*
* Lecture notes by Edward Loper
*
* Course: 9.20 (Animal Behavior)
* Professor: Petrullis
* Institution: Massachusetts Institute of Technology
*

[09/13/99 09:09 AM]

Ecology: "Behavioral Ecology and Social Biology" online at
# http://link.springer-nm.com/link/service/journals/00265/tocs.htm

> Evolution
Defintion (in our context): change in characteristics of biological
entities over generations.  Exists among populations of entities.
requirements:
	1. populations made of variable individuals
	2. variation is heritable
	3. variation is random (esp. not directed by environment)
	4. change in proportion of individuals with a characteristic
	   in a population occurs.  Dictated by chance (genetic
	   drift) and by natural selection.  Genetic drift more
           signifigant for smaller populations.  Natural selection is
           differential reproduction & survival, and accounts for
           adaptation.
	5. populations may subdivide (speciation)

> History: Pre-Darwin
      * Greeks: projection of archetype on world gives individuals.
	Basically, species don't change.
      * Early Christian: species don't change.  Also, there is a chain
	of perfection: God \gets Angels \gets humans \gets [animals].  The
        chain is unchanging.  Thus, no extinctions and no new species.
      * Linnaeus: created taxonomy based on similarity of form.  Not
	based on phylogeny.
      * De Buffon: each species has an internal model of what it
	should look like.  Closely related species may have had a
	common ancestor.  Postulated that animals could be
	differentially modified by their environment.
      * Lamark: Species naturally improve.  Speciation occurs from the
	simplest animal:

#	        /A4
#	       /
#	      / /A3
#	     / /      } species chain we see today
#	    / / /A2
#	   / / /
#	  / / / /A1
#	 / / / /
#	S S S S S \gets simplest species (spontaneous generation)

	Believed that animals adapted differentially to the
	environment.  Giraffe example.

> History: Darwin (early 19th century)
      * Travelled with the hms beagle on south american coast.  In the
	gallopogos islands, found closely related species varying from
	island to island.
      * Read pamplet by Malthus (who said that eventually there would
	be too many people to be fed -- will lead to famine).  Noticed
	that populations don't live up to reproductive limit.  So if
	some individuals are more likely to reproduce, and traits are
	heritable, you can get natural selection..
      * Spent 20 years thinking about implications etc.  Recieved
	manuscript from Wallace.  Presented both his and Wallace's
	papers, and finished his book.
      * Theses of the book:
	1. all species have arisen continuously from few 1st species
	2. similar species have diverged from a recent common ancestor
	   by accumulation of small changes
	3. different groups diverged from a more ancient ancestor.
	4. causal agent of evolution is differential reproduction.
      * Break-down of Theories presented by Darwin:
	1. Evolution (change over time) exists
	2. Common descent: even with humans
	3. Gradualism: multiple small steps
	4. Populational speciation: evolution occurs in populations
	5. Natural selection: dictates how evolution occurs

> History: Post-Darwin
      * Agreement that evolution had occured, but dispute on
	mechanisms.
      * Neo-Lamarkism: inheritance of acquired characteristics.
	Researcher showed that, even if you cut tails off mice, they
	won't ever give children with short tails.
      * Mutationism: mendellians claimed that mutation accounts for
	speciation.  Believed that any mutation that can survive will 
	persist.  Most mutations are bad, but occasionally one will
	create a new species.

> History: Modern Synthesis
      * Reconcilliation between genetics & darwinism & palentology
	etc.
      * Tennants:
	1. Phoenotype (observed characteristic) differs from the
	   genotype.  Individual differences are partially 
	2. Environmental effects on phoenotype are not transmitted to
	   genotype
	3. Phoenotype is based on genes
	4. Genes mutate to different alleles randomly
	5. Evolutionary change is a populational process.
	6. Mutation is too slow to shift a population from one type to
	   another: need some other form of selection.
	7. Natural populations are variable within a species.
	   Differences are based on genetic effects.
	8. Natural selection occurs in nature (in the present).
	   Differences in populations of species is adaptive.  (e.g.,
	   house sparrows are an English species, which varies very
	   little in traits 1..n.  But it was released into US, and
	   here there is lots of variation.  100 generations later,
	   found correlation between body mass & climate, etc.
	   Another example: in a year of drought, galopogos pigeon
	   beaks grew by 5%, since only big nuts survived drought, and
	   need big beaks to crack nuts.  Another: Annolus introduced
	   experimentally to several islands that didn't have any
	   annolus's.  Previously, annolus's that hung to big branches
	   would have big hindlimbs, small branches would have small
	   hindlimbs.  Found that the average limb side went down for
	   islands with small branches etc.
	9. There is a continuum between the small changes and
	   speciation.  Thus, genetically isolated populations can
	   diverge enough from original populations to form a new
	   species.  (Species are defined as seperate (non-breeding)
	   populations)
	10.Fossil record has gaps: there are specific conditions that
	   need to occur for fossils to be created, so there simply
	   aren't any fossils from some times & places.  Nevertheless,
	   there are still continuous changes etc.

[09/15/99 09:05 AM]

> Problems with the Modern Synthesis
     1. Traditionally, M.S. has a conception of an organism as a
	mosaic of traits, where each trait is independantly
	optimized.  But animals are built up from what ancestors have
	gone through.  The phylogeny constrains the evolutionary
	paths.  Also, traits must be integrated within an organism.
	Thus, we shouldn't expect all traits to be "useful," since it
	may take lots of energy to restructure things to get rid of
	them.  Also, traits can be byproducts of the way other things
	develop, or constraints of the way things happen.  E.g., skull
	sutures are not adaptive, but are just a byproduct of the way
	the skull develops.  Thus, it's important to consider a trait
	in the context of an organism (esp. it's phylogeny and
	ontology).  

     2. M.S. assumes that there is a short connection between genes
	and traits.  But genes just code for protiens -- many steps
	between genes and traits.  These steps interact with genome
	and environment.  Also, M.S. ignores the ability of animals to
	change their environment.  Thus, heritable doesn't really mean
	genes -- the parents give much more to the child than just the
	genes (e.g., modified environment, etc).  As a special case,
	consider culture.

> Natural Selection
>> Prerequesites for Natural Selection
	1. Variation
	2. Heritable variation
	3. Some traits are better adapted to environment than others
	   (Differential adaptation)
	4. Individuals with these traits will consistantly leave more
	   offspring in next generation (Differential reproduction)

Define fitness as the average rate of increase of a trait over
generations.

>> "Selection of" and "Selection for"
Consider a toy which filters balls through progressively smaller
holes.  But big balls tend to be a different color than small ones.
This toy selects _for_ small balls, but there is still a selection of
small-colored-balls.  E.g., there is selection for opposable thumbs,
with the incidental effect that there is selection of piano players.
Note that we select _for_ traits, not "for avoidance of extinction".

>> Levels of Selection
Group selection vs. G. C. Williams.  Altruistic genotype will leave
less offspring than non-altruists.  Also, groups succeptible to
cheaters.  Thus, altruistic trait can't evolve by Darwinian
(natural) selection.  Then we resort to group selection (i.e.,
selection between groups).  Groups with altruistic traits will tend to
survive more than selfish groups.

But there are many problems:
      - very hard to find populations which are so isolated that
	cheaters won't "contaminate" them.
      - alternate explanations possible for seemingly altruistic
	behaviors.
      - many many more individuals than populations.  Individuals are
	born & die much much faster than populations do.  Difference
	between individuals in a population tends to be much higher
	than differences between populations.  Thus, natural selection
	is much much faster than group selection.  In general, natural
	selection will almost always win over group selection.


> Adaptations
>> Definitions
Adaptation is both a feature of an organism, and a process...
     1. An adaptation is a phoenotypic variant that results in the
	highest fitness than the alternative variants.
     2. An adaptation is derived from ancestoral populations.  Define
	adapation in terms of its history
     3. Syntheis: An adaptation is a trait that has become prevalant
	or remains prevalant because of natural selection..

>> How do we identify an adaptation?
We can identify adaptations (heuristically) by:
      - complexity
      - seems to be "designed" to do something that increases
	fitness.
      - experiments: artificially modify the trait in a population,
	and see if it affects fitness etc.  (e.g., moths on polluted
	trees)
      - comparitive method: animals that have come to the same
	solution, but by different unrelated species, then we expect
	that to show a common selection pressure.

>> Caveats (What not to expect)
     1. An adaptation doesn't have to be necessary for survival.
     2. Don't expect perfection
     3. Don't expect "progress"
     4. Don't expect harmony or balance, just stability
     5. Evolution is ammoral

> The Study of Animal Behavior
>> Observation and Description

You must begin by observing and describing an animal's natural
behavior.  This both generates questions, and allows us to tell more
about what the animal is doing.  Have to deal with abstraction and
bias.  When and where to observe?  Behaviors change over time and
place.  Also, what behaviors to observe?  Distinguish function and
mechanism.  When we ask "why" an animal sings, we can talk about
evolutionary reasons, or immediate causes (mechanisms).  Also,
examine in terms of ontogeny (development) or phylogeny: birds sing
because their ancestors did; birds sing because they developed in
certain ways that lead to that.

>> Tinbergen's Questions
The following four questions are all valid, and none can be reduced to
the others.  We can answer any of them about adaptations or behavior
etc.
	1. causation: proximal mechanism
	2. ontogeny: how it developed
	3. phylogeny: how it evolved
	4. survival value: how it affects fitness

Book groups 1-3 as proximate, and 4 as ultimate (distal?).

>> The Importance of Keeping Questions Seperate
These 4 questions can't be reduced to each other.  Be sure to
differentiate them, and not confuse them.

[09/20/99 09:00 AM]

> Animal Behavior: Two Traditions

There have traditionally been two separate traditions of study for
animal behavior: ethology and animal (comparative) psychology.

Ethologists were trained as zoologists, and studied a wide variety of
animals.  Their emphasis was on instinct and evolved behavior.
Methods were careful explanation and field explanation.  Ethologists
were mainly from Europe.

Animal psychologists were trained as psychologists, studied mainly
mammals, and especially lab rats.  Their emphasis was on learning and
the development of theories of behavior.  Methods were lab work,
control of variables, and statistical analysis.

>> The Research Emphasis of Ethology

Developed in 30's \to 50's.  The study of evolution of behavior &
instinct.  Instinct = adaptive behaviors used to solve problems in the
natural world.  Instinctual behaviors were thought of as being
adapted, and not learned.  Took each species at face value (not a
model of a human behavior, & no "scala natura").  Emphasis on complex
behavior.  Documented behavior very well, but tended to avoid
experimentation.  

>> The Research Emphasis of Animal Psychology

Search for a general set of rules that could be applied to all animals
to describe their behavior.  Emphasis on the modifiability of behavior
(learning).  Early emphasis on studying many different animals, but
this went away.  Animal psychologists ultimately wanted to understand
human behavior, & incorperates the "scala natura".  Domesticated
mouse, and inbred them to reduce variability between individuals.
Wanted to eliminate subjective reporting, & advocated control.
Simplified behavior & examined one variable at a time.

