9
BEHAVIORAL ADAPTATION
TO CHANGED CONDITIONS OF LIVING
Either chance and selection can explain everything or else
behavior is the motor of evolution.
J.
Piaget
One cannot
intelligently discuss behavior and structure separately.
Behavior is what an animal does with its structure; structure
is what an animal uses to behave.
H.E.
Evans
Metazoans
respond to changed conditions of living by adaptively
modifying behavioral, morphological, physiological, and life
history characters. Being the most plastic of all the
phenotypic characters, animal behavior is the first to
adaptively change in response to the changed environment.
Behavior consists of motor responses necessary for performing
vital functions and for adapting metazoans to the changed
conditions in environment. Animal behaviors are innate or
learned. Innate behaviors can be modified by learning and
learned behaviors can evolve into innate behaviors. Although
genes and other non-genetic factors are necessary for and are
involved in performing behaviors, genes are not causes of
behavior, i.e. they are not both necessary and sufficient, for
performing animal behavior. Animal behavior is determined by
neural mechanisms, essentially related to the activity of
specific neural circuits. Neural circuits for specific
behaviors are highly conserved across metazoan taxa and they
may be conserved even after the loss of the structures that
were used for performing them. The frequently observed
temporal correlation between the appearance of adaptive
behaviors and changes in morphology suggests that a causal
relationship may exist between neural circuits determining the
animal behavior and those determining adaptive changes in
morphology. This may be of paramount importance for the
evolution of metazoans.
Adaptation to Changed
Conditions of Living
Under conditions of environmental
stress, metazoans may display two types of responses:
immediate or delayed response.
Generally, immediate
responses are neuroendocrine and behavioral responses aiming
at neutralizing or avoiding noxious factors of the changed
environment. Examples of immediate adaptive behaviors are
avoiding reflexes, relocation, and migration.
Delayed responses may lead to
changes in the morphology, physiology, and life history of
animals.
Delayed responses may be
intragenerational, i.e., taking place within the lifetime
of the affected individual, which are normally aimed at
restoring the disturbed homeostasis, repairing damaged
structure (regeneration), or adaptively changing the
morphology of the affected or challenged organs or parts. Such
adaptive changes, as a rule, are not inherited in the progeny,
which normally reverts to the pre-parental species-specific
morphology. Sometimes delayed responses may be
transgenerational, i.e., morphological and physiological
adaptive changes that arise first in (or are transmitted to)
the offspring of the affected or challenged individuals are
inherited in successive generations.
A drastically changed
environment, by disturbing the homeostasis, is a challenge to
the CNS, which is in control of the homeostasis. External
stimuli transmitted in the form of electrical spike trains
present a problem, to which the CNS, may respond in two main
ways mentioned above. It may change the behavior in order to
avoid or circumvent as much as possible the harmful elements
or effects of the adversely changed environment by settling in
a more hospitable environment. This behavioral change is
intentional: the animal avoids what is clearly harmful to it,
implying a “prediction of benefit” from its new behavior. The
second way in which the CNS expresses its problem-solving
capability is by inducing adaptive changes in morphology
(phenotypic plasticity) and/or function in order to adapt them
or their offspring (predator induced defenses and
transgenerational plasticity in general) to the changed
conditions of living. Unlike the change in behavior, adaptive
morphological and physiological changes happen “effortlessly”
and unperceived by the animal.
Due to the fact that from all the
major phenotypic features, animal behavior is beyond compare
the most plastic, the adaptive change of behavior is the first
step animals undertake under conditions of adversely changed
environment. Adaptive changes in morphology, physiology, and
life history generally come later, if ever.
Neural Basis of Animal Behavior
Types of animal behavior vary
over a wide range, from strictly innate behaviors to those
that, although innate, can be modified by learning, to
behaviors that are exclusively learned. Innate behaviors, with
instincts as their extreme form, appear in fully functional
form for the first time since they were performed. Instinctive
behaviors of an animal are just as stereotyped and
characteristic for its species as its morphology is, hence one
might expect to find a similar logic underlying the hereditary
mechanisms that specify behavior and morphology. Yet, whereas
morphological development has now largely succumbed to the
attack of the classical forward genetics in a few model
organisms, the same approach has made only modest inroads into
the developmental origins of complex innate behaviors.
Innate behaviors are automatic
stereotyped actions of the organism in response to external
releasing stimuli, which trigger innate releasing
mechanisms (from the German angeborenes auslösendes
Schema – innate releasing schema)
to produce motor patterns generally known as fixed
action patterns (FAP). The pathway from reception of the
releasing stimulus, via the central nervous system, to motor
neurons represents the neural circuit responsible for FAP.
Any FAP is based on the presence and activation of a specific
neural circuit.
In his time, Darwin observed:
As in repeating a well-known
song, so in instincts, one action follows another by a sort of
rhythm; if a person be interrupted in a song, or in repeating
anything by rote, he is generally forced to go back to recover
the habitual train of thought: so P. Huber found it was with a
caterpillar, which makes a very complicated hammock; for if he
took a caterpillar which had completed its hammock up to, say,
the sixth stage of construction, and put it into a hammock
completed up only to the third stage, the caterpillar simply
re-performed the fourth, fifth, and sixth stages of
construction. If, however, a caterpillar were taken out of a
hammock made up, for instance, to the third stage, and were
put into one finished up to the sixth stage, so that much of
its work was already done for it, far from deriving any
benefit from this, it was much embarrassed, and, in order to
complete its hammock, seemed forced to start from the third
stage, where it had left off, and thus tried to complete the
already finished work. (Darwin, 1859g1)
The nervous system is organized
in neural circuits, which represent functional units rather
than anatomic structures. It is generally believed that
Neural circuits are the basis of
all behavior, from simple reflex withdrawal away from a
noxious to a complex mating dance. (Delcomyn, 1998c)
Various complex behaviors may
result from the interaction of various functionally (not
necessarily anatomically) linked circuits. Many neurons and
their gene products (neuropeptides, neurotransmitters, etc.)
are involved in performing these behaviors and a neuron may be
involved in more than one behavior, i.e. in more than one
circuit.
