s. 
(0.39) 
0.104 ± 
0.054 
0.089 ± 
0.034 
n.s. 
(0.82) 
TA 0.068 ± 
0.039 
0.036 ± 
0.011 
n.s. 
(0.46) 
0.09 ± 
0.01 
0.043 ± 
0.008 
* 
(0.009) 
0.1 ± 
0.034 
0.058 ± 
0.024 
n.s. 
(0.35) 
VH 0.114 ± 
0.064 
0.067 ± 
0.019 
n.s. 
(0.5) 
0.18 ± 
0.045  
0.102 ± 
0.028 
n.s. 
(0.18) 
0.197 ± 
0.08  
0.197 ± 
0.08 
n.s. 
(0.67) 
Vs 0.082 ± 
0.044 
0.063 ± 
0.02 
n.s. 
(0.71) 
0.0135 
± 0.042 
0.062 ± 
0.015 
n.s. 
(0.14) 
0.055 ± 
0.028 
0.058 ± 
0.024 
n.s. 
(0.94) 
Vv 0.079 ± 
0.045 
0.061 ± 
0.022 
n.s. 
(0.73) 
0.101 ± 
0.022 
0.055 ± 
0.016 
n.s. 
(0.13) 
0.06 ± 
0.041 
0.061 ± 
0.029 
n.s. 
(0.97) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 106
Table 3.5: Neuroligin-3 optical density (mean ± SE) comparisons between “high” (> 
median) and “low” (< median) glides. 
** indicates significance after correcting for multiple hypotheses; * indicates significance 
that does not survive multiple hypothesis testing; n.s., not significant 
 
Brain 
Region 
LL LS FF 
High Low p-value High Low p-value High Low p-value 
Dm 0.012 ± 
0.001 
0.031 ± 
0.085 
n.s. 
(0.05) 
0.024 ± 
0.009 
0.041 ± 
0.015 
n.s. 
(0.37) 
0.024 ± 
0.014 
0.026 ± 
0.01 
n.s. 
(0.92) 
Dl 0.01 ± 
0.0003 
0.028 ± 
0.01 
n.s. 
(0.1) 
0.018 ± 
0.005 
0.027 ± 
0.008 
n.s. 
(0.38) 
0.019 ± 
0.009 
0.019 ± 
0.006 
n.s. 
(0.99) 
Cb 0.042 ± 
0.014 
0.074 ± 
0.028 
n.s. 
(0.33) 
0.081 ± 
0.03 
0.05 ± 
0.014 
n.s. 
(0.38) 
0.082 ± 
0.035  
0.038 ± 
0.011 
n.s. 
(0.28) 
GC 0.005 ± 
0.0008 
0.007 ± 
0.002 
n.s. 
(0.44) 
0.006 ± 
0.002 
0.005 ± 
0.0007 
n.s. 
(0.57) 
0.009 ± 
0.003 
0.004 ± 
0.0004 
n.s. 
(0.24) 
POA  0.028 ± 
0.004 
0.106 ± 
0.055 
n.s. 
(0.19) 
0.111 ± 
0.04 
0.092 ± 
0.034 
n.s. 
(0.75) 
0.104 ± 
0.054 
0.116 ± 
0.053 
n.s. 
(0.88) 
TA 0.026 ± 
0.011 
0.077 ± 
0.036 
n.s. 
(0.22) 
0.066 ± 
0.017 
0.066 ± 
0.012 
n.s. 
(0.98) 
0.1 ± 
0.034 
0.065 ± 
0.03 
n.s. 
(0.46) 
VH 0.054 ± 
0.019 
0.125 ± 
0.061 
n.s. 
(0.31) 
0.151 ± 
0.053 
0.131 ± 
0.028 
n.s. 
(0.75) 
0.197 ± 
0.08 
0.132 ± 
0.052 
n.s. 
(0.52) 
Vs 0.028 ± 
0.003 
0.117 ± 
0.038 
n.s. 
(0.05) 
0.073 ± 
0.022 
0.124 ± 
0.043 
n.s. 
(0.33) 
0.055 ± 
0.028 
0.082 ± 
0.043 
n.s. 
(0.63) 
Vv 0.022 ± 
0.003 
0.118 ± 
0.039 
* 
(0.04) 
0.072 ± 
0.028 
0.084 ± 
0.014 
n.s. 
(0.72) 
0.06 ± 
0.04 
0.086 ± 
0.043 
n.s. 
(0.68) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 107
References 
Adkins-Regan E (2011) Neuroendocrine contributions to sexual partner preference in 
birds. Front Neuroendocrinol. 
Ambroggi F, Ishikawa A, Fields HL, Nicola SM (2008) Basolateral amygdala neurons 
facilitate reward-seeking behavior by exciting nucleus accumbens neurons. 
Neuron 59:648-661. 
Andersson MB (1994) Sexual selection: Princeton University Press. 
Anken RH, Rahmann H (1994) Brain atlas of the adult swordtail fish: Xiphophorus 
helleri and of certain developmental stages: Gustav Fischer. 