> Conceptual Stances of Ethology and Animal Psychology

>> Ethology
Interested in mechanism and by evolution of behavior.  

>>> Fixed Action Patterns
A stereotyped sequence of actions that happen in a rigid, predictable,
highly structured way.  Complex patterns, not just simple reflexes.
Characteristic of a single species, and true of all animals of that
species (under the correct circumstances).  Behavior elicited by
specific and simple stimuli.  (self-exhausting: eventually you'll stop
doing it) Behavior is triggered, and once begun will occur
independantly of the environment.  Behavior is independant of
experience.

e.g., egg rolling: if a goose spots an egg out of its nest, it
orients, goes through a series of motions to roll the egg back to the
nest.  This movement is very stereotyped -- will continue without
feedback (even if egg is removed).  (however, the stabalizing aspect
of the behavior wasn't as stereotyped).

>>> Sign Stimulus
Simple but specific stimulus that elicits a specific response.
Also known as a "releaser".  e.g., the egg in goose example.
A "super-normal" stimulus (e.g., a really big egg) may elicit a
stronger response.  Ethologists studied what aspects of a sign
stimulus were the crucial ones.  

>>> Innate Releasing Mechanism
"Innate releasing mechanism" is responsible for detecting the sign
stimulus and releasing the fixed action pattern (no one's found one
yet).

>>> The Umwelt
"The world around me": each animal has its own world view, and
different animals are sensitive to different things in their
environment.  e.g., humans are blind to many stimuli that other
animals use.  Bees can detect uv and some fish, electric signals.

>>> Function or Survival Value of Behavior

What's the survival value of a behavior?  E.g., egg-shell removal
experiment.  White on eggs increases predation.  Eggs with shells near
them are more likely to be eaten.  Thus, function is to decrease
predation.

>> Animal Psychology: Themes and Personalities

>>> Process of Learning (Thorndike's Law of Effect)
Thorndike put cats in puzzle boxes, where the cat must do something to
open the door.  Looked at how long it took animals to figure out to
escape.  Over time, animals got better at escaping.  Thorndike's law:
if an action A is followed by a reward R, then A will be more likely
to occur in the future (operant conditioning).

>>> Insight/Animal Intelligence (Kohler)
Kohler thoguht animals have insights that allow them to solve
problems.  E.g., chimpanzees stacking boxes to get banannas.  More of
a cognitive approach, and less of a stimulus-response one.

>>> Content of Learning and Animal Thought
What types of things can animals learn?

Lashley tested whether mice could distinguish different visual cues.
Used doors with different pictures, and even relatively similar
pictures (dots of 2 sizes) could be distinguished.  Also, they could
learn to pick the odd-door-out of 3 doors.

Tolman said we should be asking how a behavior allows an animal to
accomplish its goals.  Animals learn things other than
stimulus-response connections.  Latent learning: animals learn (e.g.,
spacial information) even when there's no reward or punishment: If you
train a rat in a maze with a cheese reward, they'll get better and
better.  However, if you just place a rat in a maze without cheese,
and then place it in a maze with a cheese reward, they'll get better
much faster.  Tolmen claimed animals have complex representations of
their environment (e.g., a "map" of the spacial layout).  In contrast
to learning things like "left right right left left".  Tolman tested
this with a plus-maze (tubes to N, S, E, W).  Blocked off N, placed
rat in S, and put food in E.  Later blocked off S, placed rat in N,
and food in E.  According to map theory, rat will go to the right
place.  According to the stimulus-response theory (rat just learned
"turn right"), it will go to the wrong place (W).  (But if you give
the rats no visual cues outside the maze, they will go to the wrong
place (W).)

>>> Behaviorism and its demise
Behaviorism ran opposite to ethology.  Claimed:
	* only observable behaviors should be theorized about
	* measure only external variables (not internal states)
	* all behaviors are learned S \to R associations
	* operant conditioning
	* equipotentiality: all behaviors follow the same principles
	  of learning (regardless of animal & task)

Behaviorism died out (replaced by cog. psych.).  Equipotentiality is
false (e.g., much easier to associate tone w/ shock or food with
indigestion).

> Comparative Psychology and Ethology

>> Interactions and Conflict
Cog. Psychologists interested in some of the same things as
ethologists.  Psychologists didn't like the fact that ethologists
didn't do experiments.  Ethologists didn't like the lack of variety in
cog. psych., nor their lack of interrest in adaptation and evolution.

>> Common Ground
Eventually realized that much of their conflict came from confusing
what questions they were trying to answer.  Cog. psych. studied mainly
mechanism, and ethology mainly evolution.

> The Splintering of Animal Behavior: Tinbergen's 4 Questions

>> Survival Value -- The Rise of Behavioral Ecology
Understanding the relationship between an animal's ecology and its
behavior.  
	1. Economic Models of Behavior: the design of a particular
	behavior is designed to be an optimally efficient solution.
	Must consider the trade-offs between different costs & rewards
	of all the different behaiors
	2. Evolutionarily Stable Strategies: the behavior of one
	animal depends heavily on the behaviors of the animals in its
	environment. 
	3. Evolution of Social Behavior (Sociobiology)

>> Immediate Causation -- Neuroethology and Physiological Psychology
Neuroethology -- studying how neural configurations cause behavior
physiological psychology -- studies of hormones, etc.

>> Ontogeny -- Developmental Psychobiology

>> Evolution (or Phylogeny) -- ignored till recently

> The Beginning of a Re-Integration

[09/22/99 09:08 AM]

> The Comparative Approach

Why do we find the differences between species that we do?  (e.g., why
is species A solitary and species B gregarious).  Compare groups of
related species, and figure out how differences in behavior reflect
differences in Ecology.

>> Useful in the search for Adaptations

Differences between related species suggest adaptations.

>> Early Studies (John Crook, 1964)

Studied about 90 species of weaver bird.  Varying social
organizations.  Some build individual nests, others build huge group
nests.  Some species monogomous, some polygomous.  Looked for
correlations between ecology and social organizations: types of food,
abundance of food, predatory pressure, nest-site location.  Showed
that weaver birds could be classified in 2 basic categories:

Forest-dwellers eat insects, live in solitude, defend large
territories, build hidden cryptic solitary nests, and are monogomous,
and are monomorphic (both sexes look the same).

Savanah-dwellers eat seeds, feed in flocks, nest colonially in the
open (not hidden), and are polygomous.  Males tended to be more
brightly colored and ornamented.

Hypothesized that predation and food drive the social organization.
In forest, insects are dispersed food resources.  Need large territory
to get enough food.  Need two parents to provide enough food for
children (\to monogomy).  Both sexes are camoflogued, as is nest.  In
savanah, seeds are patchy and super-abundant locally.  Finding patches
best done in a group.  Enough food in the patch to go around.  In the
open country, you can't hide nest very easily.  Safe nesting sites are
a scarce resource -- colonial sites.  Within a nest site (tree), there
are some better nest sites than others.  Competition for the best nest
sites leads to polygony: females tend to mate with males in better
nesting sites.

>> Problems with this approach:

>>> Alternative Hypothesis
It's a little too easy to explain the data without testing -- there
are many different possible explanations, and even though one sounds
reasonable, it may not be correct.

>>> Cause and Effect
Can't tell which is cause, and which is effect.  For example, diet of
seeds could be a cause instead of an effect.

>>> Confounding Variables
If A correlates with B, it may just be because they both correlate
with C.  E.g., body size can be a confounding variable.

>>> Different Solutions
Differences you see between species may not reflect differences in
ecological pressure -- may just represent two different solutions to
the same problem.  E.g., ram horns derived from skin, and deer horns
from bone; but this is probably not caused by different ecological
pressures.

>> The Big Problem: "We forgot about Phylogeny"
The evolutionary history of an animal needs to be taken into acount.
This has been called "the problem of independance of data points."

>>> Hoglund, 1989
Testing the hypothesis that differences were caused by Lek-ing: large
area, but during mating all males congregate in a small location, and
females come to observe females.  Found that in Lek-ing species, males
were more likely to be larger and to have plumage than in non-lek-ing
species.  However, when he examined phylogeny, he found a very high
correlation between lek-ing and dimorphism in one particular group of
animals -- biased the whole analysis.  Basically, he showed that the
correlation between lek-ing and dimorphism might be explained by
common ancestors.

>>> How do we fix it?
Conservative approach -- use higher taxanomic groups.  That way, we
can hopefully rule out the common-ancestor explanation.  But this
disregards lots of useful data.

Independant contrasts; e.g., DeVoogd et al, 1992: First, construct a
phylogenetic tree (don't use the characteristics you're looking at).
Take contrasts between two different nodes (D&E or F&G or DE&FG).
# D  E  F  G
#  \/    \/
#   \    /
#    \  /
#     \/
# (This is a Cladogram)
In particular, compare how differences in the characteristic you care
about relate to the "distances" between different species.  

For example, for songbirds, compare the differences in HVC volume and
in song repetoire sides (after compensating for brain size etc)
between different nodes.  Then look for correlation.  For example,
with Leking and dimorphism, measure both between each pair of nodes,
and then plot the correlation.

> Phylogeny and the Evolution of Behavior

Phylogeny:
- Allows you to generate hypotheses for future studies.
- Allows you to test hypotheses with variables you can't really
    manipulate (polygomy)
- Find out how a trait evolved

>> Behavior does not fossilize

It's very difficult to find any information about the behavior of
previous species.  

>> Cladistics
>>> Reconstruct Phylogeny
But you can try to use behavior to reconstruct a phylogeny.

Trace evolution of behavior from phylogeny (assumes that most change
occurs from speciation, not from slow change over time).

Create cladogram (using molecular techniques, etc).  Then create
monophyletic groups (all species that share a common ancestor).  Do
NOT group animals just based on their features: paraphyletic groups.
In the cladogram above, the groupings [D] [EFG] are paraphyletic.
Also, avoid polyphyletic groups: [D][EF][G] in the cladogram
above. (eg., grouping birds and bats)

Define an ingroup.  Then define the sister group and the outgroups
(sister group is just the closest outgroup).  

>>> Some Concepts in Cladistics

>> The Problem of Homology
Homology: a character of 2 or more taxa which is found in the common
ancestor; a trait shared by common descent.  Note that this can be
tricky: bird wings and bat wings are homologous at the level of both
deriving from forelimbs, but are not homologous as wings.

Homoplasy: similarity in form without common ancestor.  Arises from
convergent evolution, or because of parallelism.

Always begin by assuming that traits are homologous.  If not, then
we'll find out after our analysis.  

>> Reconstruct History of Trait
>>> Parsimony
If you're confronted by 2 competing theories, choose the simpler one.
In particular, try to figure out how to get your traits with the
fewest changes.  Consider, in the cladogram above, A=X, B=X, C=-, D=X.
Comparing C and D, look at outgroups, and come up with the hypothesis
that the character was gained at [ABCD] or below, and was lost at C:
this gives 2 transitions.  Lets us infer that [CD]=C and [ABCD]=X.
Note that if a trait is common to every animal in the group, this
doesn't tell us anything.

>>> Outgroups
Note that we cannot ignore any outgroups -- doing so might give
misleading results.

