166 BIEB Lecture 5
BIEB 166: ANIMAL BEHAVIOR & COMMUNICATION
LECTURE 5 READING: Behavioral Endocrinology
I.
Hormonal control of behaviors
When behavior is hormonally triggered, there is typically a greater delay in the onset of a
behavior than when it is triggered by neurons alone. Wigglesworth (blood-sucking bugs,
Rhodnius) was able to experimentally locate the source of hormones by taking advantage of
the critical period or delay between hormone secretion and physiological consequence. In
Rhodnius, distension of the gut after a blood meal signals the brain to produce moulting
hormones that cause the insect to moult. Wigglesworth is famous for showing that whether
an insect molts to another juvenile instar, or whether it molts to a winged adult stage,
depends on the concentration of a hormone, which he called juvenile hormone in the
insect's blood. Juvenile hormone normally decreases in the final instar before the larvae
becomes an adult. Kopec demonstrated this quite convincingly by implanting a gland from a
juvenile during the final molt and this led to a giant juvenile, not an adult. Kopecs research
led to the discovery of the molting hormone, or ecdysone. A hormone, secreted by the brain,
is necessary for the gypsy moth to pupate into a chrysalis. The brain hormone acts between
the seventh and tenth day of the instar.
II.
Hormones
In many animals, motivation is principally controlled by a combination of neural
and glandular mechanisms called the endocrine system. The endocrine system
operates by the secretion of hormones, chemical messages from neurons or
glands that influence target organs. The nervous system stimulates only the cells
that it innervates and produces only short-term behavioral responses. Hormones,
however, travel through the circulatory system to every cell in the body, and can
produce both short and long term changes throughout the organism.
The most important and complex function of the endocrine system is the regulation of
reproductive behavior. At the start of the breeding season, environmental stimuli such as
increasing day length and higher temperatures are perceived by the sensory system and
processed by the brain. This triggers the hypothalamus, which causes the anterior pituitary to
release a class of hormones called gonadotropins that influence the gonads, the testes and
ovaries. In response to gonadotropins the gonads begin to develop from their non-breeding state,
and in turn produce steroid hormones, testosterone from the testes or estrogen and progesterone
from the ovaries. In some cases, these circulate back to the hypothalamus, providing negative
feedback to tell it that the gonads have gotten the message and completed
Page 1
BIEB 166 Lecture 5
their development, so further secretion of gonadotropins should be turned off. Feedback
loops, both positive and negative, are characteristic of endocrine regulatory mechanisms.
Steroid hormones also affect the morphological characteristics of males and females, leading to
the seasonal development of secondary sexual characteristics. Most importantly, steroids
circulate back to the higher centers of the brain, where they influence sexual behavior.
Finally, after all of these internal processes, we get back to the external behavior of the
animals. The sexual responses of both male and female in turn act as stimuli, perceived by
and processed in the brains of members of the opposite sex, that stimulate further
hypothalamic and pituitary activity, the secretion of additional hormones, and the regulation
and coordination of behavioral interactions between mates.
III.
Reproductive feedback loops:
The comparative psychologist Daniel Lehrman studied the ring dove as a model of
reproductive regulation, and thus helped found the study of behavioral endocrinology. His
work is an interesting example of progress in science. When he began his work in the 1950's
and 1960's the technology for directly monitoring hormone levels in living animals did not yet
exist. So he investigated hormones and behavior by looking at anatomical indicators of
hormones, for example the development of reproductive traits, and by injecting animals with
hormones. Through these methods he worked out detailed story of the regulation of sexual
and parental behavior in ring doves that became a textbook example. Some of his
intellectual descendants, particularly Rae Silver and Mei Cheng of Columbia and Rutgers,
respectively, have modified and contributed to this original work.
This reproductive cycle is divided into several phases. Pairs of ring doves will not begin
courtship until presented with two essential stimuli: (1) the appropriate light-dark cycle
(photoperiod) and (2) some suitable nest material, usually bits of straw or paper.
Exposure to these stimuli initiates phase I of the reproductive cycle, causing the
hypothalamus to trigger the anterior pituitary to secrete gonadotropins, which in turn
stimulate gonad growth, the maturation of eggs in the female's ovaries, and the secretion
of steroid hormones by the gonads. The steroids produce behavioral changes, causing
both birds to initiate a stereotyped courtship display involving cooing sounds and bowing,
for about two days. Each bird also stimulates the other. The female especially stimulates
the male, probably as much as his photoperiod-induced hormone changes.
In phase II, the birds construct a nest together over a period of about a week, during
which they mate. Rising progesterone levels in the female precede egg-laying and
ready the female for incubation, once the eggs are laid, turning off courtship. Both
female estrogen and progesterone drop after the eggs are laid, as does male
testosterone. Hormonal changes in the female are known to play a causal role in the
transition to incubation. Male incubation, however, seems to be controlled by cues
from the female, not acting through any obvious hormonal intermediates.