Bailey NW, Zuk M (2008) Acoustic experience shapes female mate choice in field 
crickets. Proceedings of the Royal Society B: Biological Sciences 275:2645-2650. 
Bailey NW, Zuk M (2009) Field crickets change mating preferences using remembered 
social information. Biology Letters 5:449-451. 
Ball GF, Balthazart J (2010) Sexual arousal, is it for mammals only? Horm Behav. 
Balthazart J, Castagna C, Ball GF (1997) Differential effects of d1 and d2 dopamine-
receptor agonists and antagonists on appetitive and consummatory aspects of male 
sexual behavior in japanese quail. Physiol Behav 62:571-580. 
Barron AB, Sovik E, Cornish JL (2010) The roles of dopamine and related compounds in 
reward-seeking behavior across animal phyla. Front Behav Neurosci 4:163. 
Basolo AL (1990) Female preference predates the evolution of the sword in swordtail 
fish. Science 250:808-810. 
Basolo AL (1995) Phylogenetic evidence for the role of a pre-existing bias in sexual 
selection. Proc Biol Sci 259:307-311. 
Benjamini Y, Drai D, Elmer G, Kafkafi N, Golani I (2001) Controlling the false 
discovery rate in behavior genetics research. Behavioural Brain Research 
125:279-284. 
Biswas S, Reinhard J, Oakeshott J, Russell R, Srinivasan MV, Claudianos C (2010) 
Sensory regulation of neuroligins and neurexin i in the honeybee brain. PLoS 
ONE 5:e9133. 
Blaustein JD, Erskine MS (2002) Feminine sexual behavior: Cellular integration of 
hormonal and afferent information in the rodent forebrain. In: Hormones, brain 
and behavior (Pfaff DW, Arnold AP, Etgen AM, Fahrbach S, Rubin RT, eds), vol 
1, pp 139-214. New York: Academic Press. 
Borgatti SP, Everett MG, Freeman LC (2002) Ucinet for windows: Software for social 
network analysis. Harvard, MA: Analytic Technologies. 
Breden F, Novinger D, Schubert A (1995) The effect of experience on mate choice in the 
trinidad guppy, poecilia reticulata. Environmental Biology of Fishes 42:323-328. 
Broglio C, Gomez A, Duran E, Ocana FM, Jimenez-Moya F, Rodriguez F, Salas C 
(2005) Hallmarks of a common forebrain vertebrate plan: Specialized pallial areas 
for spatial, temporal and emotional memory in actinopterygian fish. Brain Res 
Bull 66:277-281. 
 108
Brooks R, Endler JA (2001) Female guppies agree to differ: Phenotypic and genetic 
variation in mate-choice behavior and the consequences for sexual selection. 
Evolution 55:1644-1655. 
Brooks R (2002) Variation in female mate choice within guppy populations: Population 
divergence, multiple ornaments and the maintenance of polymorphism. Genetica 
116:343-358. 
Bruce LL, Bradford MR (2009) Evolution of the limbic system. In: Encyclopedia of 
neuroscience (Squire LR, ed), vol 4, pp 43-55. Oxford: Academic Press. 
Burmeister SS, Jarvis ED, Fernald RD (2005) Rapid behavioral and genomic responses to 
social opportunity. Plos Biology 3:1996-2004. 
Chih B, Engelman H, Scheiffele P (2005) Control of excitatory and inhibitory synapse 
formation by neuroligins. Science 307:1324-1328. 
Clayton DF (2000) The genomic action potential. Neurobiol Learn Mem 74:185-216. 
Coleman SW, Patricelli GL, Borgia G (2004) Variable female preferences drive complex 
male displays. Nature 428:742-745. 
Crapon de Caprona MD, Ryan MJ (1990) Conspecific mate recognition in swordtails, 
Xiphophorus nigrensis and X. pygmaeus (poeciliidae): Olfactory and visual cues. 
Animal Behaviour 39:290-296. 
Cummings M, Mollaghan D (2006) Repeatability and consistency of female preference 
behaviours in a northern swordtail, xiphophorus nigrensis. Animal Behaviour 
72:217-224. 
Cummings M, E. , Rosenthal G, G. , Ryan M, J. (2003a) A private ultraviolet channel in 
visual communication. Proceedings of the Royal Society B: Biological Sciences 
270:897-904. 
Cummings ME, Rosenthal GG, Ryan MJ (2003b) A private ultraviolet channel in visual 
communication. Proceedings of the Royal Society of London Series B-Biological 
Sciences 270:897-904. 
Cummings ME, Garcia de Leon FJ, Mollaghan DM, Ryan MJ (2006) Is uv ornamentation 
an amplifier in swordtails? Zebrafish 3:91-100. 
Cummings ME, Larkins-Ford J, Reilly CR, Wong RY, Ramsey M, Hofmann HA (2008) 
Sexual and social stimuli elicit rapid and contrasting genomic responses. Proc 
Biol Sci 275:393-402. 
Cummings ME, Gelineau-Kattner R (2009) The energetic costs of alternative male 
reproductive strategies in xiphophorus nigrensis. J Comp Physiol A Neuroethol 
Sens Neural Behav Physiol 195:935-946. 