>>> Example: Agression in Sticklebacks

Tinburgen hypothesized: the more agressive a species, the more likely
to find ritualization: allows species to form hierarchies using ritual
fighting instead of real fighting.  Phylogenetic tests supported this
hypothesis.  

>>> Example
Hypothesis: sexual selection can cause speciation.  In many groups
that speciate a lot, dimorphism is common.  Test it with phylogeny.

> Behavior and Evolution

[09/27/99 09:08 AM]

> Basic Concepts in Behavioral Ontogeny

Animals inherit more than just genes -- they inheret a species-
specific environment.

Originally, the study of ontogeny was regarded as the study of
"instincts."  

>> What is an Instinct?

Instincts are thought of as adaptive traits that an animal is born
with; they are species-specific and don't require learning.  Usually
it responds to a very narrow class of stimuli.  It occurs the very
first time you need it.

People often conflate "genetic" with "innate" or "instinctive."
Often, people also think that instincts ...  The idea of instinct also
suggests that there is a strict genetic program that doesn't get
affected by the environment.  People think calling something
instinctive "explains" it.

You can ask where variation comes from (genes or environment), but you
can't ask whether genes cause something or whether the environment
does -- they always work together..

We need to know how behaviors arise -- we can't just decide something
is innate and then leave it at that.  

>> Relationship between Genes and Behavior

The relationship between genes & behavior is moderated through many
levels: genes, chromosome, nucleus, cytoplasm, tissue, organ,
organism, environment..  All of these layers can interact with each
other, and interact within the layers.  "Probablistic epigenesis" --
we don't develop in a lockstep, mechanical fashion.  Instead,
different layers interact in a nondeterministic way to give rise to
animals.

All species-specific behaviors have developmental histories, and
change over time.  Furthermore, different factors can affect them at
different points of development.

>> Ontogenetic Niche

Animals are formed and grow up in a specific environmental and social
setting: breeding grounds, food, breeding grounds, parents, etc.  This
environmental information is "inhereted" along with the genes.  Thus,
species-typical behaviors may depend (in part) on the fact that there
is a species-typical environment.

>> Ontogenetic Adaptations

Animal must adapt and survive in its ontogenetic niche.  Animals often
need very specific adaptations for different points in development.
Thus, some adaptations exist because they are necessary at one
specific point in an animal's development.

For example, in rats, we find that rats develop lactase early, but it
drops at about 15 days of age.  They develop other enzymes at about 15
days.  Thus, presence of lactase is an ontogenetic adaptation --
needed for nursing, not needed later.

At any point, an animal must develop both to survive its current
conditions, and to prepare to be able to survive later conditions.

> Egg & Uterus as Habitat

>> Features of Uterine Life

Tethered by an umbilical chord that provides nutrients and oxygen.
Live in a fluid-filled environment.  With growth, movement becomes
restricted.  For animals with multiple offspring, uterine is shared
with other embryos.  Stimuli available are mainly chemosensory and
touch.

>> Embryonic Behavior

Embryos do have organized behavior -- sequential organization,
temporal organization, spacial organization.

n.b.: the way you asses a problem can affect the answers you get.
E.g., if we study embryo behavior in utero, it increases and
decreases.  Might think this is a "genetic program," but is actually
caused by constriction of movement -- if we take the embryo out, it
still shows behavior..

>> Antecedents of Post-Natal Behavior
>>> Facial Wiping

If you insert small amounts of food into an embryo's mouth, it will
wipe its mouth -- antecedent of grooming.  Wiping behavior seems to
depend on amnionic fluid -- thus it drops off at birth, once the
animal is born.  Also, it can depend on the animal's position (suppine
vs. prone), since grooming can conflict with staying balanced.

>> Adaptations to Uterine Life

Some patterns of behavior we see in embryos don't seem to be
antecedents of post-natal behavior: ontogenetic adaptations

>>> Response to Umbilical Occlusion
If a fetus accidentally occludes the umbilical chord, it has a very
particular response.  First, suppression of fetal activity and
depression of heart rate.  Then increase of activity (head toss,
flexing body), then drop in behavior.  You don't find this behavior
after the rat is born.

>> Sensory Stimulation and Future Behavior
While in the uterus, the embryo has sensory stimulation (mainly sounds
and smells).

>>> The Duck Assembly Call -- Gottlieb
The assembly call is used by the mother to keep the flock together.
Basically it means "come here kids."  How do children recognize the
call?  If you put eggs in incubator (away from the mother), they'll
still prefer their mother's call.  Is the call hard-wired?  Over
development, the child starts vocalizing before birth.  The children
can hear both their own vocalizations and those of other eggs.  These
can prime a chick to follow assembly call..  If you put an egg alone,
and don't allow the embryo to vocalize, then the chick won't have a
preference for one call or another.

> The Transition to Being a Neonate
>> How Rat Pups Find their Mother's Nipple

Behavior is determined mainly by experience with the mother's odors:
if you wash the nipples, then the children won't attatch.  Odor source
comes from 2 things: mother's saliva, and amnyonic fluid.  Effects
very specific: e.g., the saliva of a virgin mother won't elicit
attachment.  After 4-5 days, pup saliva also causes attaching.

>> Learning about Odors In Utero

How does the pup know which odor cues to use?  Hypothesis: exposure to
amnyonic fluid causes this preference.  Try injecting an odor into
amnyonic fluid.  Then stroke the animal (simulates mother licking its
children) either in the presence of the odor or not.  The animals will
attach if the nipples smell like citrol.  But this effect doesn't
occur without the stroking.  Thus, child is learning that stimulation
(licking) in the presence of an odor means that the child should find
that odor and attatch.

> Mother as Habitat: The Consequences of Suckling

Young rats learn about who they should mate with in the future based
on what their mother smells like.  Rats which are capable of mating
smell similar to rats which just gave birth.  Rats smell their mother
to learn what rats that are capable of mating smell like.  Male rats
mate more efficiently with rats that smell like their mother.  To some
extent, the rats imprint their mother's odor & use it to find rats for
mating.

[09/29/99 09:08 AM]

!! First Exam is Next Wednesday !!
!! Paper Topics Due Next Wednesday !!

> Basic Features of Bird Song

>> What is Bird Song?

Defined by a long, often complex sequence of vocalizations.  Usually
produced by males in breeding season.  Different from "calls," which
are short, simpler, produced by both sexes, and produced outside
breeding season.   Bird Song is a broadcast signal -- not directed at
a particular recipient.  Used for teritory marking and mating calls.
"Oscine"=song bird.  Oscines make up about 1/2 the species of birds.

>> How do we study it?

Use sonogram.  Fourier transform over time.  Songs can be described in
a hierarchy.  Several song types make up a repetoire.  Each song has
distinct sections (=phrases)  phrases broken into syllables, which are
broken into elements (or notes).

> Bird Song is Learned

Other bird vocalizations (calls) usually aren't learned, but bird song
is.

>> Clues from the Field

Geographical variation of bird song within a species: "dialects."

>> Isolation Studies

If you acoustically isolate a bird from hearing bird song, they will
develop a degraded version of bird song.  Song is similar to adult
song, but is "messy," and lacks complexity.

>> Hearing Adults is Critical for Song Learning

A bird will sing the dialect of the adults around it.  In some
instances, you can teach bird A to learn bird B's song.

>> Deafening Studies

If you deafen a bird before it hears song, it produces very messy
song.

> The Model of Song Learning

>> Acquisition of Song (Sensory Stage)
>>> Auditory Template model of song development
Bird has template; it is matched to hearing its own species song.
This allows the bird to create an exact template.  The animal then
starts singing, and tries to match its song to the exact template.

>>> Evidence for a Template
  1. There is substantial structure in isolated bird song.
  2. Some structre is retained in deafened bird.
  3. Very little exposure is needed to learn bird song (song sparrows
	can learn in about 30 exposures in 5 min; nightingale can
	learn to 75% in 10 presentations)
  4. There are species-wide universal features (despite dialects)
  5. If you give a child a choice between its own species' song and
	another species' song, it will preferentially learn its own.
  6. Birds heart rate is higher when hearing their own song than when
	hearing another bird's song.

conspecific = a bird's own speceis' song; heterospecific = another
bird's species' song..

>>> What features do song birds use to learn?
Computer generate and manipulate syntax and syllables.  Some species
will mainly just learn syllables, and then put them into their own
syntax.  Others will learn both syllables and syntaxes..

>>> Timing of Acquisition (Sensitive Period)
The song must be presented in the "sensitive period."  Before or after
that time, the bird will not learn the song (originally known as
"critical period").  But if the bird is deprived of song past their
normal song learning period, they can extend the sensitive period..
Some birds (close ended learners) learn songs as juveniles & then keep
that song for the rest of their life. (to test, play different songs
at different points in development, and then record which song they
learn)  Others change their songs from year to year (open ended
learners).  Usually change to match the songs of other animals near
them.  For open-ended learners, there is an annual sensitive period.

>>> Social Effects
There are differences in learning between learning from a tape and
from a live tutor.  White crown sparrows will learn song sparrow songs
from a live tutor, but not from a tape.  Social stimulation also
broadens the sensitive period.  These effects vary based on species:
some species can learn equally well from tapes & tutors; others seem
to only be able to learn from live tutors.

Experiment: bird 1 can push 2 buttons to make 2 different songs play.
Bird 2 listens to the same songs as bird 1, but has no control.  Then
bird 1 learns song much better than bird 2.

Birds tend to learn better from aggressive tutors -- increased
stimulation and salience.

>>> How Accurate is song learning?
Animals can sometimes combine different phrases from different
tutors.  

>> Storage
Animals store information about songs that they learned for a long
time before they actually produce the song (as much as 300 days).

>> Retrieval and Production
Birds seem to pay attention to their own feedback.  If you deafen a
bird after acquisition, but before production, it will only acquire a
very degraded song.  Timing of stages of production are usually
species-typical. 

>>> Subsong
Initial bird vocalizations are fairly messy.  Structure is present,
but noisy.  People believe that in this stage, the bird is learning
to coordinate different systems (breathing, beak, tounge, syrinx,
etc).  There are species-specific differences in subsong.  These
differences may depend on auditory feedback.  Very hard to recognize
final song from sub song.

>>> Plastic Song
Plastic song changes over time, and approaches the tutor's song.
Characterized by overproduction of syllables.  Then syllables are
removed from the song: selective attrition.  If a bird has heard
conspecific and heterospecific songs, then it will loose the
heterospecific syllables more readily than the conspecific syllables.
Birds seem to get rid of songs that don't match their neighbor -- keep
the song that most closely matches their most active neighbor.

>>> Crystallization
Eventually, song will stop changing.  Deafening bird after
crystallization doesn't have as much of an effect.  Crystallization
correlates well with increasing testosterone.  Experiments suggest
that testosterone is necessary for crystallization.

#     subsong     plastic song   crystalized
#                   \_\_\_\_
#                  |C    \
#                  |B  B   \
# \_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_|A\_\_A\_\_A\_\_\_A\_\_A\_\_A\_\_A\_\_A
# ^  ^  ^  x  x     ^
# |  |  |  |  |     |
# A  B  C  D  E     A'

> Distribution of Song Learning

> Why Learn Songs?
We don't really know.. mating, territory marking?