Konrad Lorenz observed that upon
seeing an egg outside its nest, goose tries to roll it back to
the nest. The release of this FAP is triggered by a “sign
stimulus”, which may be represented not only by the goose egg
but also by other objects even those remotely resembling it.
FAPs are hardwired in the brain
as it has been experimentally demonstrated by Evon Balaban
(1997). By transplanting various parts of the neural tube of
the Japanese quail (Coturnix coturnix) into domestic
chicken (Gallus gallus domesticus) embryos, he
succeeded in producing chimerae chicken exhibiting quail
crowing and head movements, both of them subcomponents of the
innate behavior originating from 2 different regions of their
brain (Balaban, 1997). Using similar brain transplant
techniques investigators have been able to transfer an inborn
perceptual auditory preference between the same above species
(Long et al., 2001).
Innate behaviors may heritably
change without changes in genes. One-century-long attempts to
find genes responsible for particular behaviors have failed.
In the meantime, successes in understanding the nature of
learning and memory show that animal behavior may be
determined by special processing and organizing properties of
the nervous system. All these complex and flexible functions
of the nervous system cannot be reduced to the structure and
frozen protein-coding capabilities of one or any number of
genes. Even the simple fact of the extreme modifiability of
the animal behavior within animal’s lifetime, rejects any
reductionist concept of genes as determinants of animal
behavior.
Many innate behavior patterns are
active immediately after birth. So, e.g. after giving birth to
its pups, the mother rabbit, Oryctolagus cuniculus,
releases from its nipples a pheromone and the pups are born
with a fully established neural circuit for identifying that
chemical cue and respond by searching behavior for grasping
the nipples. That the maternal CNS is involved in the cue
release is proven by the fact that injection of prolactin
stimulates maximum release of the chemical cue (Moncomble et
al., 2005).
In the marine mollusc Aplysia,
egg-laying consists of a number of FAPs (fixed action
patterns): extrusion of a long string of eggs from the
reproductive duct, taking of the egg string in the mouth,
stereotypic head movements intended at pulling the egg string
from the duct, coiling it into a mass glued together by
secretions from its mouth, affixing the entire mass of eggs on
a solid substrate with a strong head movement.
It was discovered that essential
for performing this complex egg-laying behavior is a
neuropeptide composed of 36 amino acids, the egg-laying
hormone, secreted by certain neurons in the nervous system of
the mollusc. Later it turned out that the egg-laying hormone
was only a part of a precursor molecule composed of almost 300
amino acids from which other neuropeptides are synthesized,
which serve as neural signals controlling other FAPs of
egg-laying behavior (Purves et al., 1992f).
The overwhelming majority of
behaviors studied in vertebrates are related with the function
of neural circuitries in the hypothalamus and with the
hypothalamic-pituitary-target endocrine gland axes:
Specialized neuroendocrine
circuits for innate behaviors thus seem to process sensory
information relevant to ethological contexts and influence
sensory perception and processing; integration by these
circuits of multiple pathways of information relevant to
different behaviors determines the behavioral state of the
animal. (Manoli et al., 2006)
A typical example of an innate
behavior in mammals is suckling reflex, involving the
hypothalamic-pituitary axis. Stimulation of nipples by young
mammals initiates sensory impulses that reach the CNS and via
the hypothalamus end in the PN (pars nervosa) of the
pituitary stimulating secretion of oxytocin in the blood
within a few seconds. Oxytocin causes contraction of cells in
the mammary gland, squeezing milk down to the nipple in less
than 1 minute (Gorbman and Davey, 1991a).
One of the most widespread innate
behaviors is the annual migration and repatriation in large
number of animals, invertebrates (insects, crabs, etc.) as
well as vertebrates (fish, reptiles, birds and mammals). Many
of the animals take the round trip journey although a visible
reason or evolutionary advantage for undertaking the trip
cannot always be identified:
The young of the bronzed cuckoo,
one month after being left behind by their parents in New
Zealand, fly 1200 miles over water to Australia and then 1000
miles north to the Solomon and Bismarck islands. The journey
seems unnecessary, since there is no great need to escape the
mild winter in New Zealand, and it is unclear why the parent
birds are required their offspring to navigate by themselves.
(Wesson, 1991b)
Indeed, it is not easy to imagine
an evolutionary pressure that would reasonably be responsible
for the evolution of the this migration instinct. Emergence of
such amazing innate behaviors seems to exclude gradual
evolution.
Following Darwin, most biologists
believe that innate behaviors have evolved from learned
behaviors (migration enforced by various environmental
factors). Migration of birds and other vertebrates and
invertebrates, the complex spider web-building behavior, or
social behavior in ants, bees and other animals, are complex
innate behaviors, which could reasonably be explained as
resulting from learned antecedents:
Learning…is one of the standard,
off-the shelf programming tricks available to evolution - and
despite the usual dichotomy, this kind of learning is the
epitome of the instinct. (Gould, 1982d).
Birds are known to have an innate
ability to recognize their conspecific song and to respond
more strongly to the song of conspecifics than to the song of
other birds, even when not all phrase types are present in the
song and when the song is played in reverse. Identification of
specific neurons responding differently to conspecific and
heterospecific songs suggests that circuits for song
recognition in these birds are established
experience–independently, i.e. during the embryonic life,
probably perfected by learning during the post-natal life
(Whaling et al., 1997). Evidence also has been presented
showing that an innate song recognition and preference at a
subspecific level also exists (Nelson, 2000).
Behavior patterning is determined
in the CNS according to the sensory input from the animal’s
periphery and the environment, but in the rhythmic behaviors
(swimming, flight, and chewing), their patterning is totally
of central origin and unmodified by sensory input. Their
neural circuits are relatively hardwired (Gillette, 1991).
This has led to the concept of the CPG (central pattern
generation) as basis of FAPs (fixed action patterns).
However, not all neural pattern
generators are central. So, e.g., the pattern generator of
feeding behavior in gastropods is distributed between the
buccal and cerebral ganglia and is, therefore, modified by
sensory input. Besides, the pattern generation of the buccal
ganglion can sustain more than one type of rhythmic patterns (Croll
et al., 1985a; Suesswein and Byrne, 1988).