Curtis JT, Wang Z (2003) Forebrain c-fos expression under conditions conducive to pair 
bonding in female prairie voles (microtus ochrogaster). Physiol Behav 80:95-101. 
Curtis JT, Liu Y, Aragona BJ, Wang Z (2006) Dopamine and monogamy. Brain Res 
1126:76-90. 
Darwin C (1871) The descent of man, and selection in relation to sex. London,: J. 
Murray. 
Dean C, Dresbach T (2006) Neuroligins and neurexins: Linking cell adhesion, synapse 
formation and cognitive function. Trends Neurosci 29:21-29. 
 109
Demir E, Dickson BJ (2005) Fruitless splicing specifies male courtship behavior in 
drosophila. Cell 121:785-794. 
Desjardins JK, Klausner JQ, Fernald RD (2010) Female genomic response to mate 
information. Proc Natl Acad Sci U S A 107:21176-21180. 
Dong S, Replogle KL, Hasadsri L, Imai BS, Yau PM, Rodriguez-Zas S, Southey BR, 
Sweedler JV, Clayton DF (2009) Discrete molecular states in the brain 
accompany changing responses to a vocal signal. Proc Natl Acad Sci U S A 
106:11364-11369. 
Dufour S, Sebert ME, Weltzien FA, Rousseau K, Pasqualini C (2010) Neuroendocrine 
control by dopamine of teleost reproduction. J Fish Biol 76:129-160. 
Dugatkin LA (1996) Interface between culturally based preferences and genetic 
preferences: Female mate choice in poecilia reticulata. Proceedings of the 
National Academy of Sciences of the United States of America 93:2770-2773. 
Earley RL, Edwards JT, Aseem O, Felton K, Blumer LS, Karom M, Grober MS (2006) 
Social interactions tune aggression and stress responsiveness in a territorial 
cichlid fish (archocentrus nigrofasciatus). Physiol Behav 88:353-363. 
Elias DO, Hebets EA, Hoy RR (2006) Female preference for complex/novel signals in a 
spider. Behavioral Ecology 17:765-771. 
Ellingsen E, Agmo A (2004) Sexual-incentive motivation and paced sexual behavior in 
female rats after treatment with drugs modifying dopaminergic 
neurotransmission. Pharmacol Biochem Behav 77:431-445. 
Elmer KR, Lehtonen TK, Meyer A (2009) Color assortative mating contributes to 
sympatric divergence of neotropical cichlid fish. Evolution 63:2750-2757. 
Everitt BJ (1990) Sexual motivation: A neural and behavioural analysis of the 
mechanisms underlying appetitive and copulatory responses of male rats. 
Neurosci Biobehav Rev 14:217-232. 
Filippi A, Mahler J, Schweitzer J, Driever W (2010) Expression of the paralogous 
tyrosine hydroxylase encoding genes th1 and th2 reveals the full complement of 
dopaminergic and noradrenergic neurons in zebrafish larval and juvenile brain. J 
Comp Neurol 518:423-438. 
Fitzpatrick MJ, Ben-Shahar Y, Smid HM, Vet LEM, Robinson GE, Sokolowski MB 
(2005) Candidate genes for behavioural ecology. Trends in Ecology & Evolution 
20:96-104. 
Garber K (2007) Neuroscience. Autism's cause may reside in abnormalities at the 
synapse. Science 317:190-191. 
Gentner TQ, Hulse SH, Duffy D, Ball GF (2001) Response biases in auditory forebrain 
regions of female songbirds following exposure to sexually relevant variation in 
male song. Journal of Neurobiology 46:48-58. 
Godin J-GJ, Herdman EJE, Dugatkin LA (2005) Social influences on female mate choice 
in the guppy, poecilia reticulata: Generalized and repeatable trait-copying 
behaviour. Animal Behaviour 69:999-1005. 
Goodson JL (2005) The vertebrate social behavior network: Evolutionary themes and 
variations. Horm Behav 48:11-22. 
 110
Goodson JL, Kabelik D (2009) Dynamic limbic networks and social diversity in 
vertebrates: From neural context to neuromodulatory patterning. Front 
Neuroendocrinol 30:429-441. 
Goodson JL, Kabelik D, Kelly AM, Rinaldi J, Klatt JD (2009) Midbrain dopamine 
neurons reflect affiliation phenotypes in finches and are tightly coupled to 
courtship. Proc Natl Acad Sci U S A 106:8737-8742. 
Graham MD, Pfaus JG (2010) Differential regulation of female sexual behaviour by 
dopamine agonists in the medial preoptic area. Pharmacol Biochem Behav 
97:284-292. 
Guiot Y, Rahier J (1997) Validation of nonradioactive in situ hybridization as a 
quantitative approach of messenger ribonucleic acid variations - a comparison 
with northern blot. Diagn Mol Pathol 6:261-266. 
Hahn MW, Kern AD (2005) Comparative genomics of centrality and essentiality in three 
eukaryotic protein-interaction networks. Mol Biol Evol 22:803-806. 
