General and Comparative Endocrinology 157 (2008) 249–253
Contents lists available at ScienceDirect
General and Comparative Endocrinology
journal homepage: www.elsevier.com/locate/ygcen
Minireview
Endocrinology in field studies: Problems and solutions for the experimental design
Leonida Fusani *
Department of Biology and Evolution, University of Ferrara, Via Luigi Borsari 46, 44100 Ferrara, Italy
a r t i c l e
i n f o
Article history:
Received 15 January 2008
Revised 18 April 2008
Accepted 30 April 2008
Available online 6 May 2008
Keywords:
Hormone treatment
Gonadectomy
Hormone replacement
Methods
Field study
Testosterone
Hormone-dependent traits
a b s t r a c t
The increasing interest in hormones among field biologists can be frustrating because of the difficulties of
applying classical endocrinological methods to natural settings. A few thoroughly tested methods have
become popular because of their simplicity of use. This does not mean that such methods are the best
or the appropriate ones for all studies. In this brief review I will examine some common problems
encountered by field biologists who want to study the relationships between a morphological, behavioral,
or physiological trait and a hormone. First, I will discuss why questions asked in the field often differ substantially from those asked in the laboratory, and how to adapt the design of the experiment accordingly.
Second, I will review alternative methods to study hormone–trait relationships and how to combine
them to strengthen the conclusions that can be drawn from the study. Then, I will discuss how to find
the right control for a hormonal manipulation. Finally, I will examine the pitfalls associated with longterm hormonal treatment and the available methods for such types of studies.
Ó 2008 Elsevier Inc. All rights reserved.
1. Introduction
There is an increasing interest in hormones among ecologists,
ethologists, and other researchers who work with free-living species. The birth date of field endocrinology can be traced back to
1802, when George Montagu observed that in male songbirds singing activity was higher in the periods during which birds had larger
testes (Montagu, 1802, cited in Armstrong, 1963). By the 1950,
there were several studies on hormones in free-living animals,
and particularly in birds (Collias, 1950). However, the major impulse to the development of the field came from the studies of John
C. Wingfield in the 1970s. Wingfield developed methods to measure hormone concentration in small blood samples taken from
wild birds without the need for killing the animals (Wingfield and
Farner, 1975) and used the technique to analyze the circulating
concentration of androgen, estrogen, and corticosteroids in relation
to season, territorial behavior, and life cycle stages. Beside his fundamental contributions to the theoretical aspects of wildlife endocrinology (Wingfield et al., 1990, 1997, 2001; Wingfield and Farner,
1993; Wingfield and Moore, 1987), Wingfield has been particularly
successful in developing methods to address questions that had
been previously investigated only in laboratory animals. The great
success of field endocrinology in the last years has resulted in an increase in the number of researchers from other areas who study
how hormones modulate behavior, developmental stages, life history stages, and immunological parameters. Very often these
* Corresponding author. Fax: +39 0532 207143.
E-mail address: [email protected]
0016-6480/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.ygcen.2008.04.016
researchers look for simple methods to study the action of hormones, and realize that most of classical laboratory studies are
based on methods that cannot be easily applied in the field. This
led to the strategy of sticking to few published methods. However,
this approach might result in adopting methodologies that are not
the most appropriate ones for the specific aim of the study. In addition, studies involving hormones are often published in journals
without a focus on endocrinology, which might favor the publication of studies based on uncomplicated tests of hormone action.
In this brief article, I would like to address some common problems
encountered by field biologists when designing experiments
involving hormones. Because my work has focused mainly on
androgen and estrogen, I will base my review on these hormones.
However, most of the concepts developed here refer to hormones
in general. A more general review on methodological issues in field
endocrinology has been published recently (Fusani et al., 2005).
2. With or without gonads?
Classic studies of behavioral endocrinology typically involved
the removal of the natural source of the hormone and the replacement with exogenous hormone. This approach served mainly two
types of studies, those testing hormone-dependent traits, and
those testing the effects of hormone agonists and antagonists.