Feeding (ingestion)
behavior in Pleurobranchaea marine slugs is under
control of 18 descending interneurons playing the role of
command neurons. These include paired MCG (metacerebral giant
neuron) and the PCNs (paracerebral neurons) of three types (PCp,
PSE, ETI1) (figure 9.1).
Figure 9.1. A partial,
hypothetical model for how motor program switching is
accomplished in the buccal motor system of Pleurobranchaea.
According to this model, separate command systems for
ingestion (top left) and egestion (top right) converge on a
common central pattern generator, which provides oscillatory
feedback (excitation and inhibition) to the command neurons.
The command pathways also provide appropriate biases to the
motor pools. Solid triangles, excitatory connections; solid
circles, inhibitory connections.
Abbreviations: VWC,
ventral white cell; AV, anterior ventral neuron; MCG,
metacerebral giant neurons; PCp, phasic paracerebral neurons;
PSE, polysynaptic excitors of the PCs; ETII , type
II electrotonic neurons (From Croll et al., 1985b).
Progress has also been made in
identifying interneurons and motor neurons in the circuits for
the basic locomotory movements (swimming, crawling,
shortening, and bending) in leech (Fan et al., 2005).
In some neural centers, in
invertebrates as well as in vertebrates, certain neurons show
after-discharge or repetitive firing as part of a reflex
response. The animal can swim, fly, or chew in the complete
absence of sensory feedback, as it is demonstrated in
deafferented animals (Delcomyn and Prosser, 1991).
Among the
best known examples of CPGs (central patterning generators)
for FAPs is that of locomotion in molluscs. The CPG for swim
escape of the marine mollusc, Tritonia diomedea is
activated as soon as the slug comes in contact with the
predatory seastar Pycnopodia helianthoides and it can
also be activated by stimulating any of a group of peripheral
nerves (Frost et al., 2001; figure
9.2).
However, artificial injection of depolarizing current pulses
into the interneuron C2 (a crucial member of the swim escape
central pattern generator) cannot stimulate the swim escape
CPG because it does not mimick the interneuron’s own inherent
spike frequency adaptation (SFA). In order for the swim escape
to occur it is necessary to change the SFA. This property of
the circuit changes, the firing rate of C2 is regulated and
swim escape behavior occurs when serotonergic interneurons DSI,
intrinsical to the circuit, are stimulated (Katz and Frost,
1997).
Figure 9.2. The
Tritonia escape swim and its underlying circuit. A.
Upon contact with a suitably aversive stimulus, such as the
tube feet of the seastar Pycnopodia helianthoides,
Tritonia respond with an escape swim consisting of a
series of alternating ventral and dorsal whole-body flexions.
The photograph shows an animal at a moment of maximum dorsal
flexion. B. The known swim circuit. Solid lines
represent direct, monosynaptic connections, broken lines
represent indirect connections, or connections not yet
confirmed to be monosynaptic. Synaptic symbols: lines,
excitatory; black circles, inhibitory; lines and circles,
multiple component monosynaptic connections. "VSI" represents
both VSI-A and VSI-B; the exact connectivity shown is for VSI-B
only. The known number of neurons of each type on each side of
the brain are: S-cells, 80;
Tr1, 1; DRI, 1; DSI, 3; C2, 1; VSI, 2; FNs, 55.
Abbreviations: S,
(sensory) afferent neurons; TR1, pre-CPG trigger type1
interneuron; DRI, dorsal ramp interneuron; DSI, dorsal swim
neuron; CPG, central pattern generator; C2, cerebral cell 2;
VSI, ventral swim interneuron (From Frost et al., 2001).
Neural circuits often interact
and may interfere with each other’s activity. Sometimes
activation of a circuit may automatically inactivate the
circuit for another behavior. Such is the case of the dominant
escape swimming behavior in the marine predator sea slug,
Pleurobranchaea. Escape swimming is an avoidance behavior
in the predatory sea slug Pleurobranchaea californica.
Stimulation of its swim escape circuit leads to automatic
inhibition of the feeding circuit. The swim central pattern
generator (CPG) consists of neurons A1, A3, A10, and IVS
, which produce the swim motor pattern, and serotonergic
neurons with modulator arousal functions on the pattern
generator. Activation of the circuit suppresses feeding in
this slug by inhibiting feeding command neurons (figure 9.3)
(Jing and Gillette, 2000).
In response to environmental
stimuli, i.e. to particular types of sensory information,
animals secrete neurohormones and other hormones that act as
gain-setting devices, biasing the animal behavior toward
particular stereotypical responses (male/female behavior,
fight or flee, explore, search for food, etc.). By acting both
in the nervous system and on effector organs, hormonal
substances can modify the input processing, and the output of
specific subsets of neurons to enhance the probability of
specific outcomes (Kravitz, 1988).
Changes in animal behavior result
from reversible neuromodulations rather than any changes in
genes. Similar behaviors may be produced by convergent rather
than homologous circuits. Changes in the configuration of the
neural network may change the behavioral output of
circuitries; neuromodulation, the state of the neural network,
thus, provides behavior with flexibility and small changes in
neural pathways, involving no changes in genes, may result in
dramatic changes in behavior (Nishikawa, 2002).
......................................................................
......................................................................
Learned Behaviors Evolve into Innate Behaviors
Despite the possibility of being modified later during the
life, innate behaviors are prenatally hardwired in the nervous
system. Learned behaviors as well, are determined by neural
circuits, often in appropriate responses to changes in the
environment. So, e.g., some nonvolatile male mice pheromones
are innately attractive to female mice but volatiles from
male-soiled bedding are not. However, repeated exposure of
female mice to male-soiled bedding “learns” them to confer
pheromonal properties to the male urine-borne volatiles.
Learning, thus, might confer
“pheromonal” properties to every odorant, as a result of the
ability of the CNS “to relate virtually any stimulus to any
adaptive result” and this suggests that innate and
acquired olfactory attractiveness relies on similar neural
mechanisms (Moncho-Bogani et al., 2002).