[10/04/99 09:07 AM]

> Quantitative Genetics

>> Strain Studies

Inbred strains created through selective breeding -- each strain is
essentially genetically identical within the strain.

Inbred strains can tell us whether variation in a given trait is based
on variation in genes -- if animals from 2 strains raised in the same
situation always develop differently, that difference must be
genetically based...

Contextual fear -- animal will condition to its environment as well as
to the conditioned stimulus.  Do tone \to shock conditioning in setting
A, then check whether association is stronger or the same in setting A
or in some other unrelated setting.  Contextual fear varies based on
strains -- thus, it's affected by variations in genes.

Marr's watermaze: there is a vat of water with a hidden platform.  How
quickly will mice learn to swim to the platform?  Note that the vat is
surrounded by consistant visual cues.  Time to find platform, how long
mice remember where the platform is, varies based on strains -- thus,
it's affected by variation in genes.

>>> Environment also has an influence

Enriched environment vs. impovrished environment.  Enriched
environment improved performance of mice that are bad at maze
learning, didn't affect mice that were good.  Impovrished environment
degraded performance of mice that were good, didn't affect mice that
were bad.  

Even if environment is -very- carefully controlled for, which site an
experiment is run at makes a signifigant difference in the results.

In mice, which mother rears a mouse & which mother carries a mouse
(when it's pregnant) doesn't seem to have much effect on the mouse's
development.. (?)

>> Artificial Selection

If we do artificial selection on a trait whose variance is controlled
by 1 gene, we should see a bimodal distribution very quickly (within
the first few generations).

If we do artificial selection on a trait whose variance is controlled
by several different genes, we should see a much more gradual
divergence, and there should be lots of intermediates.

# R=h^2(S)
#  R = difference in parents (?)
#  S = difference in children (?)

> Forward Genetics

Makes no assumptions (?)
Develop a good behavioral profile, then randomly change the genes
(induce mutations)  See which changes cause deficits in behavior.
Examine what changed in the genome, & find which genes are necessary
for a behavior.

Forward genetics is more expensive, since you need a lot of subjects.

>> Bacteria

Bacteria will move towards certain chemicals (chemotaxis).
Hill-climining, tumble & roll.  
  - Detecting the odor
  - Movement generation

Induce random changes in genes, and see which changes cause deficits
in behavior.
About 20 genes involved in detection of the odor.

>> Flies

>>> How do you make the changes?

Three primary ways to induce mutations:
  - Chemical mutagens.  Generally cause very small deletions.  Single
    mis-pairings in base-pairs.
  - Radiation.  Causes rearangements, deletions, replications, etc.
    Causes a lot of different changes.
  - P-element (transposable).  P-elements can jump around in the
    genome.  Animals with p-elements have repressors in cytoplasm.
    But if you cross them with an animal with no p-elements, then
    there will be no repressors in cytoplasm, so p-elements will jump
    around and cause mutations.

>>> What are the results?  Odor -- shock association

You can train flies to associate odor with shock.  Flies that
associate odor with shock will tend to avoid that odor.  

It's possible to dissociate different types of memory using this
paradigm (9.03 whee).  Flies have STM, LTM, MTM (for odor-shock
association).  You can selectively eliminate them.  These experiments
let us determine more about the biochemical pathways involved.

>> Mice

>>> QTL: Quantitative Trait Loci

Find 2 strains that are particularly good and particularly bad at a
particular trait/behavior.

E.g., pre-pulse inhibition: if you get a little shock before a big
shock, you won't respond to the big shock as much as you would
otherwise.  One parent has high pre-pulse inhibition, the other has
low pre-pulse inhibition.  The F1 generation tends to have high
pre-pulse inhibition.  The F2 generation has bimodal distribution --
examine the genotypes of animals on either end of the bimodal
distribution, & compare.  Somehow this can tell us about what genes
are involved.

>>> Spontaneous mutations

We can do Mendelian-type studies on spontaneous mutants to see if
they're caused by 1 gene, or maybe 2 genes..

>>> Induced mutant screen

Induce mutations in a strain, and look for changes in a particular
behavior (esp. effective with robust behaviors).  When you find a
change (e.g., circadian rythm gets signifigantly longer), clone the
animal & look at its genes to try to figure out what the mutation was
that caused it...

> Reverse Genetics

 * Good way to target genes.
 * Effects tend to be global.
 * Can end up with compensation, etc.
 * Expensive.
 * Problems with multiple genes involved in behavior.

>> Flies

>>> How do you do it?

  - Directed mutagnesis
      Take a p-element, and use a targeted vector to insert it in a
      particular point.
  - Conditional expression
      Use promotors (only activated in some contexts)
  - Tissue-specific expression

>>> What do you see?  Courtship behavior

>> Mice

>>> Knock-outs

Use a vector to insert a marker and something to disrupt the gene.
Then inject these cells back into a blastocyst.  This will create a
mouse where part of it has the gene active and parts have the gene
disabled.  Furthermore, which part is which is marked.  "chymera"

>>> Transgenics

Same idea, but add new genes to an egg.  That way we can create a new
mouse with an added gene.

>>> Variations on a theme

We want to know what the gene that we changed does on a global scale
-- therefore we want to do a whole battery of tests, to characterize
the change more completely.  Example tests: water maze, contexutal
fear, activity, rotorod, wire hang, footprint (=test gait), order
discrimination, etc.

>>> What do you see?

We see that we're out of time.

> Conclusions

Not today.

[10/13/99 09:09 AM] Sexual Selection I: Intrasexual Competition

> The Basics of Sexual Selection

>> Problems faced by Linnaeus and Darwin
Linnaeus thought that there were two "species" of duck which looked
very different.  Later it was realized that one was the males and the
other was the females.  Males and females can vary dramatically in
both appearance and behavior.

Darwin noticed that there are certain traits (singing, bright
displays) which appear to reduce the probability that the a particular
animal will survive.  How do such traits evolve?  Make distinction
between natural selection and sexual selection.  Sexual selection
depends, not on a struggle for existence, but on a struggle for access
to females.  The consequence of failure is not death, but very few
offspring.  Note that normal natural selection and sexual selection
can push in opposite directions.

Divide sexual selection into 2 parts: intersexual selection and
intrasexual selection.  Intersexual selection = female choice;
intrasexual selection = male-male competition.  (in less common
situations, intersexual=male choice, intrasexual=female-female
competition).

>> Why does Sexual Selection exist?

>>> Differences in Reproductive Potential
Difference in the size of gametes.  Females produce few eggs and
produce large ones.  Thus, females invest more.  Males produce more
sperm than females produce eggs.  Thus, males could potentially
reproduce much faster than females.  A male could increase his
reproductive success by finding & fertalizing lots of females.  Thus,
females are a scarce resource.   In contrast, a female can only
increase her reproductive success by speeding up her production of
young.  Generally this means getting more food.

>>> Differences in Parental Investment
Females tend to invest more energy in increasing the welfare of their
offspring; males tend to invest more in trying to make more offspring.
Where one sex invests signifigantly more than the other, the latter
sex will compete for access to the former.

>> The Intensity of Sexual Selection
The intensity of sexual selection depends on how easily one sex can
get access to the other sex.

>>> Differential Effort between the Sexes
If males & females invest the same amount in young, sexual selection
won't be as intense.

>>> Operational Sex Ratio
Operational Sex Ratio = the ratio of # of males & # of females who
come into breeding condition at any given time.  If all females become
available for breeding at the same time, then more males will have
access.  If females become available in waves, one group of males can
monopolize each wave.  To the extent that the operational sex ratio is
closer to 1, sexual selection is less intense.

>> Intrasexual vs. Intersexual Selection
>>> Problems
It's hard to distinguish intrasexual from intersexual selection.  For
example, there are pretty birds in some birds.  Seems at first that
these are probably just for intrasexual selection.  But they use them
to fight.  Thus, they're also intersexual selection.  Another example:
bird with spurs.  The spurs look like they'd be use for male-male
competition.  But spur length doesn't correlate with dominance.  It
does, however, correlate with female choice.  Thus, spur length is
probably involved in intrasexual selection, not intersexual
selection.   Moral: you have to do experiments to tell if something's
intersexual or intrasexual; you can't just tell by looking at the
trait.

>>> Within-species Variation
Within some species, there can be 1 population that uses more
intrasexual selection, and another that uses more intersexual
selection.  For example, to the extent that there are fewer nesting
sites available, we expect stronger intersexual selection, since males
must compete for nest sights.  This leads to populations that are more
dimorphic where nests are less common, and populations that are less
dimorphic where nests are more common.  In summary: increase nests,
and there's more room for intrasexual selection (females can be choosy
about which male to mate with); decrease nests, and there's more room
for intersexual selection.

> Intrasexual Selection

>> Scramble Competition
When females are scattered in space, main intersexual selection
involves _finding_ the females to mate with them.  

>>> Traits Selected for 
Intersexual selection will lead to persistance in searching, good
spacial memory, good ability to percieve females, mobility.

>>> Thirteen-lined Ground Squirrels
track ground squirrels before, during, and after breeding season.
Area covered increases during mating season, drops off before & after
it.  The amount of area covered accounted for almost 1/2 the
reproductive success in reproduction.

>>> Common Toads
Explosive breeder (i.e., reproduces in a very short time frame).  All
reproduction must be done in a very short time period.  Females come
to pond to mate.  Males can either intercept females before they get
to the pond, or can displace other males once the females get to the
pond.  Larger males are not more successful at intercepting females,
but they can interrupt other matings.  The larger you are, the less
likely you are to be displaced.  

Thus, there are 2 mating phases: scramble phase, where you try to get
to the female first, and other phase, where you try to use your size
to displace other frogs & get mating.

>> Endurance Rivalry
The male who stays around longest (at breeding sites) is the most
likely to get mates.

>>> Traits Selected for
Intersexual selection will lead to male energetic effeciency,
endurance.  It should select for larger (fatter) animals, and for the
ability of animals to fast.

>>> Grey Seals
Copulations per day correlates with initial body mass (body mass at
the beginning of the reproductive season).  The larger they are, the
longer they can spend at the mating site, & the less time they have to
spend feeding.

>>> Elephant Seals
One of the most sexually dimorphic species among mammals.  Males are
very very large.  Females must give birth on land (they mate there
too).  Very few good spots on land.  Thus, females cluster on these
breeding areas.  Males stay on the breeding area for the entire
breeding season (3 months), and don't eat at all.  But they must have
enough energy to defend the females and to mate.  The dominant males
loose their mass at higher rate than less dominant males.  Often the
dominant male will loose 1/2 his mass by the end of the breeding
season. 

>> Contests or "The Law of Battle"
Males can compete with each other directly for access to females:
either attack each other or have contests & let females choose.  Males
can either compete directly for males or can compete for some resource
that females need (e.g., breeding sites)

>>> Traits Selected for
Size, weaponry, threat signals, alternative mating strategies.