The oldest example of the first type of studies is the pioneering
work of Berthold on caponization (Berthold, 1849). Farmers have
known for centuries that if the testicles are removed from young
male chicks, masculine traits will fail to develop, which in the case
of fowls is called caponization. Berthold demonstrated that if the
250
L. Fusani / General and Comparative Endocrinology 157 (2008) 249–253
testicles were reimplanted in castrated chicks they will develop
normal masculine traits (Berthold, 1849). Later it was discovered
that the testicles are the main source of androgen in males, thus
the traits that were lost as a result of castration but restored by
androgen treatment were called androgen-dependent. This experimental protocol has been used successfully in a large number of
key studies, including the classical paper of Phoenix and coworkers
describing the principles of sexual differentiation of brain and
behavior in mammals (Phoenix et al., 1959). However, there are
several reasons for which such an approach might be difficult or
inappropriate in field endocrinology. First, gonadectomy might be
challenging to perform in small animals, particularly under field
conditions. Secondly, it is an invasive operation that requires some
recovery period, which might expose the individuals to increased
predation risks and is undesired when behavioral tests have to
be performed shortly afterwards. Finally, because of its surgical
nature gonadectomy is not always allowed by licensing authorities.
The problem of studying if a trait is androgen-dependent when
gonadectomy is not possible or desirable has been addressed in
several ways. Correlational evidence can be collected for example
by showing that the trait is modified seasonally in concurrence
with changes in androgen levels. However, this kind of observations can be only indicative as other seasonal factors like the photoperiod could be directly responsible for the variation of the trait
(Dawson et al., 2001; Kirn and Schwabl, 1997). The latter hypothesis can be tested by treating the animals with the hormone when
its natural levels are low and see whether the trait changes in the
expected direction. Alternatively, an androgen antagonist can be
used to inhibit receptor-mediated responses. Also this approach
is not immune to the effects of confounding variables, and can be
used only for short-term experiments because receptor inhibition
alters the regulatory feedback on the hormone. The real problem,
however, is not how to find an alternative to gonadectomy and
hormone replacement but rather whether this protocol is the
appropriate one for the study in question. The removal of the gonads has several side effects and a gonadectomized animal is not
simply an animal without gonadal hormones. Moreover, we now
know that ‘gonadal hormones’ can be produced in large amounts
also at extragonadal sites (Callard et al., 1978; Naftolin et al.,
1975; Schlinger and Arnold, 1991). Thus, unless the study is specifically aimed to know what happens when the gonads are removed,
other experimental approaches might be preferable. In fact, a
behavioral ecologist or an ecological immunologist are probably
more interested in asking how the studied trait is modulated by
the hormone, i.e. how variations in the trait are related to hormone
variations. In this perspective, the alternative approaches listed
above and discussed in the next section are likely to be more effective than the castration-replacement protocol.
3. Alternative approaches to studying hormone function
Let’s say we have decided to adopt one or a combination of the
alternative approaches presented above: seasonal or life history
stage correlations between the trait(s) and the hormone, hormone
treatment in periods and/or physiological conditions of low hormone levels, and hormone or hormone receptor manipulation in
intact animals. Each of these approaches has some pitfalls that deserve to be examined in details. Correlational evidence should always imply a good knowledge of the seasonal or ontogenetic
stages for a correct interpretation of the data. For example, male
canaries show large changes in androgen and androgen receptor
levels between autumn and spring which are correlated with differences in song activity and song structure (Fusani et al., 2000;
Gahr and Metzdorf, 1997; Leitner et al., 2001a,b; Nottebohm
et al., 1986). Therefore it would be tempting to conclude that
low androgen levels together with low androgen receptor expression are related to song instability. However, this correlation holds
for the early autumn but not for the later autumn: in November
canaries have very low levels of androgen yet the song is already
stable and the expression of androgen receptor in the neural song
system does not differ from spring (Fusani et al., 2000). In fact, the
decrease in circulating androgen, brain androgen receptors, and
song stability in the early autumn are all expression of a general
‘shutdown’ of reproductive traits which is typical of the molt period (Dawson, 2006; Dawson et al., 2001; Nicholls et al., 1988). All
these data do not challenge the notion that song is regulated by
androgen in male canaries, on the contrary, they illustrate how dynamic the system is. At the same time though, they tell us that it
would be inexact to conclude that song stability is proportionally
related to concentrations of androgen in the blood.