Both innate and learned behaviors are products of the activity
of neural circuits. Although differences between these two
types of animal behavior do exist, they do not amount to a
“Chinese Wall” separating them. For a number of biological
phenomena (imprinting, modification of song circuits and
circuits of other behaviors dealt with earlier in this
chapter), it has been shown that both these forms of behavior
are based on essentially similar neurobiological mechanisms.
Like evolution of innate behaviors, the learning of new
behaviors is associated with structural and functional changes
in neurons and neural circuitries, as pointed out earlier in
examples of the song circuitries of zebra finches.
The fact that similar neural mechanisms determine both
learned and innate behaviors arises the question whether
learned behaviors may evolve into innate behaviors. Darwin
believed that the learned behaviors could be inherited and
lead to evolution of innate behaviors:
If we suppose any habitual action to become inherited - and it
can be shown that this does sometimes happen - then the
resemblance between what originally was a habit and an
instinct becomes so close as not to be distinguished. (Darwin,
1859g2)
Even more
explicitly, he states:
Some intelligent actions, after being performed during several
generations, become converted into instincts and are
inherited. (Darwin, 1874)
Moreover,
Darwin believed that learning of new behaviors may even modify
innate behaviors:
Habit no doubt sometimes comes into play in modifying
instincts; but it certainly is not indispensable, as we see,
in the case of neuter insects, which leave no progeny to
inherit the effects of long-continued habit. (Darwin, 1859j1).
Presently as well, it is generally believed that innate
behaviors have evolved from learned behaviors. However,
contrary to Darwin’s view that cerebral organization is
involved in evolution of instincts, the prevailing
neoDarwinian view is that, in order for a learned behavior to
be transmitted to the offspring and become an innate behavior,
spontaneously arising gene mutations must occur. It is quite
ironical that on Darwin’s behalf, neoDarwinians reject
Darwin’s idea that variations of innate behaviors are not
accidental, that randomness and spontaneousness of the
variations in instincts is product of “our ignorance”. More
than adequate experimental evidence proves that no changes in
genes are necessary for evolution of new innate behaviors and
later in this work (see Epigenetics of Sympatric Speciation,
chapter 20) additional evidence will be presented in support
of this idea.
Being a response to changed conditions of living, often a new
behavior is associated with a stress condition. It is possible
that neuroendocrine mechanisms of stress facilitate the
transmission of the learned behavior to the offspring
as an innate behavior. Reliable experimental evidence
in support of that hypothesis is modest, but now it is almost
consensually admitted that any instinct or innate behavior has
evolved from a learned behavior at some point in time in the
past. Hence, it is worth reviewing the modest evidence on
transformation of learned behaviors into innate behaviors.
Example 1. Although evolution of behavior of
unicellulars is beyond the scope of this work, an experimental
example of transmission of a learned behavior into an innate
behavior in a unicellular, involving no change in genes,
bears some relevance to the study of the possibility of
transformation of learned behavior into an innate behavior.
For, if one believes in the origin of multicellulars from
unicellulars then it is beyond imagination, and incompatible
with the principles of organic evolution, to think that
evolution would have lost or not used such a highly adaptive
property (transmission of learned behavior to offspring) in
organisms standing higher on the evolutionary ladder.
In 1971, S.R. Bergstrom reported his experiments on learning
and transmission of learned behavior in a unicellular. Those
experiments were carried out under strictly controlled
conditions in Tetrahymena, a freshwater ciliated
Protozoa, 0.2mm long. Normally the microorganism does not
react (= reacts neutrally) to the light, but it avoids
electric impulses (current). In order to learn Tetrahymena
to avoid the light, both light flashes and electric current
impulses were simultaneously applied. After a training period
of time, the microorganism learned to associate the light with
electric impulses, so that it was able to display avoiding
behavior to the light even when the light was applied alone,
not combined, with an electric impulse. The investigator
observed that the light avoidance behavior increased with the
increase in the number of trials or training time of combined
light-electric current treatment. Over time, after the trials
were discontinued, the avoidance behavior gradually weakened.
The third and the most important observation was that after
cell division, daughter cells of the animals which learned to
avoid the light displayed the same avoiding behavior when
exposed to light only (Bergstroem, 1970).
Bergstroem’s experiments on Tetrahymena demonstrated
that, the unicellular organism transmitted to the first
generation offspring (two daughter cells), a new character,
without any change in genes.
Nevertheless, one must be wary of premature generalizations.
Indeed, there is a strong argument against such a
generalization: in a unicellular whose reproduction is based
on mitotic division there are no great barriers for the
transformation of a learned behavior into an innate behavior;
the acquired epigenetic (it cannot be a gene mutation since no
particular gene mutation would occur so systematically in
whole populations) structure that makes this transformation
possible would easily be divided between both daughter cells.
Unlike this, in a sexually reproducing multicellular animal, a
similar transformation of a learned behavior into an innate
behavior would require that the germ cell(s) somehow inherit
the epigenetic information for
that behavior rather than the epigenetic structure per se.
Fortunately, examples of
unambiguous sudden of evolution of instincts in metazoan
organisms, however scarce, are also known.
Example 2. When the
cane toad Bufo marinus was first introduced to
Australia it was so toxic to the Australian black snake,
Pseudechis porphyriacus that ingestion of even a small
toad was lethal to the snake. Now, about 60 years (~23
generations) after introduction of the toad in the continent,
the snake has evolved an innate avoidance behavior
toward the toxic cane toad
(Phillips and Shine, 2006).
Example 3. The solitary
sedentary form of the locust Schistocerca gregaria does
not practice flying. However, when forced to live in sites
crowded with conspecifics, the locust changes its behavior
from solitary to gregarious, preferring to live in crowd and
fly over distances with other locusts of its species. This
behavioral transformation is related to perception, under
conditions of crowding, by the locust of the aggregation
pheromone, which binds specific proteins on dendrites of the
olfactory receptor neurons. Perception of aggregation
pheromone in the brain induces mutual attraction of locusts,
while cuticular hydrocarbons perceived via antennal olfactory
neurons stimulate insects activity and group formation. The
behavioral change is correlated with secretion by the brain of
a factor (agoratropic factor), whose injection in
locusts also induces transition from the solitary to the
gregarious behavior. Transition to the larval gregarious
behavior occurs within 0.5 to 4 hours but the reverse
transition to the solitary behavior is slower (Applebaum and
Heifetz, 1999).