>>> Red Deer of Rhum
Males and females differ signifigantly in size.  Males have large
antlers.  Deer are polygenous.  Breeding season is very short.  In
females, reproductive success is mainly dependant on how long a female
lives.  Male reproductive success is mainly dependant on ability to
get access to females.  Male reproductive success doesn't seem to be
determined by female choice.  Fighting success correlates very well
with how long a harem can be held.  Fights mainly occur when males are
of similar size.

Assesment behavior: roaring, parallel walk.  Animals holding
territories tend to roar more.  Animals that are in better condition
are much better at roaring.  Thus, it's an honest signal of your
ability to fight.

>>> Elephant Seals
Most successful holders are the largest.  Big harems are formed first,
smaller ones later.  Fighthing increases directly after formation of
harems.  

>>> Common Toads
Male toads assess the size of another toad before they attempt to
dislodge them.  Smaller males are unlikely to try to displace larger
ones.  How does assesment work?  Frog croak becomes lower-pitched as
the frog gets bigger.  Thus, croak is a good signal to use to assess
the size of frogs.  If we silence frogs, and either play high-pitched
croaks or small-pitched croaks, competitor males's attack rate will
depend on the recorded croaks.  It also depends on the size, though,
so they must have some other ways of assessing size, too.

>>> Something Different: female-female competition
If males are a scarce resource (esp. if males provide parental care),
then females may compete with each other for access to males.  E.g.,
in some hen, males incubate the eggs, & females compete for good
males.  What makes a good male?  Fatter males -- must incubate for a
long time without eating much.  Also short males.  (unclear why they
want short males)

>> Sperm Competition
Competition between the sperm of several males for the ability to
fertalize a female's egg.  Only present when females mate with
multiple males.  In general, there is a last-male-preference: the last
male to copulate will have the most offspring.

>>> EPC's: Extra Pair Copulations
Sperm competition selects for copulation outside of normal pairings.
EPC's are quite common in nature.

>>> Mate Guarding
Selects for guarding your mate from having any EPCs.  

>>> Retaliatory Copulation
Retaliatory copulation: if a male is guarding a female, and sees
another male copulate with her, the guarding male will copulate with
the female a lot to try to decrease the chances that the female will
have the userper's children.

>>> Cloacal Pecking in Dunnocks
Males force females to eject sperm from previous males.

[10/18/99 09:13 AM]

> Female choice
>> Examples
>>> Peacocks
Peacocks have lots of eye spots.. Correlation between # of eye spots
and # of matings.

>>> Other birds
There seems to be female choice in a bird w/ long tails.  Seems that
males don't use tails except to display to females -- no male/male
aggression effects.  Text experimentally: if we shorten their tail,
they get less mates; if we lengthen their tail, they get more mates.
Thus, females are choosing males on the basis of tails.  Evidence that
there are no male/male agression effeects: when tail lengths were
modified, sizes of territories didn't change any.

>>> Guppies
Where there's lots of predation, guppies are duller.  When there's not
as much, male guppies have bright spots.  Hypothesis: females choose
on the basis of male spots.  Find proportion of male displays that
elicit reactions.  Test males both under normal light & under orange
light -- under orange light, their spots are hard to see.  Also test
under blue light -- in blue light, the orange spot looks black, but
the contrast is still present.  Under blue filter, females as much as
under normal light.  So it's the presents of spots (regardless of
color) that attracts the females.

>>> Bower birds
Bower = pretty house thingy that birds make.  Male makes elaborite
bower which is basically just used for mating.  Males with the most
highly decorated bower gets the most mates.  Experimentally try
removing decorations from bowers -- males get less matings.

> Direct Selection

Direct selection = females gets direct gains by choosing certain types
of mates.  

>> Selecting on correlating traits
Males sometimes produce some sort of resource to the female.
Sometimes the amount of resource depends on the trait that females
select on.

In one species of bird, males that are more red are chosen over males
that are less red.  In nature, there's a correlation between how red a
bird is, and how well it feeds its offspring.  Thus, females select
males that are redder because it means they'll feed her offspring
better. 

>> Nuptual Gifts
Sometimes a male gives a female a "nuptual gift," and the size of the
gift determines how much the female will mate with the male.

>> Choosing a particular male can decrease costs
Choose a male that's close by -- takes less energy, & might get
eaten.  Choose a male without parasites.  

>> Complementary size/genetics
Choose a male with complementary size & complementary genetics.  E.g.,
in one animal, females choose males that are about 70% of their size
-- if male/female size is too different, then matings won't be
successful.

Complementary genetics -- in one bird, tested what animals they
prefer: siblings, 1st cousins, 2nd cousins, 3rd cousins, etc..  They
tend to spend the most time with 1st-2nd cousins.  This may be an
example of "optimal outbreeding."  If you mate with people too closely
related, you can bring out homozygous traits..  If you mate with
people too distantly related, it's bad too.. ?

>> Sensory exploitation hypothesis
Sensory systems have evolved to suit a number of purposes.  Males can
evolve traits to explit biases in sensory systems.  For example, water
bugs use water vibrations to find prey.  But males that can imitate
these vibrations get more matings.  This hypothesis makes a very
specific prediction: female preferences evolve prior to males having
the traits.  How can we test?  Look for female attraction to specific
calls..  

In a certain bird, females have optimal hearing at frequency f, but
males produce a call at frequency f+k.  Why the mismatch?  Because
males are trying to match, but haven't succeeded??  Female always
prefers smaller k (even if other traits of the call are wrong).  True
across several species.  

Take a phylogenetic tree, with 4 ancestors.  One has a call with whine
+ chuck.  Females prefer males with larger chucks.  Others have calls
with just whines.  So the question is: in that one, did the males add
the chuck after the female had the preference?  Try playing
whine+chuck to the species where males don't produce chuck.  Then the
females prefer these sounds!  So the females had preference before
males made call.  

> Indirect Selection

>> Leks
Females choose particualar mates on a lek.  But the males don't
provide anything.  Why?  

>> Runaway selection
One thesis -- runaway selection.  Assume that there is heritable
variation both in trait and in trait selection.  Self re-inforcing
selection: males tend to mate with females that have selection, and
females tend to mate with males with trait.  No real support for this
hypothesis -- maybe treat it as a null hypothesis.

For example, sticklebacks.  Differences in dull & bright males, and
differences in preference.  Found high correlation between father's &
son's intensity of red.  Also, found correlation between father's
intensity & daughter's ..?

>> Judging male by traits that corellate w/ fitness
Another idea: Females can select healthy males by seeing how much of a
handicap he can live with.  If you have the trait, & are still alive,
it shows that you can overcome selection against it -- good genes.  Or
amount that you can invest in the trait depends on how healthy you
are. 

>>> Offspring Survivability
Look at relationship between a male's trait & the offspring's
survivability.  At least in some cases, they're correlated.  Not
correlated with foster-fathers.  Thus, genetic...  Proportionately
more offspring survive if their genetic father has bigger stripe.  In
peacocks, when father has more eyespots, the (male) children tend to
grow more.  Also, size of eyespots of father correlates well with
offsprings' survivability (probably because of weight).

>>> Parasite Loads
Look at whether females choose males on the basis of parasite load.
Direct benefits: they won't get parasitized.  Indirect benefits: their
offspring may be less suceptible to parasites.  Seems to be true in at
least some species.  For example, in a bird with tail streamers, the
males with larger tail streamers do much better w/ female selection.
Also produce more offspring.  Found that, by changing parasite load,
the tail streamer length can change.  Males that find mates tend to
have lower parasite load than males that don't.  In extra-pair
copulations, males usually have very low parasite load.  Tail length
says something about how well the male can protect vs. parasites?
Correlation beteen father's tail length and offsprings' parasite
load.. even if offspring is raised away from father.  No correlation
between adopted father's tail length & offsprings' parasite load.

>>> Fluctuating Asymmetry
Look at why asymmetrical males are less preferred than symmetric ones.
Some males are more symmetric than others.  Symmetry seems to show
that you can develop well, even when there is envrionmental stress
during development.  Adding mites give asymmetric tails.  Removing
mites gives more symmetric tails.  So symmetry may be a "fair measure"
of male fitness..

[10/20/99 09:11 AM] Mating Systems

> Basic Aspects of Mating Systems
>> Definition
>> Modern conceptualization
>> Types of Mating Systems
  * monogomy
  * polygamy
    - polygany
    - polyandry
  * promiscuity

> Determinants of Mating Systems

>> Physiological constraints: birds vs. mammals
What factors effect which mating system an animal uses:

  * physiological constraints / phylogenetic constraints.  For
    example, in birds, parental care can be done by either sex.  Male
    can increase reproductive success by providing parental
    care.. Thus, about 90% of birds are (nominally) monogamous.  In 
    contrast, in mammals, the mother puts in signifigantly more energy
    into the wellbeing of their offspring: mother carries child for a
    long time, and then provides milk.  Thus, a male doesn't loose as
    much if he deserts.  About 90% of mammals are polygenous.

  * ecological factors: how are males & females dispersed in space.

>> Ecological Factors

>>> Dispersal of males and females
Females tend to disperse according to where the resources that they
need are.  Males tend to disperse according to where the females are.

>>> Example of the Grey-Sided Vole
Both males & females form territories.  Experimenter found small
island, removed all rodents, and experimentally introduced voles to
the island.  Females' territories were spread-out and were
non-overlapping.  Then add clumped food sites.	Females change their
territories to be around the food sources & to be small.  Males follow
the females.  If you artificially force females to have certain
ranges, males will go to ranges with greatest female density.  If you
artificially force males to have certain ranges, females will ignore
them.  

>>> Economics of female defence
It's easier for a male to monopolize females if they're clumped
(vs. evenly distributed).  Also, operational sex ratio is important.
Synchronous breeding gives low operational sex ratio; asynchronous
breeding gives higher operational sex ratio.

> Mammalian mating systems: Male care?

>> Monogomy: Obligate vs. Facultative
In males, monogomy is odd, since females invest so much more.
Obligate monogomy = children will only survive if both of their
parents provide them with care.  Facultative = a male isn't really
necessary for parental care, but it's mainly because the male can't
find other mates for some reason.  For example, females widely
dispersed, high predation pressure..

>>> Djungarian hamsters (Obligate)
In Djungarian hamsters, signifigantly fewer offspring live with 1
parent, even given all the food etc. that they need (raise them in the
lab).  Another experiment: put them in cold room (similar to the
environment they would normally be in).  Then children can't survive
without the thermal regulation provided by the father.

>>> Kirk's Dik-Dik (Facultative)
Small animals, widely dispersed, high predation pressure.  Monogamous.
Facultative monogomy?  Other things we can rule out (at least in this
case): males are defending food resources; ability to detect
preditors; protection against infanticide).

>> Polygyny
>>> Female solitary and defensible (~60% mammals)
Males can defend more than one female.  Males can form territories
which let them monopolize several females

>>> Female solitary and not defensible
Female ranges are too wide for males to defend them.  Males & females
mainly associate with each other during reproduction.  

>>> Females social and defensible
Male can monopolize a single group of female.