The second type of approach is to treat the animal with a hormone when its endogenous levels are naturally low. A good example is the stimulation of song development in female songbirds by
testosterone. Already in the 1939 Leonard showed that the development of masculine song could be ‘induced’ in female canaries
by injecting the recently discovered male hormone, testosterone
(Leonard, 1939). A straightforward generalization of these results
is that song development in males depends on the action of gonadal androgen. This interpretation was challenged by the discovery
that the action of testosterone within the brain is often mediated
by its conversion into estrogen by the enzyme aromatase (Hutchison, 1971; Naftolin et al., 1975), which is particularly abundant in
the brain of male songbirds (Metzdorf et al., 1999; Schlinger and
Arnold, 1991). Nevertheless, female songbirds have lower concentrations of aromatase in their brain, thus it could be concluded that
song development in females is not mediated by aromatization but
depends on the androgenic action of testosterone. Intact females,
however, do also have very high aromatase activity in their ovaries
so when they are given testosterone there is a significant increase
in the concentration of circulating estrogen which is produced by
the ovarian aromatization (Fusani, 2000; Fusani et al., 2003). This
phenomenon would last a few days but if testosterone treatment
is prolonged estrogen levels decrease to basal, probably because
of the negative feedback of estrogenic metabolites of testosterone
on LH production (Fusani et al., 2003). Thus also in female canaries
song development depends on both androgenic and estrogenic actions of testosterone. Indeed when the estrogenic action is blocked
by giving an aromatase inhibitor together with testosterone, song
development is altered (Fusani et al., 2003).
Hormone manipulation in intact animals can sometime provide
unexpected results. It is generally assumed that ‘hormone-dependent traits’ means ‘traits whose expression varies proportionally
with hormone concentration’, although we have seen above that
the definition is traditionally based on the removal-replacement
protocol. In reality, proportional relationships between hormones
and morphological or behavioral traits are less common than
thought, and in many cases a threshold mechanism seems to be involved (reviewed by Adkins-Regan, 2005; Fusani and Hutchison,
2003; reviewed by Hews and Moore, 1997). Thus it is not unusual
that the same hormonal treatment that leads to an increase in trait
expression in animals which have low endogenous concentrations
of the hormone has no effect on the trait when the animals have
high endogenous levels. For example, testosterone can induce
courtship behavior in juvenile and female Golden-collared manakins and in non-breeding males (Day et al., 2006) but does not affect courtship activity in breeding males (Day et al., 2007).
Similarly, in male ring doves courtship is testosterone-dependent
following the classical definition (Hutchison, 1970) but testosterone treatment of courting males does not result in further increases in courtship activity (Fusani and Hutchison, 2003).
Sometime administration of exogenous hormones can actually
251
L. Fusani / General and Comparative Endocrinology 157 (2008) 249–253
cause a reduction in the ‘hormone-dependent’ trait. In vertebrates,
testosterone regulates spermatogenesis, which is thus considered
an androgen-dependent mechanism. Nevertheless administration
of testosterone to intact animals can lead to a suppression of spermatogenesis because of the negative feedback on gonadotropic
hormones (Brown and Follett, 1977; Turek et al., 1976) (Fig. 1).
To further complicate the story, if the amount of exogenous testosterone is increased above a certain point the suppressive effect is
reversed (Brown and Follett, 1977; Turek et al., 1976) (Fig. 1). Thus,
one should be very careful in choosing the dosage for the hormone
treatment. The rationale of keeping the concentrations within
physiological levels is particularly important.
4. The search for the optimal control
No matter which approach is used, an inescapable problem of
hormonal manipulation methods is the choice of the appropriate
control. In the early times of behavioral endocrinology, researchers
used as control for a hormone treatment either a blank—the vehicle in case of injections or an empty tube in case of silicone implants—or a substance similar to the tested hormone but without
hormonal action. For example, cholesterol was used as the control
for steroid hormones in many early studies. In later studies, the
vehicle or empty tube option became the common one. Today,
most if not all studies use no chemical as control for hormone
treatment. For several reasons, however, this might not be the best
choice. In fact the use of a blank control derives from the removalreplacement protocol, where the question was: What happens
when a hormone that has been removed by resecting the gland
is replaced by an exogenous hormone. In many field endocrinology
studies the question is rather: what happens when the circulating
concentrations of the hormone are increased (or reduced). In the
previous paragraph I have discussed how important is to keep
the exogenous hormone within the physiological range. Nevertheless, typically the treatment is made to obtain concentrations
which are at the upper end of the range, so that there is a significant difference between the treated and the control animals. If this
is the rationale of the experiment, a low dose treatment might be a
better control for a high dose treatment compared to a blank. In
fact, the high dose treatment has a series of physiological effects
that a blank does not have (Table 1). Thus, whereas there is a qualitative difference between the high dose and the blank, the difference between a low and a high dose is mainly quantitative. Ideally,
both types of control should be used, but if there are limitations in
the number of animals that can be used, the low dose implant
might be preferable for some studies. We used this approach in a
Fig. 1. Effects of implantation of silastic capsules filled with testosterone proprionate (TP) on testicular weight, tubule diamater, and concentrations of gonadotropin in intact male quails. Testosterone caused a suppression of gonadotropin release
and a reduction of testicular weight and tubule diameter, but the effects on the
testes were reversed when the implant length reached 50 mm. Drawn from Brown
and Follett (1977).