Harvey-Girard E, Tweedle J, Ironstone J, Cuddy M, Ellis W, Maler L (2010) Long-term 
recognition memory of individual conspecifics is associated with telencephalic 
expression of egr-1 in the electric fish apteronotus leptorhynchus. The Journal of 
Comparative Neurology 518:2666-2692. 
Heimovics SA, Riters LV (2008) Evidence that dopamine within motivation and song 
control brain regions regulates birdsong context-dependently. Physiol Behav 
95:258-266. 
Heimovics SA, Cornil CA, Ball GF, Riters LV (2009) D1-like dopamine receptor density 
in nuclei involved in social behavior correlates with song in a context-dependent 
fashion in male european starlings. Neuroscience 159:962-973. 
Herdman EJE, Kelly CD, Godin J-GJ (2004) Male mate choice in the guppy ( poecilia 
reticulata): Do males prefer larger females as mates? Ethology 110:97-111. 
Hoke KL, Burmeister SS, Fernald RD, Rand AS, Ryan MJ, Wilczynski W (2004) 
Functional mapping of the auditory midbrain during mate call reception. J 
Neurosci 24:11264-11272. 
Hoke KL, Ryan MJ, Wilczynski W (2005) Social cues shift functional connectivity in the 
hypothalamus. PNAS 102:10712-10717. 
Holveck MJ, Riebel K (2010) Low-quality females prefer low-quality males when 
choosing a mate. Proc Biol Sci 277:153-160. 
Houde AE, Endler JA (1990) Correlated evolution of female mating preferences and 
male color patterns in the guppy poecilia-reticulata. Science 248:1405-1408. 
Houde AE (1997) Sex, color, and mate choice in guppies. Princeton, N.J.: Princeton 
University Press. 
Hull EM, Lorrain DS, Du J, Matuszewich L, Lumley LA, Putnam SK, Moses J (1999) 
Hormone-neurotransmitter interactions in the control of sexual behavior. Behav 
Brain Res 105:105-116. 
Hull EM, Meisel RL, Sachs BD (2002) Male sexual behavior. In: Hormones, brain and 
behavior (Pfaff DW, Arnold AP, Etgen AM, Fahrbach SE, Rubin RT, eds), vol 1, 
pp 3-137. New York: Academic Press. 
 111
Jamain S, Quach H, Betancur C, Rastam M, Colineaux C, Gillberg IC, Soderstrom H, 
Giros B, Leboyer M, Gillberg C, Bourgeron T (2003) Mutations of the x-linked 
genes encoding neuroligins nlgn3 and nlgn4 are associated with autism. Nat 
Genet 34:27-29. 
Jennions MD, Petrie M (1997) Variation in mate choice and mating preferences: A 
review of causes and consequences. Biol Rev Camb Philos Soc 72:283-327. 
Jones AG, Ratterman NL (2009) Mate choice and sexual selection: What have we learned 
since darwin? Proc Natl Acad Sci U S A 106 Suppl 1:10001-10008. 
Kallman KD (1984) A new look at sex determination in poeciliid fishes. In: Evolutionary 
genetics of fishes (Turner BJ, ed). New York: Plenum Press. 
Kallman KD (1989) Genetic control of size at maturity in Xiphophorus. In: Ecology and 
evolution of livebearing fishes (poeciliidae) (Meffe GK, Snelson FF, eds), pp 163-
184. Englewood Cliffs: Prentice Hall. 
Kaslin J, Panula P (2001) Comparative anatomy of the histaminergic and other aminergic 
systems in zebrafish (danio rerio). The Journal of Comparative Neurology 
440:342-377. 
Kendrick AM, Rand MS, Crews D (1995) Electrolytic lesions to the ventromedial 
hypothalamus abolish receptivity in female whiptail lizards, cnemidophorus 
uniparens. Brain Res 680:226-228. 
Kidd CE, Kidd MR, Hofmann HA (2010) Measuring multiple hormones from a single 
water sample using enzyme immunoassays. Gen Comp Endocrinol 165:277-285. 
Kleitz-Nelson HK, Cornil CA, Balthazart J, Ball GF (2010) Differential effects of central 
injections of d1 and d2 receptor agonists and antagonists on male sexual behavior 
in japanese quail. Eur J Neurosci 32:118-129. 
Kleitz HK, Cornil CA, Balthazart J, Ball GF (2009) Species differences in the relative 
densities of d1- and d2-like dopamine receptor subtypes in the japanese quail and 
rats: An in vitro quantitative receptor autoradiography study. Brain Behav Evol 
73:81-90. 
Kodric-Brown A, Nicoletto PF (2001a) Age and experience affect female choice in the 
guppy (poecilia reticulata). Am Nat 157:316-323. 
Kodric-Brown A, Nicoletto P (2001b) Female choice in the guppy (poecilia reticulata): 
The interaction between male color and display. Behavioral Ecology and 
Sociobiology 50:346-351. 
Lampert KP, Schmidt C, Fischer P, Volff JN, Hoffmann C, Muck J, Lohse MJ, Ryan MJ, 
Schartl M (2010) Determination of onset of sexual maturation and mating 
behavior by melanocortin receptor 4 polymorphisms. Curr Biol 20:1729-1734. 