>>> Females social and defensible: predictable vs. unpredictable mvmt.
Males can't defend the whole region.  But they can anticipate places
where females will go, e.g., watering holes.  Males can follow the
group of females \to harems.

> Leks

>> Features
Males have tiny territories with no apparent value.  Males are very
aggressive and very bright.  Females visit lek, look through the leks,
mate with one, and leave.  Huge reproductive skew: dominant male gets
almost 1/2 of all matings on the lek.  (Leks are fairly rare)

>> Why do leks occur?
When males can't defend a female, or defend her resources.  Species
that lek tend to have much larger ranges, lower density.  There is an
inverse correlation between female range size & male range size,
esp. for very large female range sizes.

>> Why do males aggregate onto leks?

>>> Hotspots
Males might aggregate at places like watering holes, where females
will come to them.  In some species, leks tend to form in places where
female ranges overlap.

>>> Avoid Predators: tundra frogs
Leks reduce predator pressure.  If you're in a crowd, you're less
likely to be picked off by a particular preditor.  For example, tundra
frogs are eaten by bats.  As the lek size increases, the predation per
individual animal decreases.  

>>> Hotshots
Males tend to aggregate around a very good male.  Other males
parasatize his ability to attract females.  "Satellites."  Then they
may be able to get matings with the females attracted by the hotshot.
  * Natterjack toad: small males can't produce loud vocalizations.  So
    they hang around bigger males.  In these toads, if a toad is 1/2
    as loud as another male, he tends to become a satellite.  
  * Sharp-tailed grouse: males seem to want to be in the middle spot.
    Females seem to be attracted to the male in the middle spot.  The
    males are competing to get to the middle spot.  So is it really
    hotshot or is it hotspot?
  * Fallow deer leks: if you force successful males to move (even to
    periphery), they still remain successful.  This is evidence for
    hotshot.  

> Avian Mating Systems: Role of Male Care
Birds are primarily monogamous.  Males provide care.

>> Monogamy: Causes?

>>> Obligate?
In some sea birds & other species, it takes both parents to produce
any offspring at all.  If you lose your mate, you may lose your
entire clutch.  

>>> Male care important?
But in many bird species, single mothers still have a fair amount of
success (although it's almost always less than a couple).  Often
females will compensate for male not being there by getting more food
(as much as the male would have), but time spent defending the nest &
caring for the young goes down.  In a few monogamous species, removing
the father doesn't decrease the success of the children.

>>> Why are males monogamous?
Often, even though a species is monogamous in its normal environment,
it will be different in other environments: if you decrease the # of
males enough, males become polygynous. 

>>> Sperm competition and egg dumping
  * Mixed strategies.  There are a lot of EPCs in some species.
    Amount of EPC varies a lot from species to species (0%-35%).
  * EPC's: Zebra finches.  Figure out who's the most attractive male.
    Randomly pair females with random males.  Females will have more
    EPCs with males that are more attractive than their own males.
    Suggests that they're looking for good genes.
  * Cliff Swallows as brood parasistes.  Females can dump eggs in
    other animals' nests, and also raise your own, you can have more
    offspring.  Egg dropping especially prevalant in large colonies.  

>> Polygyny: Causes?
Males can monopolize resources that females need.

>>> Costs or no costs?
Often males don't provide any parental care, so they don't care who
they mate with.  But sharing a male means sharing a male's parental
care and monopolized resources.

In some species, females will only mate with already-paired males if
all the males with resources are already paired.  

>>> Polygyny threshold model: evidence and problems
Sometimes there's a choice: second female on a good territory, or an
only female on a bad territory.  Basically, predict that there will
basically be a theshold where polygyny with a good territory becomes
more profitable than monogomy with a bad territory.

Problems: if you remove the first female, it doesn't result in
increased female settling.  So size of group is decreased.  But the
females' success doesn't seem to decrease.

Sometimes males set up 2 territories, far apart, and decieve females
into thinking that the male has no mate.

[10/25/99 09:09 AM] Mechanisms of Mating Behavior

!! OUTLINE DUE NEXT WEEK !!
!! EXAM ON THE 10th !!

> Neurological/Physiological Mechanisms of Mating Systems

Why, among closely related species, do some pair-bond, and others
don't?  

>> Voles as a model system

A small rodent.  Some voles are polygamous, others are monogamous.

  * The parie Vole is monogamous
  * The montane vole is polygamous

>>> Evidence for monogamy
  - Set up traps, and see how many times you catch males and females 
    together.  If you catch males & females together, then that's
    evidence that they form pair bonds.
  - Fathers show more paternal behavior in prarie vole: evidence for
    pair bonds.
  - If you raise them in a large enclosure, prarie voles spend lots of
    time in male/female pairs.
  - Put a male with a female for a while.  Then give the female the
    choice of two males: her old partner, and a new animal.
    Monogamous animals will prefer the old partner over the novel
    animal. 
  - Males can become very agressive (to intruders) after mating.
    Agression applies to any vole except the vole that was mated
    with. 

>> Differences in the brains of these two species
>>> Oxytocin and vasopressin affects pair-bonding in prarie voles
Prior evidence suggests that increasing oxytocin increases how much
animals associate with each other.  Look at oxytocin receptors in
prarie voles and montane voles.  Monogamous voles tend to have similar 
oxytocin receptor distributions, and polyganous voles tend to have
similar receptor distributions.

Another peptide: vasopressin.  Difference between vasopressin receptor 
distribution between prarie and montane voles.  These differences
appear fairly early in development.  

>>> Peptide and antagonist infusions affect pair-bonding in voles
Inject female with oxytocin.  Expose a montane vole to a male for 6
hours.  (Normally this won't lead to pair-bond formation)  Then later
(after oxytocin is gone), test preference of female.  Then the female
will show a strong preference for the male she was exposed to.  Thus,
oxytocin release during time of mating seems to be important to the
formation of female pair-bonds.  

With males, vasopressin seems to cause pair-bond formation.  Inject
males with vasopressin-blockers, then males won't show post-mating
aggression formation.  Also, they won't show preference for the female
they mated with.

Also, try putting males & females together for a short time (not long
enough for them to normally form a pair-bond), & inject then with
oxytocin(female) or vasopressin(male).  Then they'll form a pair-bond.

>>> Manipulations have little effect in monotane voles
Try these experiments with the montane vole (polyganous).  They show
no signifigant effect.  Probably because the montane voles don't have
the right receptor distribution.

>>> Genetic manipulations and affiliative behavior
Cloned the genes for oxytocin/vasopressin receptors.  There are
differences in the genes.  Try transplanting the monogamous form of
this gene into mice.  Then we get mice with similar receptor
distributions.  Now, inject vasopressin.  Then mice get higher
affiliative behavior.

> Basics of Sexual Selection and Gonadal Hormones
>> General Issues
Normal course of mating:
  - Females engage in "Proceptive behavior": behavior that makes
    mating more likely to happen (indicates her willingness to mate;
    i.e., indicates that she is receptive).
  - Males examine females, determine whether they're receptive
  - Mounting
  - Intromission (inserting penis into vagina)
  - Ejaculation

After mating, the female often becomes agressive (esp. in polyganous
species).  

>> Description of Male Rodent Copulatory Behavior: Variations
Variations: 
  - copulatory lock?
  - thrusting?
  - multiple intromission?
  - multiple ejaculation?

>> Description of Female Rodent Copulatory Behavior: Lordosis
Curvature of the spine, to allow males to mate.

>> Role of Gonadal Steriods in the Regulation of Sexual Behavior 
>>> Types of Gonadal Steroids
  - Androgens: testosterone, etc.  In general, stimulate male
    characteristics.  
  - Estrogens: estrodyle, etc.  In general, have femanizing effects.  
  - Progestons: progesterone, etc.  Acts in uterus to allow for
    successful fertilization or birth or something.

>>> Properties
All of these hormones come from cholesterol.  The steroids can be
transformed into each other fairly easily.  E.g., there are enzymes
that transform testosterone to estrodyle.  Hormones are fat soluble,
so they spread everywhere in the body.

Steroids attatch to DNA, and stimulate certain types of protein
production. 

>>> Regulation of Steroid Release
Production of gonadal steroids is regulated by lutenizing hormone
(LH).  LH is regulated by LHRH (LH releasing hormone, AKA GNRH).
This happens in the pituitary, in the brain.

The gonadal steroids act as a general overall signal which controls
mating behavior.  

>>> "male" vs. "female" hormones

Hormones aren't really "male" or "female."  They just do different
things in each sex.

>> Why do Hormones Regulate Reproductive Behavior?

> Male Sexual Behavior: Hormonal Mechanisms
>> Berthold's classic example on roosters
>> Remove and Replace Experiments
>>> Findings
If you remove testes, & replace them, then sexual behavior develops
normally (thus, not caused by neural mechanisms, because neurons were
cut).  If you remove them & don't replace them, sexual behavior goes
down.  Inject testosterone after castration, and sexual behavior
appears again.

>>> Caveats
Some animals show sexual behavior without gonads, esp. if they have
mated prior to castration.  Also, there are individual differences in
sexual behavior that are independant of testosterone.  Only a little
bit of testosterone is needed, and providing more doesn't have much of
an effect.

>> Testosterone and Estrogens
>> Variations: the Red-Sided Garter Snake
Canadian snake that hibernates for up to 9 months.  Males store sperm
from the previous year.  Mate at the beginning of the season, when
testosterone levels are low.  Then as the season goes on, testosterone 
increases, & gonads enlarge, so that at the end of the season the male 
will have sperm to store for the next year.

> Female Sexual Behavior: Hormonal Mechanisms
>> Cyclicity
Unlike males, females tend to have cyclic periods of receptivity (not
all -- reflex ovulation = ovulate just when you get male
stimulation).  

>> Estrogens and Progestins in the Regulation of Sexual Behavior
Increasing estrogen tends to increase receptivity.  In general, in
order to cause receptivity, you have to simulate the normal cycle of
hormones.  Normally, estrogen for some time, then progesterones.

>> Female Rats Control the Pace of Mating
At first, it seemed that males controlled the pace of mating.  But if
you put them in a more natural environment -- i.e., a large, social
environment -- then the females will control the pace of mating.  If
mating occurs at the pace of the female, pregnancy is much higher.

> How do Hormones Affect Sexual Behavior?
>> Act Directly on Brain and Spinal Cord
>>> Steroids and the CNS
Steroids are fat-soluble, so they can go through the blood/brain
barrier.  Receptors in hypothalamus and other brain areas.

>>> The hypothalamus: AH-POA and the VMN.

The anterior hypothalamus/pre-optic area (AH-POA) seems to be related
to male mating -- ablate it and mating behavior goes away. (but
they're still interrested in females)

ventro-medial nucleus (VMN) seems to be related to female mating --
ablate it and mating behavior goes away.

Neurons in these areas seem to be activated during copulation.  

Steroids are important for the behaviors of these areas.

The amygdala seems to be related in making males want females.  Get
rid of amygdala, and the male will mate with females that are present, 
but won't try to find females (c.f. AH-POA, where males will try to
find females, but won't mate).