Table 1
Effects of testosterone implants on physiological variables in intact animals
Changes testosterone concentration
Introduces exogenous testosterone
Alters feedback on gonadotropin
Alters release of endogenous testosterone
Affects steroid metabolism/receptor
Blank
implant
Low
testosterone
High
testosterone
No
No
No
No
No
Yes
Yes
Yes
Possibly
Possibly
Yes
Yes
Yes
Yes
Yes
Beside increasing the concentration of testosterone, the hormone implants likely
alter a series of physiological variables which are not affected by the blank.
study of the effects of testosterone on courtship behavior of male
ring doves (Fusani and Hutchison, 2003). Male courtship is abolished by castration and restored by testosterone implants in this
species (Adkins-Regan, 1981; Barfield, 1971; Cheng and Lehrman,
1975; Hutchison, 1970, 1971), and thus match the classical definition for testosterone-dependent traits. We wanted to test whether
an increase of testosterone in males already in full reproductive
conditions would lead to an increase in courtship activity. Intact
males were implanted with two different doses of testosterone in
silastic implants. After a week, the circulating levels of testosterone
were significantly different between the two groups, but courtship
behavior was not (Fusani and Hutchison, 2003). If a ‘low dose’ control is used, however, great care should be taken in the choice of
the dose, which should not modify significantly the circulating levels of the hormone. In fact, even little modifications of the hormone
levels can have significant biological effects. For example, a very
moderate increase in plasma corticosterone levels in female
white-crowned sparrows alters the offspring sex-ratio (Bonier
et al., 2007).
5. The problems with prolonged hormonal treatment
There are two major difficulties in continuing hormonal treatment for longer than a few days. First, the endocrine system is a
homeostatic system, and tends to compensate for disruption. The
secretion of steroid hormones, including androgen, estrogen, and
corticosteroids, is controlled by negative feedback systems that
stimulate hormone production when the blood concentrations
are too low and vice versa. Thus, if we treat intact animals with a
hormone, after a few days the endogenous production of the hormone will be reduced. A prolonged treatment can lead to the
regression of the endocrine gland. Thus, any treatment which lasts
longer than, say, a week, is likely to have substantial consequences
on the endocrine homeostasis. At first sight this may appear a minor problem: if the aim of the study is to test the effects of hormone
increase and the dose of exogenous hormone is high enough to replace endogenous production, the experimental animals will have
elevated hormone levels throughout the treatment period. However, it is unusual that after a hormonal treatment the target variables are recorded continuously. For example, we might implant
testosterone in male singbirds and then look at effects on song 1
week after the beginning of the treatment. Now if the time at
which we record the behavior of treated birds is postponed to 4
weeks after the treatment, we will probably obtain very different
results. This might depend not only on the method used for the
administration of the hormone (which will be discussed below),
but also on the resetting of the endocrine homeostasis. In fact
the alteration of the negative feedback will probably have additional effects beside the changes in the regulation of the endogenous hormone. For example, the antiandrogen drug flutamide, an
androgen receptor antagonist, reduces courtship activity in male
golden-collared manakins 1 week after the implantation (Fusani
et al., 2007). However, these effects disappear 2 weeks after the
252
L. Fusani / General and Comparative Endocrinology 157 (2008) 249–253
implantation and are actually reversed after 3 weeks (Fusani et al.,
2007). The reasons for the inversion are not fully understood yet,
however it has been shown that antiandrogen can lead to a substantial upregulation of androgen receptor, which results in a
hypersensitivization of tissues to very low concentration of androgen (Chen et al., 2004).