In phase III, both sexes incubate for two weeks. Incubation leads to the secretion of the
pituitary hormone prolactin, which inhibits the action of gonadotropins. In female mammals,
prolactin stimulates the development of mammary glands and secretion of milk, but the
same hormone affects pigeons of both sexes. Under its influence, cells lining the birds'
Page 2
BIEB 166 Lecture 5
throat or crop develop and begin to produce a nutritious white liquid called "crop
milk," which the birds (male & female) will regurgitate to feed their chicks.
Prolactin also appears to inhibit the further production of gonadotropins. It also
maintains incubation behavior in the later stages of incubation.
Finally, in phase IV, the eggs hatch and both male and female regurgitate crop milk to
the chicks until they are old enough for solid food, about two weeks after hatching. The
maturation and departure of the young at about 25 days of age causes prolactin
secretion to decline, terminating the inhibition of gonadotropins. Now, if the pair remains
together, gonadotropins will be secreted again and the cycle resumes. The main theme
in this story is that each behavior in the sequence sets off endocrine and/or motivational
changes that then alter receptivity of birds of each gender to the stimuli occurring in the
next step of the sequence. It is a cascade of events.
For example, the birds will ignore artificial nests or dummy eggs presented before they have
built their own nest; in fact, they often start their own nest on top of the artificial one, or take
it apart to provide material for constructing a nest that to us appears identical. But once they
are stimulated by the act of nest construction, you can give them dummy eggs and release
incubation, even before the female has laid her own eggs. However, this will only take place
if the female has been exposed to the correct hormones. Removing the female's ovaries to
prevent hormone production can test this. If you do this, but give her the correct
environmental stimuli, such as nest materials, a partner and later, eggs, the females will not
incubate. If we inject the females with estrogen alone or progesterone alone, the females
also will not incubate. However, ovariectomized females injected with both hormones at the
correct times and given the correct environmental stimuli will incubate.
It has also been shown that if the nest is removed after the female has laid eggs, so the
birds cannot incubate, prolactin production is not stimulated and the crop does not develop- individuals without crops clearly cannot proceed to feeding of hatched offspring. But if a
male is separated from his nest and mate but can watch her incubating though a glass
window, he produces crop milk on schedule. Thus the simple chains we discussed earlier of
KS --> IRM --> FAP are much expanded and complicated in the ring neck dove example.
To recap, sensory stimuli, including environmental cues like day length and social stimuli
from other animals' displays act upon neural and endocrine mechanisms to influence
behavior through an intricate series of interactions and feedback loops. Remember that
these loops often differ between the sexes.
In general, the endocrine regulation of reproduction is remarkably consistent in many
vertebrates, from fish to mammals. At the start of the breeding season, external stimuli are
received by the sensory system, processed by the brain, which stimulates the
hypothalamus to initiate the reproductive cycle. These external stimuli usually consist at
least in part of environmental changes such as increasing day length and higher
temperatures. But as we have seen in the ring dove, social stimuli can also be involved,
and can be extremely important in coordinating the behavior of the two sexes.
Page 3
BIEB 166 Lecture 5
IV. Effects of social stimulation on reproductive
biology Anolis lizards
This takes us to our next topic, the effect of social stimuli on reproductive biology, using the
anolis lizard as our first example. The regulation of reproduction in the so-called American
chameleon or Anolis lizard was the subject of a series of experiments by David Crews.
For this example, we will be using the concept of the ethogram, introduced
earlier. An ethogram is simply a record of behavioral activities.
Crews established the reproductive ethogram for this species: Males of populations living in
Florida emerge from winter dormancy in the beginning of February and establish breeding
territories by fighting and with aggressive displays, particularly the extension of the colored
throat patch or dewlap, apparently a fixed action pattern. Aggressive territory owners also
bob up and down in a stereotyped manner, raise small crests and develop dark eye spots.
An intruding male responds to this by flattening its body against the ground in a submissive
posture; if not, it is attacked. About a month after males become active, females emerge and
are courted by males, with displays that also feature dewlap extension and bobbing but
without the crest erection and dark eye spot of aggressive displays. Females repeatedly
mate and lay one egg at a time, every 10 - 14 days through the summer, until August.
Dissection reveals that males have enlarged active testes from January to July, and female
ovaries contain active egg follicles from February to August. Thus the endocrine activity of
the two sexes is synchronized; and it is synchronized not only by their common exposure to
the environmental stimuli of spring, but also by their behavior.