Larsson LI, Traasdahl B, Hougaard DM (1991) Quantitative non-radioactive in situ 
hybridization. Model studies and studies on pituitary proopiomelanocortin cells 
after adrenalectomy. Histochemistry 95:209-215. 
Laumonnier F, Bonnet-Brilhault F, Gomot M, Blanc R, David A, Moizard MP, Raynaud 
M, Ronce N, Lemonnier E, Calvas P, Laudier B, Chelly J, Fryns JP, Ropers HH, 
Hamel BC, Andres C, Barthelemy C, Moraine C, Briault S (2004) X-linked 
 112
mental retardation and autism are associated with a mutation in the nlgn4 gene, a 
member of the neuroligin family. Am J Hum Genet 74:552-557. 
Lawniczak MK, Begun DJ (2004) A genome-wide analysis of courting and mating 
responses in drosophila melanogaster females. Genome 47:900-910. 
Levinson JN, Li R, Kang R, Moukhles H, El-Husseini A, Bamji SX (2010) Postsynaptic 
scaffolding molecules modulate the localization of neuroligins. Neuroscience 
165:782-793. 
Levitt M, Spector S, Sjoerdsma A, Udenfriend S (1965) Elucidation of the rate-limiting 
step in norepinephrine biosynthesis in the perfused guinea-pig heart. J Pharmacol 
Exp Ther 148:1-8. 
Liao W, Lu X (2009) Male mate choice in the andrew’s toad bufo andrewsi: A preference 
for larger females. Journal of Ethology 27:413-417. 
Liley NR (1965) Ethological isolating mechanisms in four sympatric species of poeciliid 
fishes. Behaviour Supplement:III-197. 
Lise MF, El-Husseini A (2006) The neuroligin and neurexin families: From structure to 
function at the synapse. Cell Mol Life Sci 63:1833-1849. 
Machaalani R, Waters KA (2002) Distribution and quantification of nmda r1 mrna and 
protein in the piglet brainstem and effects of intermittent hypercapnic hypoxia 
(ihh). Brain Res 951:293-300. 
Madani R, Kozlov S, Akhmedov A, Cinelli P, Kinter J, Lipp HP, Sonderegger P, Wolfer 
DP (2003) Impaired explorative behavior and neophobia in genetically modified 
mice lacking or overexpressing the extracellular serine protease inhibitor 
neuroserpin. Mol Cell Neurosci 23:473-494. 
Maney DL, MacDougall-Shackleton EA, MacDougall-Shackleton SA, Ball GF, Hahn TP 
(2003) Immediate early gene response to hearing song correlates with receptive 
behavior and depends on dialect in a female songbird. J Comp Physiol A 
Neuroethol Sens Neural Behav Physiol 189:667-674. 
Marler CA, Foran C, Ryan MJ (1997) The influence of experience on mating preferences 
of the gynogenetic amazon molly. Animal Behaviour 53:1035-1041. 
Martinez I, Paredes RG (2001) Only self-paced mating is rewarding in rats of both sexes. 
Horm Behav 40:510-517. 
McKinnon JS, Mori S, Blackman BK, David L, Kingsley DM, Jamieson L, Chou J, 
Schluter D (2004) Evidence for ecology's role in speciation. Nature 429:294-298. 
Melis MR, Argiolas A (1995) Dopamine and sexual behavior. Neurosci Biobehav Rev 
19:19-38. 
Mello C, Nottebohm F, Clayton D (1995) Repeated exposure to one song leads to a rapid 
and persistent decline in an immediate early gene's response to that song in zebra 
finch telencephalon. J Neurosci 15:6919-6925. 
Miranda E, Lomas DA (2006) Neuroserpin: A serpin to think about. Cell Mol Life Sci 
63:709-722. 
Morris MR, Ryan MJ (1990) Age at sexual maturity of male Xiphophorus nigrensis in 
nature. Copeia 1990:747-751. 
 113
Morris MR, Rios-Cardenas O, Scarlett Tudor M (2006) Larger swordtail females prefer 
asymmetrical males. Biol Lett 2:8-11. 
Morris MR, Rios-Cardenas O, Brewer J (2010) Variation in mating preference within a 
wild population influences the mating success of alternative mating strategies. 
Animal Behaviour 79:673-678. 
Mukai M, Replogle K, Drnevich J, Wang G, Wacker D, Band M, Clayton DF, Wingfield 
JC (2009) Seasonal differences of gene expression profiles in song sparrow 
(melospiza melodia) hypothalamus in relation to territorial aggression. PLoS ONE 
4:e8182. 
Nagel L, Schluter D (1998) Body size, natural selection, and speciation in sticklebacks. 
Evolution 52:209-218. 
Newman SW (1999) The medial extended amygdala in male reproductive behavior. A 
node in the mammalian social behavior network. Ann N Y Acad Sci 877:242-257. 
Northcutt RG (2006) Connections of the lateral and medial divisions of the goldfish 
telencephalic pallium. J Comp Neurol 494:903-943. 