The second problem is how to conduct the treatment. Unless we
are looking at short-term effects, injections are not very convenient. In most cases wild animals cannot be captured every day
to give them an injection, and even if this was possible, such repeated handling would probably cause stress. The choice method
so far, at least for lipophilic hormones, has been to pack the hormone inside silicon tubing which is then closed to both ends and
implanted under the skin. The hormone then diffuses through
the semipermeable walls of the tube into the subcutaneous. Silicon
implants (better known as silastic from the name of the product of
Dow Corning Corporation) remain the method of choice because
the tubing itself is inexpensive. The release rate depends mainly
on the length of the tubing, the thickness of the tubing wall, the
chemical nature of the hormone, and the amount of hormone
packed inside. Thus, pilot studies should be conducted to make
sure that the resulting blood concentrations of hormones are within the desired range. Using implants of the same size as those used
with other hormones or in other species might in fact give very different results because of differences in the chemical properties of
the hormone or in the metabolism between species. It should be
noted that there is often confusion among the releasing properties
of such implants. Silastic implants allow a continuous release, but
this does not mean a consistent release. The hormone diffuses without interruption but the release rate might change considerably
over the time. Among the drug releasing devices alternative to
silastic tubes there are osmotic mini-pumps (ALZET, Durect Corporation, Cupertino, CA, USA) and time-release pellets (Innovative Research of America, Sarasota, Florida, USA), which offer some
advantages compared to silastic but also some disadvantages (Table 2). The major advantage of these alternative devices seems to
be the consistent release rate. For example, we compared plasma
levels of testosterone in female canaries at 1-week intervals after
subcutaneous
implantation
of
either
silastic
capsules
(8 mm 0.762 mm ID 1.651 mm OD Silastic Tubing, Dow Corning Corporation) filled with crystalline testosterone or time-release
pellets (Innovative Research of America, Sarasota, Florida, USA)
containing 1.5 mg of testosterone (60-day release, 25 lg/day).
Blood was sampled from the jugular vein and testosterone concentrations measured by radioimmunoassay (detailed methods in
Fusani et al., 2000). Values decreased significantly in both groups
during the first 3 weeks of implantation (two-way repeated measure ANOVA; F2,20 = 35.132, P < 0.0001) (Fig. 2). However, testosterone levels were more consistent in animals implanted with
the pellets than in those implanted with the silicon tubing, as
shown by the significant interaction between weeks of implantation and type of implant (F2,20 = 7.654, P = 0.003). In addition, 1
Table 2
Comparison between properties of silicon tubes, osmotic mini-pumps, and timerelease pellets
Device
Advantages
Disadvantages
Silastic
Inexpensive; flexible use
Osmotic
minipumps
Timerelease
pellets
Consistent release; high
replicability; easy to prepare
Poor replicability; only lipophilic
substances; variable release rate
Expensive; relatively large;
substances must be dissolved
Consistent release; high
replicability; most substances
Expensive; packed by producer
Fig. 2. Mean (±SEM) plasma concentrations of testosterone (T) in female canaries
implanted subcutaneously with time-release pellets or silicon tubing containing
testosterone. T levels decreased less sharply in pellet-implanted females compared
to silicon-implanted ones. See text for details.
week after the implantation testosterone levels were above the
physiological range in the animals implanted with the silicon capsules, and if one performs a non-linear regression with the data,
the concentration one day after the implantation would result to
be about 50 ng/ml! Thus, no matter what method is chosen for
the treatment, a measurement of the circulating concentrations
of the implanted hormone soon after the implantation is highly
recommended, particularly when working on a new species or
with a new implant type.
6. Conclusions
In this brief review I have highlighted some common problems
encountered when conducting hormone studies with free-living
animals and suggested some possible solutions. I have focused on
those aspects of experimental design that are more likely to affect
significatively the soundness of the experiment. In no way I
wanted to discourage biologists without an endocrinology background to embark on hormone studies. On the contrary, the variety
of behavioral, morphological, and developmental traits controlled
by hormones offers endless opportunities to apply endocrinology
research in natural settings. The line of research started by John
Wingfield more than 30 years ago has generated new, important
concepts and ideas in endocrinology. Its extension to other fields
appears similarly promising.
Acknowledgments
I am grateful to John Wingfield for proposing me to contribute
to this special issue. I thank Virginie Canoine, Wolfgang Goymann,
and Michaela Hau for reading and commenting a previous version
of this manuscript.
References
Adkins-Regan, E., 2005. Hormones and Animal Social Behavior. Princeton University
Press, Princeton.