Females collected in February and housed in a laboratory, exposed to a spring p hotoperiod
cycle of 14 hours of light and 10 of hours darkness, begin to develop their ovaries and
produce vitellogenin, the precursor of egg yolk. However only about 80% of females
produce vitellogenin they if are kept in isolation or with a group of other females or if they
kept with a castrated male who does not perform any head bobbing or dewlap extension. If
one male is added to the group, then 100% of the females' ovaries rapidly develop to
complete maturity. It is the male's courtship behavior that stimulates the female's brain and
hypothalamus to release gonadotropins, resulting in ovarian development.
You might expect that providing the stimulus of more than one male would lead to still
more rapid ovarian growth. Exactly the opposite occurs: if several males are placed in a
small cage with a group of females, they spend all their time fighting or in aggressive
displays instead of courting, and the percentage of females with developed ovaries
shows some initial increase, but then actually declines to less than the level observed in
isolated females. Thus if a wild female emerged from hibernation too early and found the
males still fighting to establish territories, her ovaries would not begin to develop until the
males were ready for mating. If a castrated male is added to a group of females, about
80% of them produce vitellogenin; the same percentage as observed when females are
isolated from males. Unable to produce testosterone, the castrated male does not court
and does not stimulate enhancement of ovary development. However, if testosterone is
injected into the castrated male, he performs normal courtship displays and induces
ovarian development in all of the females.
As a control, some males were injected with cholesterol, which happens to be chemically
somewhat similar to testosterone but has none of its effects. These control males did not
Page 4
BIEB 166 Lecture 5
enhance female ovarian development. This is a nice bioassay showing definitively that it is
the male's behavior, not just his presence that stimulates the females. Moreover, this is a
standard experimental approach in behavioral endocrinology to test for a causal relationship
between a hormone and a behavior: remove the organ that produces the hormone and see
if the behavior disappears; then replace the hormone and see if the behavior returns. It is
also very important to conduct appropriate sham treatments in the controls. Crews did this
by injecting the castrated anoles with cholesterol in order to be sure that it was not just
getting an injection that caused courtship behavior to return in castrated anolesit had to
be an injection of testosterone.
Plainfin midshipman fish
In a second example of how social conditions may influence reproductive behavior,
consider the odd case of the male plainfin midshipman, studied by Andrew Bass at
Cornell University. These are teleost fish that live in intertidal and subtidal zones along
the western coast of North America. From late spring to early summer, the "song " of the
male plainfin midshipman can be heard after nightfall as a low-frequency hum-- a
droning on to human ears, but highly attractive to female midshipman fish, who seek out
the singing males and mate with them. The singing males build nests under rocks, and
invite the females in to deposit their eggs. In addition to their singing, males also emit
grunt trains and growls to expel intruders from the nest.
Now here's the interesting twist. Not all male midshipmen are singing fish that build
nests. Only Type I males sing. However, type II males also exist in the population. These
little fish lie perched outside of the type I males' nests, and sneak into the nest to deposit
sperm after a female has spawned. These sneaky type II males resemble females-- they
are much smaller and slimmer than the singing Type I males. However, they have big
gonads. The average ratio of gonad to body weight in type II males is 9 times greater
than in type I males. Type II males may invest up to 15% of their weight in testes,
compared with only 1% in type I males. When Bass and his students looked for
hormonal differences between the different morphs, they found that Type I males contain
high amounts of a unique form of testosterone called 11-ketotestosterone. On average,
type I males contain five times as much 11-keotestosterone than they do testosterone,
whereas type II males contain only testosterone. This hormonal distribution is similar in
all teleosts with two distinct male morphs. Thus it seems likely that 11-ketotestosterone
is more potent than testosterone in supporting courtship behaviors such as humming.
Bass went on to identify the central pattern generator giving rise to the courtship
humming. To do this, he used a staining compound called biocytin, which consists of
the amino acid lysine and biotin, a naturally occurring protein in neurons. When crystals
of biocytin are applied to the cut ends of the motor neurons, the cells that innervate and
stimulate the sonic muscles, the biocytin is carried backwards from the nerve ending
along an axis to the parent cell body, and onwards, across the synaptic spaces, etc.
Since the biocytin stains the cells brown, Bass and his colleagues could thus visualize
the entire vocal neuronal circuit required for stimulating the sonic muscle.
The different morphs (Type I male, type II male and female) all contain the same circuitry,
and the same number of nerve cells devoted to that circuitry. However, the pace-maker
motor neuron circuit in type I males fires at a frequency about 15 to 20 % higher than type
Page 5
BIEB 166 Lecture 5
II males. The cell bodies, dendrites, and axons are also one to three times larger in type I
males than in type II males and females. Thus the differences within and between sexes in
the organization of the vocal motor system depend upon a divergence in the morphological
and physiological properties of individual nerve cells. The differences between morphs is
initiated very early in development, and once committed to a particular pathway, an
individual cannot shift from one morph to the next. So what controls the production of type I
or type II males? So far, it appears that there is density dependent regulation: under
sparsely populated conditions, more type I males are produced, whereas as the population
density increases, so does the percentage of type II males.