Phase transition is followed by specific changes in several
morphological, morphometrical, and physiological characters
and, what is even more surprising, the new, learned
behavior of living in group and flying is maternally
transmitted to the offspring, together with all the
morphophysiological characters acquired by the mother.
Although the acquisition of this innate behavioral may be
reversed to the original if the locust would be exposed to
respective conditions, it clearly shows that a parental
learned behavior is passed on to the offspring in an
innate behavior within a single generation, without
changes in genes.
Example 4.
The predatory snake, Natrix maura, was first introduced
to the Spanish Mediterranean island of Mallorca by Romans
about 2,000 years ago as part of an ancient Roman fertility
ritual. Ever since, the native Mallorcan midwife toad,
Alytes muletensis, inhabiting natural ponds, has evolved
an adaptive innate behavior of suppressing its movement
upon visually detecting the presence of the predator snake.
Moreover, experimentally it has been shown that the toad
displays the same adaptive behavior even when it does not see
the predator but only perceives the chemical cues the snake
releases in the water, at the same time that it does not
respond to chemical cues of other midwife toad-eating snakes
of the mainland Spain (Griffiths et al., 1998).
It may be
assumed that introduction by ancient Romans of the predator
snake in Mallorca added a threat and an environmental stress
to the population of the native A. muletensis,
which first learned that by suppressing its movements it
becomes less visible to the predator. Later it might have
learned to relate the presence of the predator snake to the
chemicals it releases in the environment (that it has
learned to recognize chemical cues of the predator is
proven by the fact that the toad does not react to chemical
cues of other nonpredator snakes). Now, both initially learned
behaviors have evolved into innate behaviors: Mallorcan
midwife toads are born with the instinct of “freezing” in the
presence of the predator .
There is no reason to believe that the above examples (as well
as some other examples from insects not presented here) of
transformation of learned behaviors into innate behaviors may
be “exception to the rule”; for from an evolutionary point of
view, i.e. from the point of view of the advantages it offers,
once evolved in unicellulars and multicellulars, the ability
to transform a learned behavior into an innate behavior would
have been conserved in multicellulars.
Behavioral Atavisms – Activation of Ancestral Behavioral
Circuitries
Essentially, evolution of behavior in metazoans is evolution
of the structure and connectivity of neural circuits. We have
shown that the same behavior may be determined by distinct
neural circuits and the same circuit may be modulated to
produce different elementary behaviors (FAPs).
All innate behaviors (avoiding behavior, mating behavior,
courting behavior, etc.) result from activation of specific
neural circuits. Under changed conditions of living, the
animal may be forced to change a particular behavior and
later, in the course of its phylogeny, the organ(s) performing
the behavior.
Let’s consider a hypothetical gradual transformation of a
terrestrial habitat into an aquatic habitat. Under such
conditions, walking will gradually become impossible and
terrestrial vertebrateswill be forced to swim. As for the
possibility of activation of a swimming motor pattern, recall
that studies in invertebrates have shown that the same CPG
(central pattern generator) might serve several different
locomotory behaviors, such as swimming, crawling and
burrowing, implying that from an evolutionary view the reverse
transition burrowing and crawling vertebrates into an aquatic
swimming-requiring habitat would not be impossible. Swimming
circuits are still functional in most terrestrial vertebrates
as suggested by the fact that most of them are able to swim
when forced to do so. The stressfully changed conditions in
the new aquatic environment may only stimulate their
modification for better swimming.
When some 380 million years ago an adventurous fish decided or
was forced to explore the land, it found locomotion in the
terrestrial environment very difficult but its fins, to some
extent, might have supported its crawling movements. In
experiments (Ayers et al., 1983), the isolated lamprey spinal
cord bathed in D-glutamic acid (an amino acid that also serves
as a neurotransmitter) generated a motor pattern that has been
assumed to represent the central motor program underlying
swimming, but their analysis shows that undulations produced
by exposure of the spinal cord to D-glutamate solution are
different from those observed during normal behavior, and the
investigators believe that this central motor program might
represent
a fundamental undulatory pattern that is modulated by
different descending systems to produce the complete
undulatory behavioral repertoire. (Ayers et al., 1983)
Thus, behaviors mediated by front-to-rear lateral undulations,
including swimming, burrowing and crawling movements of
ancestral lampreys, may be regulated by a single motor
pattern. This central motor program might have been used by
land-exploring fish to switch from swimming behavior to
crawling behavior and holding up their body. It must have been
a very painful journey but, nevertheless, extremely beneficial
and rewarding. The innate half-crawling locomotion of the
first colonizer of land might have started as a learned
behavior, based on the swimming central motor program.
Limbs of terrestrial vertebrates, as we see them presently,
have lost the hydrodynamic features of appendages of aquatic
vertebrates, such as interdigital membranes that are still
used by aquatic birds and frogs for paddling in water.
Nevertheless, there is evidence that having lost these
ancestral webbed feet, many mammals, have conserved not only
the swimming circuit and swimming behavior but also the
developmental pathways for developing webbed feet. So, for
example, embryos of all the vertebrate terrestrial classes,
including us humans, still develop webbed feet (with
interdigital membranes), which are later apoptotically
eliminated at various stages of embryogenesis.