Adkins-Regan, E.K., 1981. Effect of sex steroids on the reproductive behavior of
castrated male ring doves (Streptopelia risoria). Physiol. Behav. 26, 561–565.
Armstrong, E.A., 1963. A Study of Bird Song. Oxford University Press, London.
Barfield, R.J., 1971. Activation of sexual and aggressive behaviour by androgen
implanted into the male ring dove brain. Endocrinology 89, 1470–1476.
Berthold, A.A., 1849. Transplantation der Hoden. Arch. Anat. Physiol. 16, 42–46.
Bonier, F., Martin, P.R., Wingfield, J.C., 2007. Maternal corticosteroids influence
primary offspring sex ratio in a free-ranging passerine bird. Behav. Ecol. 18,
1045–1050.
Brown, N.L., Follett, B.K., 1977. Effects of androgens on the testes of intact and
hypophysectomized Japanese quail. Gen. Comp. Endocrinol. 33, 267–277.
Callard, G.V., Petro, Z., Ryan, K.J., 1978. Phylogenetic distribution of aromatase and
other androgen-converting enzymes in the central nervous system.
Endocrinology 103, 2283–2290.
L. Fusani / General and Comparative Endocrinology 157 (2008) 249–253
Chen, C.D., Welsbie, D.S., Tran, C., Baek, S.H., Chen, R., Vessella, R., Rosenfeld, M.G.,
Sawyers, C.L., 2004. Molecular determinants of resistance to antiandrogen
therapy. Nat. Med. 10, 33–39.
Cheng, M.-F., Lehrman, D., 1975. Gonadal hormone specificity in the sexual behavior
of ring doves. Psychoneuroendocrinology 1, 95–102.
Collias, N.E., 1950. Hormones and behavior with special reference to birds and the
mechanisms of hormone action. In: Gordon, E.S. (Ed.), Steroid Hormones.
University of Wisconsin Press, pp. 277–329.
Dawson, A., 2006. Control of molt in birds: association with prolactin and gonadal
regression in starlings. Gen. Comp. Endocrinol. 147, 314–322.
Dawson, A., King, V.M., Bentley, G.E., Ball, G.F., 2001. Photoperiodic control of
seasonality in birds. J. Biol. Rhythms 16, 365–380.
Day, L.B., Fusani, L., Hernandez, E., Billo, T.J., Sheldon, K.S., Wise, P.M., Schlinger, B.A.,
2007. Testosterone and its effects on courtship in golden-collared manakins
(Manacus vitellinus): seasonal, sex, and age differences. Horm. Behav. 51, 69–76.
Day, L.B., McBroom, J.T., Schlinger, B.A., 2006. Testosterone activates courtship
display but does not alter plumage in the tropical golden-collared manakin
(Manacus vitellinus). Horm. Behav. 49, 223–232.
Fusani, L., 2000. Action of Oestrogen on Brain Mechanisms of Behaviour in the Ring
Dove (Streptopelia risoria) and Canary (Serinus canaria). University of Cambridge,
Cambridge.
Fusani, L., Canoine, V., Goymann, W., Wikelski, M., Hau, M., 2005. Difficulties and
special issues associated with field research in behavioral neuroendocrinology.
Horm. Behav. 48, 484–491.
Fusani, L., Day, L.B., Canoine, V., Reinemann, D., Hernandez, E., Schlinger, B., 2007.
Androgen and the elaborate courtship behavior of a tropical lekking bird. Horm.
Behav. 51, 62–68.
Fusani, L., Hutchison, J.B., 2003. Lack of changes in the courtship behaviour of male
ring doves after testosterone treatment. Ethol. Ecol. Evol. 15, 143–157.
Fusani, L., Metzdorf, R., Hutchison, J.B., Gahr, M., 2003. Aromatase inhibition affects
testosterone-induced masculinization of song and the neural song system in
female canaries. J. Neurobiol. 54, 370–379.
Fusani, L., Van’t Hof, T., Hutchison, J.B., Gahr, M., 2000. Seasonal expression of
androgen receptors, estrogen receptors and aromatase in the canary brain in
relation to circulating androgens and estrogens. J. Neurobiol. 43, 254–268.
Gahr, M., Metzdorf, R., 1997. Distribution and dynamics in the expression of
androgen and estrogen receptors in vocal control systems of songbirds. Brain
Res. Bull. 44, 509–517.