>>> Stimulation Studies
>>> Neuronal Activation: c-fos studies
>>> Intracranial Implants of Steroids

>> Act on the periphery
>>> Increased somatosensory input
Hormones can increase the receptivity to certain types of
somatosensory input.  E.g., female rats will have increased input from 
stimulation on the back.

>>> The muscles Involved in Penil Function

>> Beach's Model of Copulatory Behavior
  1. Motor components are coded in the spinal chord -- if you cut the
     brain from the spinal chord, then mating behaviors persist.
  2. Brain inhibits the spinal reflexes.
  3. Sex steroids act do dis-inhibit (i.e., to inhibit the
     inhibition). 

> How Do These Sex Differences in Copulatory Behavior Arise?
>> Basics of Sexual Differentiation
Sexes have some sort of differences.  In mammals, XY vs. XX.  In many
reptiles, the temperature in which the egg was incubated leads to
male/female.  Either way, the gonads produce a signal to tell the rest 
of the body what sex to become.  Default sex (sex if gonads are
removed) is usually female.

>> The Organizational Effects of Hormones: Phoenix et al.
Exposure to hormones early in development causes different
organizations to form.  

Females exposted to high levels of testosterone as embryos/neonates
will show decreased lordosis as an adult, and will sometimes mount
other females.

If you remove gonads from neonatal males, then they will show female
sexual behavior, and no male sexual behavior.  Give them a single
injection of testosterone, and they will show normal sexual behavior.

Females that are between males in the uterus develop more masculine
characteristics, are more aggressive, develop larger territories, and
are less attractive to males.  Females that are between males give
birth to more males, and females that are between females tend to give 
birth to more females.

Similar effects for males.

>> Trans-Generational Effects
The 2m-stuff and 0-m stuff can go across generations.

[10/27/99 09:04 AM] Evolution and Mechanisms of Parental Behavior
> Differential Investments in Sons and Daughters
Parental investment is often necessary for the successful production
of offspring.  

>> Why should animals treat sons & daughters differently?

Animals often give more support for one sex than for other.  In
particular, if supporting one successor will increase your overall
success than supporting another, then support the former.

>>> The Red Deer of Rhum: males favored
Variation in male success is greater than variation in female
success.  Strong males can have many children.  Adult size of male has 
a strong effect on their success.  Therefore, feed male offspring
more.  

Breeding success of a female's son increases with her status (\to the
amount of investment she can put in children), but her daughter's
success doesn't vary much.

>>> Bonnet macaques: females favored
Breeding success of a female's daughter increases with her status (\to
the amount of investment she can put in children), but her son's
success doesn't vary much.

>> Evidence for Differential Investment?
>>> Sex differences in juvenile size and growth
Sexual dimorphism is present from birth.

>>> Sex differences in energetic costs of rearing
In sexually dimorphic species, the energy intake of males is often
greater than the energy intake of females.   Males spend more time
suckling, etc.

>>> Differences in reproductive costs of rearing males or females
If a female has given birth to a male the preceeding season, then she
is less likely to have a child in the next season.  Also, it takes her 
longer to conceive.  Also, more females die if they reared a male the
previous year.

>> Sex ratio variation
If resources are plentiful, invest in males; if resources are scarce,
invest in female.

>>> Why vary sex ratio?
If it's more costly to produce one sex than another, then it might
make sense to vary your sex ratio.  There is often a trade-off between 
the number of each sex you produce and how expensive each is.  For
example, if you're in bad shape, produce 1 daughter.  Slightly better
shape, 1 male.  Slightly better sex, 2 females. Slightly better shape, 
1 of each.  Etc..

>>> Does this happen?  Red Deer, Bonnet macaques, and Baboons
In low-dominant animals, males are much less likely to survive than
females.  In high-dominant animals, it's about the same.  Dominant
females seem to produce more males than females; subordinate females
seem to produce fewer males.

In species with matrolines, dominant females will produce more
females, and subordinates will produce more males.

>>> Scarce resources and investment
Amount of resources available affects the sex ratio.  Reduced food
intake \to smaller litter, and fewer males.  This effect holds if the
female was food-restricted during development but not during
breeding.  It also holds across generations: the daughters will have
fewer females..

>>> Mechanisms
This seems to occur because there are many more male spontaneous
abortions when food is scarce etc.  Also, if food is low, then males
have a higher mortality rate.

> Maternal Effects on Offspring
>>  Maternal Anogenital licking (AGL) in rats
>>> description and function
Mothers lick their offspring to clean them off.  Females lick the
anogenital region of offspring to reclaim water, since lactating
females need a lot of water: reclaim it from offspring's urine.  

>>> Sex-bias in AGL
Males are licked more than females.  If you put male urine on a
female, then the female gets licked more.  If you mask odors,
preference goes away.  Thus, it seems to be smell cues in male
juvenile urine that causes males to be differentially licked.  This is
a testosterone-dependant effect.

>>> Consequences
You can manipulate the amount of licking a mother does, either by
giving her dilute saline, or by eliminating her ability to smell
urine.  The males that recieve less AGL have poorer copulatory
efficiency.  This is due to a decrease in the number of motor
neurons.

Take 2 strains of rats with different salt tolerances.  One does less
AGL than the other.  The ones that are licked more succeed better in
producing offspring.

>> Differential effects on sons depending on who the father is
You can make zebra-finches more or less attractive by giving them
leg-bands.  If they're mated to a more attractive male, they'll
produce more male offspring.  They can also bias clutches in another
way: if they mate with an attractive male, then they'll add more
testosterone to their eggs.  More testosterone: grow faster, beg for
food more intensely, and are more likely to be dominant.

>> Effects of maternal care on adult "personality"
More AGL can change a rat's response to stressful situations.
Children with more AGL will have lower level of stress when put in
stressful situations.  Also, rats that have recieved more AGL explore
their environment more.  Rats that recieved more AGL will be more
likely to eat food in a novel environmnet.

> Experiential and Hormonal Mechanisms of Rat Parental Care
>> Basic types of maternal care in mammals
All mammals lactate (the only food that their offspring can consume).

Three categories
  - Female gives birth to altricial young: highly undeveloped young
    that are left in a nest.  Female provides milk, warmth, protection
    from predators. 
  - Females give birth to procotial young: children are fairly well
    developed, have independant mobility.  Usually have a nest.
    Either (1) the offspring hides in the nest & the mother comes back 
    with food etc; or (2) the offspring follows around the parent.
  - Females give birth to semiprecotial young: in between.  Like
    humans.  

>> Effects of experience on rat parental behavior: concaveation
concaveation: if you put a pup with an animal long enough, it will
eventually show parental behavior.  Not hormonally caused: still
occurs if you reduce gonads, etc.

Present pups to an animal, and see how long it takes for them to show
any parental behavior.  If a mother has just given birth, she will
show parental behavior almost immediately.  This is caused hormonally.
If you've never given birth, then it will take you a while to show
parental behavior.

>> Effects of hormones on induction of rat parental behavior
>>> Clues from natural fluctuations
Over pregnancy, estradiol increases right before birth; progesterone
drops right before birth.  Check to see if these changes cause the behavior?

>>> Parabiotic evidence
Connect the blood supplies of maternal and non-maternal females (thus
all signals must travel via blood \to hormones).  Then the non-maternal 
female will be paternal more quickly.

>>> Hormonal manipulation and hysterectomy
See if pregnancy termination causes female to act parentally.  Remove
placenta \to progesterone drops quickly.  Conditions seem to be an
increase in E followed by a drop of P.

>> What maintains parental behavior?
>>> Non-hormonal mediation
Even though the rapid induction of parental behavior by mothers who
just gave birth is hormonal, the maintanance of that behavior isn't.
Test with parabiotic experiments.  Remove hormonal organs of a
rat w/ maternal behavior, & parental behavior persists.

>>> Pups control duration of parental behavior
The pup's behavior seems to be what causes parental behavior to be
maintianed: if you rotate through young pups, so the mother has the
illusion that her pups are staying the same age, then the mother will
stay maternal.

>>> Suckling and temperature
Cut the sensory nerves around the mouth, or give litocane injections,
then mothers will become less maternal: she tells how old her pups are 
by sensory stimulation (nuzzling, suckling)

If you cool down pups, they stop moving, and the mother will stop
showing maternal behavior.  Also occurs if you anesthatize the pups.

Inject litocane into pups' mouth, so they stop suckling.  Then the
mother will stop showing maternal behavior.

>> What terminates parental behavior?
If you increase stress hormone, it decreases the amount of time spent
breeding.  Stress hormone increases mother's temperature.  This
increases her temp.  So she can't stay on pups as long, because she'll 
overheat.  When her brain temp is lower, she shows less maternal
behavior.  As pups get older, they start regulating their own
temperature \to they become heat sources.  So the mother spends less
time being maternal because mother's brain temp gets higher.

> Neural Mechanisms of Rat Parental Care
>> The preoptic area of the hypothalamus
>>> Lesions
If you lesion this area, then some parental behaviors are greatly
reduced (e.g., nest-building, retrieval).

>>> C-fos experiments
Use C-fos to map out activity in the brain.  A female that becomes
maternal in response to pups have more activity in pre-optic area than 
females that don't become maternal.

>>> estrogen receptors
Estrogen will trigger early onset of maternal behavior.

>> Sensory mechanisms for induction of parental behavior
Certain sensory mechanisms affect how long it takes animals to show
parental behaviors.  E.g., ablate olfactory bulb \to rats become
maternal faster.

> The Sheep: formation of selective bond with offspring
Sheep are herd animals, and young must follow their mother around from
pasture to pasture.  Sheep form selective bonds between young and
mother.  Sheep will not become maternal unless they go through the
physical act of giving birth.

>> Hormonal effects on maternal behavior
You can use estrogen/progesterone to mimick the effect.

>> Vaginocervical stimulation and maternal behavior
Stimulation in the vaginocervical area is critical to showing maternal
behavior.  If you give stimulation to non-pregnant sheep, you can
induce maternal behavior.  Also, if you give additional genital
stimulation to a pregnant sheep, she'll show more maternal behavior.

But they don't require as much vaginal stimulation if they've given
birth before.

>> Olfaction, amniotic fluid and the selective bond
It's the olfactory cues that are responsible for forming selective
bonds.  Reject sheep on the basis of their smell in close proximity.
This depends on the amniotic fluid.  Dry amniotic fluid off a newborn, 
then they won't be accepted as well.

>> Vaginocervical stimulation and olfactory processing
Vaginocervical stimulation increases the chance that she'll accept
alien lambs etc.

>> Natural mechanism of the selective bond
vaginocervical stimulation increases hormones that increase plasticity 
in the olfactory bulb.
coop 3rd floor

[11/01/99 09:11 AM] Communication II: Birdsong
!! Exam next Wed!!

> Mechanisms of Song Production and Perception
>> Song Production
>>> The Syrinx: 2 songs - one throat
The syrinx is a structure that only appears in song-birds.  It's
located at the junction of the broncheal tubes.  There are two MTM's,
which are stretched membranes.  As air passes over the MTMs, they
vibrate.  Thus, there are 2 potential sources of sound -- two songs
are produced and combined by songbirds.