Let’s also remember that the evolution of webbed feet, while
adapting them for swimming, would not prevent animals from
walking, as it can be seen in modern ducks and other
webbed-feet birds. Terrestrial mammals that had to adapt to
aquatic conditions of life stepwise returned to ancestral
webbed feet by simply preventing apoptosis of interdigital
membranes they develop as embryos. In all the likelihood, this
is what actually is happening in mammals such as otters, sea
lions, etc that seem to be in a transition stage of adaptation
to aquatic habitat.c
Reversal of lost ancestral behaviors is not only a theoretical
possibility. Konrad Lorenz obtained a hybrid duck by crossing
two races, Chiloe teals and Bahama pintails. To his surprise,
he observed that elements of the courtship display of the
hybrid duck resembled those of neither of two parental
species, but a primitive precursor of both parental species.
According to J. L. Gould, the ancestral behavior had not been
replaced, but simply repressed (Gould, 1982e).
Another experimental example of reversion of an ancestral
behavioral response: when female guppies of the species
Poecilia reticulata visually detect the presence of their
cichlid predator, Cichlasoma biocellatum in the
environment, they reverse to the ancestral preference for
duller instead of bright-colored males (Gong and Gibson,
1996).
In the course of their phylogeny animals have shifted to
different habitats. In the process of evolving new behaviors
and structures, animals lose previous behaviors that are no
longer adaptive to the new habitat, but they conserve the
circuits for the “lost” behaviors. Species that happen to
return to ancestral or quasi-ancestral habitats might activate
the conserved ancestral circuit and reverse to the lost
ancestral innate behavior.
This is not a purely speculative idea. There is evidence that,
while losing particular behaviors and even the structures
performing these behaviors, metazoans retain the structural
basis of the “lost” behaviors. The flightless grasshopper,
Barytettix psolus, and Schistocerca locusts are in
possession of two similar large interneurons, the descending
contralateral movement detector (DCMD) and the tritocerebral
commissure giant (TCG). These interneurons, which are
homologous in both species, enable locusts to fly while the
grasshopper cannot fly. It has been observed that differences
exist only in the connections made by a single first-order
axonal branch of the DCMD interneuron with the flight
motoneurons: while in locusts the DCMD sends branches to the
dorsolateral neuropile and forms synapses with flight motor
neurons, such branches form only in 52% of cases in the
grasshopper and all of the connections have abnormal
projections in comparison with locusts:
The differences in DCMD projection suggest that a discrete set
of output connections may have been modified in Barytettix
by the alteration of a single first-order axonal branch. (Arbas,
1983)
The grasshopper, Barytettix psolus, lost the ability to
fly as a result of the loss of hind wings, reduction of
immobile forewings to a vestigial state, and the loss of the
indirect flight muscle, the metathoracic dorsal longitudinal
muscle, which develops in the nymphal stage but is lost in
adult grasshoppers. However, the motor neurons for flight
muscles and wings are retained although their target muscles
are lost (Arbas, 1983; Arbas and Tolbert, 1986). The
conservation of motor neurons that are directly related to the
lost structures (muscles and wings) and functions (flight)
suggests that, despite the changed connections, the flight
circuitry in nonflying grasshoppers is preserved.
We have pointed out that the structure of the nervous system
and even neural circuits are conserved to a considerable
extent in the course of the evolution of metazoans. Here is an
impressive illustration. Innate fear of snakes is common not
only among humans but among other primates as well. Eleven
species of primates exhibit fear-related responses (avoidance,
alarm calls, mobbing, etc.) in virtually all instances in
which they were observed confronting large snakes (Oehman and
Mineka, 2003). Limbic structures, related to the snake fear
neural circuits, emerged during the evolutionary transition
from reptiles to mammals (first mammals were small
insectivorous tetrapods) but before the evolution of neocortex,
as can be concluded by the irrationality of the snake fear
observed under natural and experimental conditions and the
immodifiability of the snake aversiveness. Although the snake
fear circuit evolved long before the evolution of man, during
transition from reptilians to mammals, the circuit still
exists in most mammals but has been later modified in mammals
that preyed on reptiles, including snakes (Oehman and Mineka,
2003). For more than 200 million years the “snake fear”
circuit is still functional in the rest of mammals!
The fact that the structure of the nervous system in metazoans
and even specific neural circuits are conserved to a great
extent represents a crucial premise of the evolutionary
ability of metazoans to revert to ancestral behaviors and may
be an important asset of the metazoan evolutionary-adaptive
strategy.
If we should believe, as most biologists do, that the swim
bladder in teleost fish evolved from the lung of ancestral
lung fish, the breathing CPG (central pattern generator) must
have been modified when they substituted the swim bladder for
the lung. The re-evolution of lungs in tetrapods required
re-invention of the lost circuitry for breathing for replacing
the “buoying” motor pattern used for controlling the water
depth-graded swimming by inflating and/or deflating the swim
bladder in teleost fish. In the Urtetrapod, this might have
required just reactivation of an ancestral silenced
“breathing” circuitry for terrestrial life. Recall that even
the same circuit in metazoans (e.g. Aplysia circuitry)
may be modified to generate more than one motor pattern.
Developmental and Evolutionary Relationship between
Behavior and Morphology
Each phenotypic structure is related to one or a number of
innate behaviors: a fin to swimming, a wing to flying, a
digestive tract to eating, a lung to breathing, secondary
sexual traits - to the sexual behavior of males and females,
and so on. For, ultimately, the structure evolves not for its
own sake but essentially for the sake of the function, of the
behavior it has to perform. Behavior is the ultimate cause of
the structure. In the first part of this work I have presented
empirical evidence that the CNS and neural circuits are
essentially involved in the development of morphological,
physiological, and behavioral characters. The question now
arises whether neural circuits for behavioral and morpho-physiological
characters may be developmentally and evolutionarily related.
Adaptation of metazoans to sudden environmental changes and
the ensuing environmental stress begins with adaptive changes
in behavior, which precede all other forms of the phenotypic
(morphological, physiological and life history) adaptation.
From an evolutionary point of view, it is plausible that a
pressure for evolving a relationship between the changes in
behavior, on the one hand, and the following adaptive changes
in morphology and physiology, on the other, would have always
been present. In view of the fact that, ultimately, the
information for both performing behaviors and for developing
animal morphology comes from the nervous system, it is to be
expected that in the course of evolution, causal relationships
might have been established between behavioral circuits and
circuits involved in the development of morphological traits.