Hews, D.K., Moore, M.C., 1997. Hormones and sex-specific traits: critical questions.
In: Beckage, N.E. (Ed.), Parasites and Pathogens Effects on Host Hormones and
Behavior. Chapman & Hall, New York, pp. 277–292.
Hutchison, J.B., 1970. Differential effects of testosterone and oestradiol on male
courtship in Barbary doves (Streptopelia risoria). Anim. Behav. 18, 41–52.
Hutchison, J.B., 1971. Effects of hypothalamic implants of gonadal steroids on
courtship behaviour in Barbary doves (Streptopelia risoria). J. Endocrinol. 50, 97–
113.
253
Kirn, J.R., Schwabl, H., 1997. Photoperiod regulation of neuron death in the adult
canary. J. Neurobiol. 33, 223–231.
Leitner, S., Voigt, C., Gahr, M., 2001a. Seasonal changes in the song pattern of the
non-domesticated island canary (Serinus canaria), a field study. Behaviour 138,
885–904.
Leitner, S., Voigt, C., Garcia-Segura, L.-M., Van’tHof, T., Gahr, M., 2001b. Seasonal
activation and inactivation of song motor memories in wild canaries is not
reflected in neuroanatomical changes of forebrain song areas. Horm. Behav. 40,
160–168.
Leonard, S.L., 1939. Induction of singing in female canaries by injections of male
hormone. Proc. Soc. Exp. Biol. Med. 41, 229–230.
Metzdorf, R., Gahr, M., Fusani, L., 1999. Distribution of aromatase, estrogen receptor,
and androgen receptor mRNA in the forebrain of songbirds and nonsongbirds. J.
Comp. Neurol. 405, 1–15.
Montagu, G., 1802. Ornithological Dictionary, J. White, London.
Naftolin, F. et al., 1975. The formation of estrogens by central neuroendocrine
tissue. Recent Prog. Horm. Res. 31, 295–319.
Nicholls, T.J., Goldsmith, A.R., Dawson, A., 1988. Photorefractoriness in birds and
comparison with mammals. Physiol. Rev. 68, 133–176.
Nottebohm, F., Nottebohm, M.E., Crane, L.A., 1986. Developmental and seasonal
changes in canary song and their relation to changes in the anatomy of songcontrol nuclei. Behav. Neural. Biol. 46, 445–471.
Phoenix, C.H., Goy, R.W., Gerall, A.A., Young, W.C., 1959. Organizing action of
prenatally administered testosterone propionate on the tissues mediating
mating behavior in the female guinea pig. Endocrinology 65, 369–382.
Schlinger, B.A., Arnold, A.P., 1991. Brain is the major site of estrogen synthesis in a
male songbird. Proc. Natl. Acad. Sci. USA 88, 4191–4194.
Turek, F.W., Desjardins, C., Menaker, M., 1976. Antigonadal and progonadal effects
of testosterone in mae house sparrows. Gen. Comp. Endocrinol. 28, 395–402.
Wingfield, J.C., Farner, D.S., 1975. The determination of five steroids in avian plasma
by radioimmunoassay and competitive protein-binding. Steroids 26, 311–327.
Wingfield, J.C., Farner, D.S., 1993. Endocrinology of reproduction in wild species. In:
Farner, D.S. et al. (Eds.), Avian Biology. Academic Press, San Diego, pp. 163–327.
Wingfield, J.C., Hegner, R.E., Dufty, A.M., Ball, G.F., 1990. The challenge hypothesis—
theoretical implications for patterns of testosterone secretion, mating systems,
and breeding strategies. Am. Nat. 136, 829–846.
Wingfield, J.C., Jacobs, J., Hillgarth, N., 1997. Ecological constraints and the evolution
of hormone-behavior interrelationships. Integrative neurobiology of affiliation.
Ann. N. Y. Acad. Sci. 807, 22–41.
Wingfield, J.C., Lynn, S.E., Soma, K.K., 2001. Avoiding the costs of testosterone:
ecological bases of hormone-behavior interactions. Brain Behav. Evol. 57, 239–
251.
Wingfield, J.C., Moore, M.C., 1987. Hormonal social and environmental factors in the
reproductive biology of free-living male birds. In: Crews, D. (Ed.), Psychobiology
of Reproductive Behavior an Evolutionary Perspective. Prentice Hall, New
Jersey, pp. 149–175.