Northcutt RG (2008) Forebrain evolution in bony fishes. Brain Res Bull 75:191-205. 
O'Connell LA, Hofmann HA (2010) Genes, hormones, and circuits: An integrative 
approach to study the evolution of social behavior. Front Neuroendocrinol. 
O'Connell LA, Fontenot MR, Hofmann HA (2011a) Characterization of the 
dopaminergic system in the brain of an african cichlid fish, astatotilapia burtoni. 
The Journal of Comparative Neurology 519:75-92. 
O'Connell LA, Matthews BJ, Crews D (2011b) Neuronal nitric oxide synthase as a 
substrate for the evolution of pseudosexual behaviour in a parthenogenetic 
whiptail lizard. Journal of Neuroendocrinology 23:244-253. 
O'Connell LA, Hofmann HA (accepted) The vertebrate mesolimbic reward system and 
social behavior network: A comparative synthesis. Journal of Comparative 
Neurology. 
O’Connell LA, Matthews BJ, Crews D (2011) Neuronal nitric oxide synthase as a 
substrate for the evolution of pseudosexual behaviour in a parthenogenetic 
whiptail lizard. Journal of Neuroendocrinology 23:244-253. 
Ojanguren AF, Magurran AE (2004) Uncoupling the links between male mating tactics 
and female attractiveness. Proceedings of the Royal Society of London Series B-
Biological Sciences 271:S427-S429. 
Parafati M, Senatori O, Nicotra A (2009) Localization of tyrosine hydroxylase 
immunoreactive neurons in the forebrain of the guppy poecilia reticulata. J Fish 
Biol 75:1194-1205. 
Paredes RG, Agmo A (2004) Has dopamine a physiological role in the control of sexual 
behavior? A critical review of the evidence. Prog Neurobiol 73:179-226. 
Paredes RG (2009) Evaluating the neurobiology of sexual reward. ILAR J 50:15-27. 
Patricelli GL, Uy JA, Walsh G, Borgia G (2002) Male displays adjusted to female's 
response. Nature 415:279-280. 
 114
Pawlisch BA, Riters LV (2010) Selective behavioral responses to male song are affected 
by the dopamine agonist gbr-12909 in female european starlings (Sturnus 
vulgaris). Brain Res 1353:113-124. 
Pfaff DW, Schwartz-Giblin S, McCarthy MM, Kow L-M (1994) Cellular and molecular 
mechanisms of female reproductive behaviors. In: The physiology of reproduction 
(Knobil E, Neill JD, eds), vol 2, pp 107-220. New York: Raven. 
Pfaus JG (1996) Frank a. Beach award: Homologies of animal and human sexual 
behaviors. Hormones and Behavior 30:187-200. 
Plath M, Seggel U, Burmeister H, Heubel K, Schlupp I (2006) Choosy males from the 
underground: Male mating preferences in surface- and cave-dwelling atlantic 
mollies (poecilia mexicana). Naturwissenschaften 93:103-109. 
Radyushkin K, Hammerschmidt K, Boretius S, Varoqueaux F, El-Kordi A, Ronnenberg 
A, Winter D, Frahm J, Fischer J, Brose N, Ehrenreich H (2009) Neuroligin-3-
deficient mice: Model of a monogenic heritable form of autism with an olfactory 
deficit. Genes Brain Behav 8:416-425. 
Ramsey ME, Wong RY, Cummings ME (2011) Estradiol, reproductive cycle and 
preference behavior in a northern swordtail. Gen Comp Endocrinol 170:381-390. 
Rebar D, Zuk M, Bailey NW (2011) Mating experience in field crickets modifies pre- 
and postcopulatory female choice in parallel. Behav Ecol 22:303-309. 
Renn SC, Aubin-Horth N, Hofmann HA (2008) Fish and chips: Functional genomics of 
social plasticity in an african cichlid fish. J Exp Biol 211:3041-3056. 
Rink E, Wullimann MF (2001) The teleostean (zebrafish) dopaminergic system 
ascending to the subpallium (striatum) is located in the basal diencephalon 
(posterior tuberculum). Brain Res 889:316-330. 
Rink E, Wullimann MF (2002) Connections of the ventral telencephalon and tyrosine 
hydroxylase distribution in the zebrafish brain (Danio rerio) lead to identification 
of an ascending dopaminergic system in a teleost. Brain Res Bull 57:385-387. 
Rios-Cardenas O, Tudor MS, Morris MR (2007) Female preference variation has 
implications for the maintenance of an alternative mating strategy in a swordtail 
fish. Animal Behaviour 74:633-640. 
Riters LV, Olesen KM, Auger CJ (2007) Evidence that female endocrine state influences 
catecholamine responses to male courtship song in european starlings. Gen Comp 
Endocrinol 154:137-149. 
Riters LV (2010) Evidence for opioid involvement in the motivation to sing. J Chem 
Neuroanat 39:141-150. 
Riters LV (2011) Pleasure seeking and birdsong. Neurosci Biobehav Rev. 