Is there evidence on a causal or noncausal relationship
between the evolution of behavior and the evolution of
metazoan morphology?
If evolutionary change is transmission to the offspring of a
character that the parents have not inherited but have
acquired during their life time or appears for the first time
in the offspring, modern biology offers a considerable number
of demonstrated and demonstrable cases of evolutionary change,
which now are object of an interesting field of the study, the
transgenerational developmental plasticity. Based on the fact
that such cases meet the basic criterion of the evolutionary
change, that is transmission of the new character to the
offspring, identification of the mechanisms of their emergence
may provide important clues to understanding mechanisms of
metazoan evolution. While human life is too short to witness
possible evolutionary relationships between behavior and
morphology in nature, the study of transgenerational
developmental plasticity, and developmental plasticity in
general, might reveal that relationship, if it exists at all.
Lex parsimoniae tells us that there is no reason to
suspect that the mechanisms of transgenerational developmental
plasticity might be different from mechanisms of long-term
evolutionary changes. Evolution is too economical to waste
resources for the luxury of evolving two different mechanisms
for a single end, that is generation of inherited changes.
It is generally believed that evolution of behavior precedes
evolution of the structure for performing the behavior. In an
overused aphorism, “Behavior evolves first”.
From an evolutionary point of view, a causal relationship
between evolution of behavior and morphology that might arise
from performing the new behavior would clearly be
advantageous. And if such a relationship indeed exists, it
probably would be expressed at the level of responsible for
behaviors and morphology.
Evidence on a close relationship between evolution of behavior
and animal morphology and physiology has been presented
earlier by a number of authors. In experiments on functional
mechanisms of predator-induced changes in morphology and
behavior of Hyla versicolor tadpoles, van Buskirk and
McCollum have observed that changes in behavior, on the one
hand, and the color and relative length and depth of tadpole
body and tail, on the other, vary as an integrated unit and
conclude that behaviour, colour and morphology are highly
correlated in naturally occurring tadpoles (van Buskirk and
McCollum, 2000). Fuchs et al. also have described the
existence of a relationship between the behavior and
morphological and physiological changes and have pointed out
the role of behavior in inducing physiological changes in the
case of phase transition in locusts:
Locusts are capable of extreme behavioral plasticity; in
response to changes in population density, they dramatically
alter their behavior. These changes in behavior facilitate the
appearance of various morphological and physiological changes,
cumulatively termed density-dependent phase characteristics…
the behavioral changes are, on the one hand, a response to
specific environmental changes, and on the other,
stimulant-catalysts of various other environmentally induced
physiological changes. (Fuchs et al., 2003)
Theoretically, it might be argued that the experimentally
confirmed correlation between changes in behavior and
morphology is inherently determined by the fact that
morphologies in general are means for performing specific
functions and behaviors such as feeding, preying, hiding,
aquatic, air or terrestrial locomotion (swimming, flying, or
walking), etc. The correlation could have been established in
the course of phylogeny and is based on the fundamental fact
that both the behavior and morphological-physiological
characters develop under control of the CNS (see chapter 1,
Control Systems in Metazoans ).
As shown, I cannot claim to be the first to have presented
evidence on the existence of a close relationship between
behavior and animal morphology and physiology. What I claim
here, instead, is that adequate empirical evidence exists for
validating my hypothesis on the existence of a causal
relationship in the evolution of metazoan behavior and
morphology.
It was pointed out
earlier (and will be discussed in some details in chapter
chapters 11 and 12) that, in response to specific stimuli,
locusts of the species Schistocerca gregaria
(Forskål)
switch between two behaviorally and morphologically distinct
forms in a phenomenon known as phase transition. The solitary
form, which lives isolated, away from other conspecifics, when
put under crowding conditions or under influence of pheromones
or tactile stimuli, switches to the gregarious form, which
displays not only several new behavioral traits [tendency to
swarm and fly with locust crowds, to feed on a
toxic alkaloid-containing plant that avoided before (Despland
and Simpson, 2005),
etc.] but also exhibits several changes in
morphology, morphometry and body coloration (characteristic
color change from cryptic green to warning brown coloration).
Behavioral changes may appear within one to several hours and
are reversible.They precede the morphological, morphometric
and color changes during phase transition.
All the phase
change-inducing factors act via the insect central nervous
system [crowding in this insect is a stressor that also acts
via the CNS as is concluded by the experimental evidence that
antennectomized locusts do not change phase under conditions
of crowding (Applebaum and Heifetz, 1999)]. Moreover,
the stressed locusts transmit the acquired traits to the
offspring. The full scale phase transformation takes several
generations and occurs probably only in nature (Pener et al.,
1997).
At a neuroendocrine
level this transformation is related to an elevated level of
JH (juvenile hormone) under stimulation of neurohormones
allatotropins and nerves innervating the corpora allata as
well as cerebral secretion of [His7]-corazonin (Grach
et al., 2003), also known as DCIN (dark-color-inducing
neurohormone).
Intense changes are also observed in the levels of numerous
neurotransmitters in the locust brain (Rogers et al., 2004).
All the changes
during transition to the gregarious phase are triggered by
sensory stimuli (visual, olfactory, and tactile), which are
perceived in the insect brain where the information for
activating the signal cascade for changes in behavior and
morphology is generated by processing the afferent neural
input from sensory neurons. Aggregation pheromones received by
the olfactory neurons are converted into electrical spike
trains in which form they are transmitted for processing first
to the frontal antennal lobe then to the mushroom body and
further to the lateral protocerebrum (Anton and Hanson, 1996).
Tactile stimuli (touch on the outer side of
the upper portion of a hind leg, for instance) from
mechanosensory trichoid sensillae on
the hind limb, via metathoracic
nerve 5, are also transmitted to the CNS (Rogers et al.,
2003).
Transmission of the
gregariousness to the offspring is correlated with deposition
of a 5-10-fold greater amounts of ecdysteroids in the eggs of
the gregarious locusts than in the eggs of the solitarious
locusts (Tawfik et al., 1999; Tawfik and
Sehnal, 2003; Hagele et al., 2004; Tawfik et al., 1999).