V.
Field behavioral endocrinology
Most of our knowledge of the importance of hormones in regulating behavior comes
from experiments conducted in the lab in which the effects of non-hormonal cues in
the social and physical environment have been intentionally removed or controlled.
But in the last few years a new area of research has emerged: field behavioral
endocrinology. The biologist, Michael Moore (1987), decided to see what would
happen if you studied animals whose hormone levels had been manipulated under
field conditions, and for a start, he looked at the effects of castration on male spiny
lizards, Sceloporus jarrovi in their natural Arizona habitat.
He found that castration reduced dramatically, but did not completely abolish,
territorial and sexual behavior in the experimental males when compared with sham
operated controls. Certain behaviors, such as the establishment and occupation of
territories, were relatively unaffected by castration. This work is important because it
had long been believed that elevated levels of gonadal sex steroid hormones are
necessary prerequisites for the expression of adult reproductive behaviors. This and
related work on other organisms showed that in some cases, non-hormonal cues
such as social and environmental cues can either override or substitute for
hormonal cues when the latter are removed. So the mechanisms controlling
aggressive and reproductive behavior are more diverse than was once thought.
VI.
Hormones, behavior, and life histories
In fact David Crews (1987) has pointed out that different relationships between
hormones and behavior may be adaptive, associated with the evolution of different life
histories. An animal's life history is essentially its pattern of growth and reproduction,
with the emphasis on how it allocates its limited resources throughout its life to
maximize lifetime reproductive success. The pattern we have emphasized so far is the
associated reproductive pattern, in which reproductive behavior and gonadal activity
peak at the same time. This pattern is expected in species living in environments in
which favorable breeding conditions are predictable and/or prolonged, such as
organisms living in some moderate temperate zone environments.
But what if an animal lives in an environment with a very narrow window for breeding?
In this case it might not have time to produce eggs and/or sperm in the period just
before breeding must occur. Consequently, it may have its peak period of gonadal
activity at a different time and store gametes in readiness for mating. Since mating
does not coincide with peak gonadal activity, we would not expect it to be controlled
by gonadal hormones. Crews refers to this as a disassociated reproductive pattern.
Page 6
BIEB 166 Lecture 5
Just this sort of pattern occurs in the red-sided garter snake, Thamnophis sirtalis in Northern
Manitoba, Canada. These snakes spend much of the winter in limestone caverns
(hibernacula) that serve as subterranean hibernacula. The snakes emerge in the spring,
which can be remarkably late in the year. They engage in most of their mating activity in the
next few weeks. Following mating, they move off to feeding grounds to fatten up in the brief
summer before heading underground again in the fall. In this species, male gonadal activity
peaks in the summer, well after breeding, when there is plenty of food available to fuel the
gonads; sperm are then stored through the winter for spring matings. Thus androgen, or
male steroid, levels are low at the time of mating. Numerous studies have been conducted
looking for a role for steroids in the short-term activation of male reproductive behavior, but
none has been found. Rather, the control of reproductive behavior appears to be strictly
neuronal, controlled from the brain. Again, we see that the relationship between hormones
and behavior can be adaptively modified.
Crews does not claim that testosterone plays no role whatsoever in garter snakes. We
know that testosterone is important in the expression of garter snake reproductive
behavior over the medium term: exposure to testosterone in the summer is important if
full reproductive behavior is to be released the following spring, even though no
testosterone need be present then. And it is almost certainly true that testosterone has
an organizational effect on the development of the brain of the garter snake, as it does
in so many other organisms, particularly mammals. Activating effects of a hormone
(such as testosterone) are those in which the expression of a certain behavior requires
the constant and immediate presence of the hormone.
VII.
Forms of hormonal action
However, hormones are not the complete story when it comes to explaining reproductive
behavior. A student of Michael Moore's, Catherine Marler (Physiological Zoology 1989)
performed an experiment in which she elevated testosterone levels of free-living spiny
lizards by giving males subcutaneous implants of testosterone and then releasing them
back into the field. So she really pumped these little guys up. She found that more
aggressive T-implanted males expended more energy by having a longer daily activity
period and performing more territorial displays and movements. These males made fewer
foraging attempts, caught fewer insects, and had lower gut content massall suggesting a
trade-off between territorial aggression and foraging ability. Implanted males also have
higher mortality and may suffer higher predation due to their conspicuous territory defense.