Robbins E, Baldino F, Jr., Roberts-Lewis JM, Meyer SL, Grega D, Lewis ME (1991) 
Quantitative non-radioactive in situ hybridization of preproenkephalin mrna with 
digoxigenin-labeled crna probes. Anat Rec 231:559-562. 
Robinson GE, Grozinger CM, Whitfield CW (2005) Sociogenomics: Social life in 
molecular terms. Nat Rev Genet 6:257-270. 
Robinson GE, Fernald RD, Clayton DF (2008) Genes and social behavior. Science 
322:896-900. 
 115
Rosenqvist G, Houde A (1997) Prior exposure to male phenotypes influences mate 
choice in the guppy, Poecilia reticulata. Behavioral Ecology 8:194-198. 
Rosenthal GG, Wagner WE, Ryan MJ (2002) Secondary reduction of preference for the 
sword ornament in the pygmy swordtail Xiphophorus nigrensis (pisces: 
Poeciliidae). Animal Behaviour 63:37-45. 
Rutledge JM, Miller A, Uetz GW (2010) Exposure to multiple sensory cues as a juvenile 
affects adult female mate preferences in wolf spiders. Animal Behaviour 80:419-
426. 
Ryan MJ, Wagner WE (1987) Asymmetries in mating preferences between species - 
female swordtails prefer heterospecific males. Science 236:595-597. 
Ryan MJ, Causey BA (1989) Alternative mating-behavior in the swordtails Xiphophorus 
nigrensis and Xiphophorus pygmaeus (pisces, Poeciliidae). Behavioral Ecology 
and Sociobiology 24:341-348. 
Ryan MJ, Hews DK, Wagner WE (1990) Sexual selection on alleles that determine body 
size in the swordtail Xiphophorus nigrensis. Behavioral Ecology and 
Sociobiology 26:231-237. 
Ryan MJ, Pease CM, Morris MR (1992) A genetic-polymorphism in the swordtail 
Xiphophorus nigrensis - testing the prediction of equal fitnesses. American 
Naturalist 139:21-31. 
Ryan MJ, Rosenthal GG (2001) Variation and selection in swordtails. In: Model systems 
in behvioral ecology (Dugatkin LA, ed), pp 133-148. Princeton: Princeton 
University Press. 
Salzburger W, Niederstatter H, Brandstatter A, Berger B, Parson W, Snoeks J, 
Sturmbauer C (2006) Colour-assortative mating among populations of tropheus 
moorii, a cichlid fish from lake tanganyika, east Africa. Proc Biol Sci 273:257-
266. 
Schlupp I, Marler C, Ryan MJ (1994) Benefit to male sailfin mollies of mating with 
heterospecific females. Science 263:373-374. 
Schluter D (2009) Evidence for ecological speciation and its alternative. Science 
323:737-741. 
Schultz W, Dayan P, Montague PR (1997) A neural substrate of prediction and reward. 
Science 275:1593-1599. 
Schultz W (1998) Predictive reward signal of dopamine neurons. J Neurophysiol 80:1-27. 
Schultz W (2007a) Multiple dopamine functions at different time courses. Annu Rev 
Neurosci 30:259-288. 
Schultz W (2007b) Behavioral dopamine signals. Trends Neurosci 30:203-210. 
Scott AP, Hirschenhauser K, Bender N, Oliveira R, Earley RL, Sebire M, Ellis T, 
Pavlidis M, Hubbard PC, Huertas M, Canario A (2008) Non-invasive 
measurement of steroids in fish-holding water: Important considerations when 
applying the procedure to behaviour studies. Behaviour 145:1307-1328. 
Searcy WA (1992) Song repertoire and mate choice in birds. American Zoologist 32:71-
80. 
 116
Servedio MR, Saether SA, Saetre GP (2009) Reinforcement and learning. Evolutionary 
Ecology 23:109-123. 
Shoemaker CM, Queen J, Crews D (2007) Response of candidate sex-determining genes 
to changes in temperature reveals their involvement in the molecular network 
underlying temperature dependent sex determination. Molecular Endocrinology 
21:2750-2763. 
Sockman KW, Gentner TQ, Ball GF (2002) Recent experience modulates forebrain gene-
expression in response to mate-choice cues in european starlings. Proc Biol Sci 
269:2479-2485. 
Sockman KW, Gentner TQ, Ball GF (2005) Complementary neural systems for the 
experience-dependent integration of mate-choice cues in european starlings. J 
Neurobiol 62:72-81. 
Sockman KW, Salvante KG (2008) The integration of song environment by 
catecholaminergic systems innervating the auditory telencephalon of adult female 
european starlings. Dev Neurobiol 68:656-668. 
Stockinger P, Kvitsiani D, Rotkopf S, Tirián L, Dickson BJ (2005) Neural circuitry that 
governs drosophila male courtship behavior. Cell 121:795-807. 
Sudhof TC (2008) Neuroligins and neurexins link synaptic function to cognitive disease. 
Nature 455:903-911. 
Svec LA, Lookingland KJ, Wade J (2009) Estradiol and song affect female zebra finch 
behavior independent of dopamine in the striatum. Physiol Behav 98:386-392. 