Observations on locust phase
transition show that a single stimulus, visual-social
(crowding), olfactory (aggregation pheromone), or tactile (touch
on the outer side of the upper portion of a hind leg) is both
necessary and sufficient for stimulating impressive behavioral
and morphological changes of phase transition within a few to
24 hours.
The fact that the circuitry for
gregarious behavior and circuitries for gregarious morphology
in locusts are activated by the same stimulus, and that always
changes in behavior are followed by changes in morphology and
body coloration, suggest that at some level of the brain
function or structure, behavioral circuits are related to
circuitries that, via signal cascades, determine the
development of gregarious morphologies. While the fact that
behavioral change precedes the appearance of morphological
changes suggests that the induction of the circuit for changed
behavior may somehow influence the circuit(s) determining
changes in the color and morphology, the possibility of an
independent, parallel activation of the latter by the same
stimuli cannot be excluded.
The temporal correlation between
changes in behavior and changes in morphology observed during
phase transition in locusts is not unique. Examples of such
correlations abound in the field of developmental plasticity.
In most cases of predator-induced developmental plasticity,
changes in morphology are also preceded by, or accompanied
with, changes in behavior. For example, larvae of the pipevine
swallowtail butterfly, Battus philenor, show a
phenotypic plasticity in the South West of the United States:
in California they are predominantly black, while in western
Texas and Arizona - predominantly red. Recently, investigators
have observed that California butterfly larvae, in an adaptive
response to the higher summer temperature, exhibit a double
(behavioral and morphological) phenotypic plasticity. In order
to avoid the excessive summer heat, they switch to a new
climbing behavior by climbing higher on non-host plants and
change their body color from black to red (figure 9.8).
These changes are adaptive, for both color change and climbing
allow the larvae to escape the higher temperatures. The
critical temperature for the onset of the polyphenism lies
between 300C and 360C and the
polyphenism is reversible. Both red color and climbing
behavior are components of a thermoregulatory strategy
intended
to avoid internal temperatures
above the thermal maximum temperature for growth and
development in B. philenor or to maintain body
temperatures in the optimum range for facilitating maximum
growth rate… The maintenance of maximum growth rate may be
critical for insect larvae susceptible to larval predators or
parasites. (Nice and Fordyce, 2006)
The systematic correlation of the
change in climbing behavior with the change in body color
suggests the existence of a causal relationship between them.
Figure 9.8. The black
(left) and red larval phenotypes of B. philenor
observed in two half-siblings from Texas. The black larva was
reared at 300C and the red larva was reared at 360C
(From Nice and Fordyce, 2006).
The
neotropical tadpole,
Rana palmipes,
in response to the presence of its predator water bug, or even
of predator cues alone, changes its behavior by strongly
reducing its activity, darkening its body color and increasing
the size of muscle and tail (McIntyre et al., 2004).
In response
to the presence of its predators, the freshwater snail,
Helisoma trivolis, simultaneously changes its behavior
(preference for a particular habitat and the timing of the
onset of the reproductive behavior) and morphology (the form
of the shell) (Hoverman et al., 2005).
Acyrthosiphon
pisum
(Harris, 1776) is a pea aphid that in the presence of
predators emits a volatile alarm pheromone, which, when
perceived in the brain of females, induces the latter not only
to shift to walking behavior and drop off the plants but also
to increase the proportion of winged morphs in the offspring
(Dixon and Agarwala, 1999; Kunert and Weiser, 2003).
North American frogs of the genus
Scaphiopus are omnivorous amphibians that, as tadpoles,
inhabit ephemeral ponds and flooded areas, which only
exist for short periods of time, often before the tadpoles
could develop into adult terrestrial individuals. These
species exhibit an adaptive strategy, a developmental
plasticity that enables a proportion of tadpoles to develop an
alternative carnivorous behavior and mouth morphology.
According to D. Pfennig (1990), the proportion of tadpoles
that develop carnivorous behavior and mouth morphology depends
on the amount of shrimp they eat and shrimps are more abundant
in ephemeral ponds (Pfennig, 1990). The carnivorous tadpole
morphology is similar to mouth adaptations of
Hoplobatrachus tadpoles. Tadpoles of both groups have
longer intestines than those of other carnivorous species (Grosjean
et al., 2004). All these facts support the hypothesis that
carnivorous behavior and mouth morphology in anuran tadpoles
evolved in correlation, as an adaptation to the temporal
unpredictability of desiccation of the pond. Tadpoles of
desert amphibians live in temporary ponds that contain water
for unpredictable periods of time. In the years of low
precipitations the pond dries up earlier than usually. This
causes a habitat stress to which the tadpoles of that and some
other species respond by changing their behavior and speeding
up their metamorphosis to transform into adult amphibians,
able to live on dry land.
I have already mentioned the
example of the Mallorcan midwife toad, Alytes muletensis,
which in response to the presence of its viperine predator,
and even upon detecting a chemical released by the predator,
induces rapid changes in its behavior and later changes in
morphology, which make the toad less vulnerable to the snake.
Konrad Lorenz has shown that
birds which make nodding movements while courting eventually
develop highly colored feathers or crests, which draw
attention to these movements - not the reverse (Taylor,
1983d).
It is believed that a tendency to
reptate (wriggle), instead of walking on their reduced limbs,
which is observed in some lizards, is an indication that this
“wriggling” behavior is causally related to the reduction of
their limbs and may predict their future evolutionary loss
(Taylor, 1983e).
The systematic correlation of
specific behavioral changes with specific changes in
morphology in all the above cases strongly suggests the
existence of a causal relationship between the evolution of
the behavior and the structure(s) performing it.
What seems to have in common all
the examples of correlated change in behavior and morphology
is the fact that these changes are stimulated by
drastic changes in the environment, which trigger a stress
response, a response that, as has been shown, is neurally
determined. The immediate change in behavior often is itself
an intergral part of the stress response. As it will be later
shown, the stress response leads to developmental instability
that is an important permissive factor for ensuing morpho-physiological
changes.
|