Tabuchi K, Blundell J, Etherton MR, Hammer RE, Liu X, Powell CM, Sudhof TC (2007) 
A neuroligin-3 mutation implicated in autism increases inhibitory synaptic 
transmission in mice. Science 318:71-76. 
Tinbergen N (1963) On aims and methods of ethology. Zeitschrift für Tierpsychologie 
20:410-433. 
Tracy AL, Jarrard LE, Davidson TL (2001) The hippocampus and motivation revisited: 
Appetite and activity. Behav Brain Res 127:13-23. 
Tudor MS, Morris MR (2009) Experience plays a role in female preference for symmetry 
in the swordtail fish xiphophorus malinche. Ethology 115:812-822. 
Tye KM, Janak PH (2007) Amygdala neurons differentially encode motivation and 
reinforcement. J Neurosci 27:3937-3945. 
Varoqueaux F, Aramuni G, Rawson RL, Mohrmann R, Missler M, Gottmann K, Zhang 
W, Sudhof TC, Brose N (2006) Neuroligins determine synapse maturation and 
function. Neuron 51:741-754. 
Vecchiola A, Collyer P, Figueroa R, Labarca R, Bustos G, Magendzo K (1999) 
Differential regulation of mu-opioid receptor mrna in the nucleus accumbens shell 
and core accompanying amphetamine behavioral sensitization. Brain Res Mol 
Brain Res 69:1-9. 
Walling CA, Royle NJ, Metcalfe NB, Lindstrom J (2007) Green swordtails alter their age 
at maturation in response to the population level of male ornamentation. Biology 
Letters 3:144-146. 
 117
Walling CA, Royle NJ, Lindstrom J, Metcalfe NB (2008) Experience-induced preference 
for short-sworded males in the green swordtail, Xiphophorus helleri. Animal 
Behaviour 76:271-276. 
Westneat DF, Walters A, McCarthy TM, Hatch MI, Hein WK (2000) Alternative 
mechanisms of nonindependent mate choice. Anim Behav 59:467-476. 
Whitfield CW, Cziko A-M, Robinson GE (2003) Gene expression profiles in the brain 
predict behavior in individual honey bees. Science (Washington D C) 302:296-
299. 
Wilczynski W, Lynch KS (2010) Female sexual arousal in amphibians. Horm Behav. 
Wise RA (2002) Brain reward circuitry: Insights from unsensed incentives. Neuron 
36:229-240. 
Wong BBM, Candolin U (2005) How is female mate choice affected by male 
competition? Biol Rev 80:559-571. 
Wong RY, Hofmann HA (2010) Behavioral genomics: An organismal perspective. In: 
Encyclopedia of Life Sciences,  Chichester: John Wiley & Sons Ltd. 
Wong RY, So P, Cummings ME (in review) How female size and male displays 
influence female mate preference behavior in the northern swordtail, Xiphophorus 
nigrensis 
Wong RY, Ramsey M, Cummings ME (in review) Neural correlates of female mate 
preference behavior: A candidate gene approach 
Woodgate JL, Bennett ATD, Leitner S, Catchpole CK, Buchanan KL (2010) 
Developmental stress and female mate choice behaviour in the zebra finch. 
Animal Behaviour 79:1381-1390. 
Woolley SC, Sakata JT, Gupta A, Crews D (2001) Evolutionary changes in dopaminergic 
modulation of courtship behavior in cnemidophorus whiptail lizards. Horm Behav 
40:483-489. 
Woolley SC, Doupe AJ (2008) Social context-induced song variation affects female 
behavior and gene expression. PLoS Biol 6:e62. 
Yamamoto K, Tohyama M, Shimizu N (1977) Comparative anatomy of the topography 
of catecholamine containing neuron system in the brain stem from birds to 
teleosts. J Hirnforsch 18:229-240. 
Yamamoto K, Ruuskanen JO, Wullimann MF, Vernier P (2010) Two tyrosine 
hydroxylase genes in vertebrates new dopaminergic territories revealed in the 
zebrafish brain. Mol Cell Neurosci 43:394-402. 
Yamamoto K, Ruuskanen JO, Wullimann MF, Vernier P (2011) Differential expression 
of dopaminergic cell markers in the adult zebrafish forebrain. J Comp Neurol 
519:576-598. 
Young KA, Gobrogge KL, Liu Y, Wang Z (2011) The neurobiology of pair bonding: 
Insights from a socially monogamous rodent. Front Neuroendocrinol 32:53-69. 
Zhang L, Ravipati A, Joseph J, Roth GS (1995) Aging-related changes in rat striatal d2 
receptor mrna-containing neurons: A quantitative nonradioactive in situ 
hybridization study. J Neurosci 15:1735-1740. 
 118
Zhao C, Fujinaga R, Yanai A, Kokubu K, Takeshita Y, Watanabe Y, Shinoda K (2008) 
Sex-steroidal regulation of aromatase mrna expression in adult male rat brain: A 
quantitative non-radioactive in situ hybridization study. Cell Tissue Res 332:381-
391.