Are the two MTMs ever used seperately?  Yes.  Airflow through each
broncheal passageway is controlled independantly, as is tightness to
MTM.  Evidence: cutting nerves innervating each side of the broncheal
tube gives different effects.

The rest of the vocal tract filters the sound.

>>> CNS mechanisms of song production
There is a system that seems fairly dedicated to song production.  HVC 
sits at the top of the system (high vocal center).  Basic motor
circuit is:

# HVC \to RA \to crainal nerve nucleus nXII in the hindbrain

How do we know this is important for song production?  Destroy HVC,
RA, or nXII, and the animal will stop singing.  He gets in the posture 
for singing, and open their mouth to sing, but don't sing.

Birds can still produce vocalizations without (some of) these
structures -- for example, calls.  In fact, for calls that are
partially learned, lesioning this pathway will selectively remove the
learned components.  Thus, this structure is probably involved in
learned songs.  

Also, tight correlations between neurons and song.  In HVC,
correlation between neurons and syllables.  In RA, correlation between
neurons and notes.

If the animal is singing, then production of Zenk (some random
protein) increases in HVC (and RA?)

If we stimulate HVC or RA electronically, you can elicit a song.
Songs are fairly close to their normal songs.  In contrast, in
midbrain stimulation elicits calls.

Structures that are important for learning songs:
# HVC \to X \to DLM \to LMAN \to RA

If you lesion this pathway, it doesn't impair production in adults.
But in juveniles who are still learning their song, it degrades
learning.  Seems that the song crystallizes too fast. 

>>> Structural features of song system

The HVC\to RA\to nXII structures seem to be specific to song-birds.  Other 
birds that learn sounds (parrots, etc.) use different structures.
Males have more primarily developed HVCs etc. (since primarily males
produce song, we would expect this).  The differences are formed
early.  Look across species to see if sexual dimorphism in volumes of
HVC etc tracks dimorphism in song production.  Indeed, in species
where both sexes sing, the HVCs etc are similar size, and when only
males sing, HVCs are larger in males.

There is a correlation between how many song syllables you produce and
how big your production pathway (HVC, RA, etc) is.  (note correlation
-- we don't know which way it goes yet).

There is a correlation between ?? and how big your learning pathway
(X, DLM, etc) is.  (note correlation -- we don't know which way it
goes yet).

There is a correlation between how many song types of song you produce
and how big your production pathway (HVC, RA, etc) is.  (note correlation
-- we don't know which way it goes yet).

Examine which way the causation goes for these correlatoins: is the
size of the structures the consequence of singing?

We can experimentally produce birds where some are only taught 5
songs, others are taught about 45.  Then compare voulmes -- they don't 
differ.  This suggests that it's not that sining more makes your HVC
bigger, but rather that bigger HVCs allow you to learn more.

Sizes of HVCs etc change over time, and over seasons.  You can
manipulate this in the lab by manipulating photoperiod -- expose
animals to short days, and their HVC & RA will get smaller.  HVC kills 
neurons when it doesn't need them, and then makes new ones when it
needs them (neurogenesis).

HVC size changes also correlate well with testosterone levels.  HVC
etc are affected by testosterone etc.  Singing is experinmentally
dependant on testosterone.  Castration decreases singing, then
testosterone injection restores it.

>> Song Perception
>>> Perception of Song and Field L
How do birds percieve songs?  Field L is a part of the brain that
projects near HVC.  Field L seems to be involved in the perception of
song.  Field L seems to be interrested in any song.

>>> Perception in the Song Production System?
Structures that produce vocalizations are also sensitive to their own
vocalizations.  HVC seems to be interrested only in the animal's own
song -- feedback for learning.

Females use HVC to recognize songs.  If you lesion HVC in females,
then they'll respond to other species' songs.

>>> Song perception and gene expression
NCM is innervated by Field L.  Zenk production is increased in NCM by
conspecific song.  MCM projects to HVC.  NCM may be involved in
recognizing conspecific song.

> Getting The Message Across
>> Attenuation
Sound gets weaker the farther you go.  High frequencies attenuate more
-- atmospheric turbulance, etc.  But low frequencies interact in bad
ways with the ground.  Thus, birds sould try to produce low-freq songs
and get high off the ground to sing.  But birds have small vocal
tracts, so it's hard to make low frequency songs.  Also, low freq
songs are more expensive.

>> Degradation
Sound gets degraded (the environment acs as a filter).  It's mostly
high-frequency sounds that get degraded.  Reverberation causes signals 
to blur.

>> Adaptations to noisy environments
>>> Singing position
Birds try to sing from high positions.  Males compete for good
positions from which to sing.

>>> Correlation with habitat
In open country, birds tend to produce rapid frequency modulations.
In dense forests, birds use low frequencies.  Some songs have
different types of songs that are suited to different environments.

>>> Using degradation of song as a cue
Animals use the degredation of a song to judge whether the producer is 
nearby.  Males will respond more agressively to undegraded song than
to degraded song. (??)  

>>> Predation pressure
Calls can be affected by predation pressure.  For example, alarm calls 
are similar in most species.  The features of these common calls may
make it difficult for the predator to find the prey that's making the
alarm call.

> What are the Functions of Song?
>> Territorial Devence
>>> Effects of muting male songbirds
If you mute a male, it gets challenged more, and it usually loses his
territory.  Also, muted males take longer to get territories.  Muted
males need to engage in much more direct agressive behavior, since
they don't have a way of anouncing their status.

>>> Speaker exeriments: Removing the male
Replace the male with a speaker that makes the male's song.  Then
territory shrinks much slower than if you just remove the male.

One idea about why a bird sings multiple songs: it's pretending to be
multiple males ("Beau Geste" hypothesis -- guy left alone at a castle
propped up his dead comrades against the walls to make it look like
the castle was manned & so he could scare off the intruders).  Areas
where speakers play bigger repetoires get filled up slower.

>> Recognition
>>> Species
Recognize whether birds are conspecific or heterospecific.  Important
for competition and for finding mates.  Sometimes invariant features
are the important features in species recognition.

>>> Parent-offspring recognition
Birds show preference for their parents' calls and their offsprings'
calls, especially in social birds.

>>> Mate recognition
Monogamous birds show preference for their mates' calls.

>>> Neighbors and Strangers
Males have territorial neighbors.  Males distinguish neighbors from
strangers.  You've already settled on boundries with neighbors, but
strangers generally don't have territories yet, so respond more
strongly to strangers' calls.  Birds keep track of where neighbors are 
-- if you present a neighbor's call from a new area, then respond
agressively (that neighbor is expanding?).  Birds even remember their
neighbors from previous years.

>> Attraction of Females
Muted males don't attract females.  Speakers producing male calls
attract females.  Females aren't fooled by mimick birds (even though
males are).  Hearing songs to femalse increases females' reproductive
behaviors.  Fast singers selected over slow ones.  Ability to feed
young increases with increasing song rate.  How quickly you pair
increases with repetoire size.  

>>> Muted Males
>>> Attraction
>>> Reproductive Activity
>>> Female Choice: Output and Repertoires

[11/03/99 10:00 AM] Communication II: Chemosignals in Vertebrates
> Production of Chemical Signals
Many animals learn about other animals through chemical signals.  This 
method of communication is very phylogenetically old.  It's present in 
most animals.  True of almost all nocturnal mammals (and most mammals
are nocturnal).

Olfaction/chemical communication harder to study because we can't
detect it as directly.

Some lemurs that rub their tail with their arms -- puts their scent on 
their tail.  Then they wave their tail at competitors.

Chemical signals are different than most signals because they persist
in the environment -- you can leave a signal and go away.  Also, they
can provide information about how long that scent can be there.
c.f. visual signals.

>> Sources of Odor
>>> Glands
Many odors are produced by specialized scent glands.

  - Location: glands can be found just about anywhere on an animal's
    skin.  Heads/body/arms/legs/tail/etc.  Can be big or small.  Often 
    found in andogenital area.  Also can be located inside the
    animal.
  - Structure: usually modified hair folicles (sebatious).  Produce
    sebum, which is a type of oil.
  - Sexual Dimorphism: many scent glands are sexually dimorphic in
    size.  Sometimes glands are bigger or only present in males,
    sometimes in females.
  - Regulation by Gonadal steroids and seasons: The glands are
    regulated by hormones (males' usu. regulated by testosterone,
    females' usu. regulated by ovarian hormones).  There are also
    seasonal differences.  Often become bigger when animals are
    competing for territories.  
  - Social Status: Glands are larger in dominant glands.
  - Bacteria: bacteria proliferate in warm wet places.  Often it's the 
    bacteria working on the products of the odor glands that produces
    the actual scent.  Often it's harder for animals to discriminate
    odors that come from animals w/o bacteria -- maybe different
    bacteria produce different scents?

>>> Excreted "By-Products": Urine, feces, and saliva
Urine is a very common scent-marking device.  It's cheap to make,
since animals urinate anyway.  Ungulates (antelopes etc) often use
feces.  Urine can allow animals to learn about the reproductive state
of the urinating animal -- animals with more progestine in blood
stream (i.e., ready to mate) will often have more progestine in their
urine.  Same with other hormones.  Males can pick up on these scents
in the urine to determine how receptive females will be.  In pigs etc, 
saliva is used to scent mark.  Saliva contains sex hormones that
increase female receptivity.

>>> Different glands -- same/different information
Many species have multiple sources of odor.  Often the different
sources of odor contain different information and serve different
purposes.  Can tell whether an odor is used by testing whether animals 
can distinguish odor from different animals and whether they respond
differently to male & female odors.

>> The odors themselves
>>> Complex Mixtures
In vertebrates, the odors are huge mixtures of molecules -- there is
usually no single phermone molecule that accounts for the effect.  Use 
gas chromatogram to seperate out different molecules -- allows us to
compare odors produced by different animals, and differences in male
vs. female.  Often there are compounds that only one sex has.

>>> Changes with sex and reproductive status
The scents put out by females changes depending on how receptive they
are.  Often females have new chemicals appear in their odors when they 
are receptive.  Also, dominant animals often have different odor
compositions than subordinate animals.  We can use gas chromatograph
to figure out which odors might designate dominance or receptivity.

>>> Changes with social status and testosterone levels
  - Tree Shrews: castate males, and they no longer have a compound in
    their scent that males normally have.  Give them testosterone and
    they'll start having that compound.  Experiment: first, males
    scent mark more if they smell normal scent than if they smell
    castate scent.  But if we put the compound there, then they will
    scent mark more.
  - House Mice: castrate male urine doesn't elicit the responses that
    normal male urine does (agression from males, female attraction,
    induce estrus, etc.).  Examine urine of castates and normals.
    There are 2 chemicals missing from castrate urine.  If we add
    those chemicals, then agression increases.  Also, if we add those
    chemicals, females are attracted & estrus is induced.

>>> Peptides as social odors in newts
Males produce a peptide odor attractive to females.  Female has to be
in reproductive condition to be responsive.  Odor production dependant 
on testosterone.  Detected via olfaction.  Peptides can function to
distinguish species (a fairly similar peptide is attractive to females 
of a closely related species).

>> Deposition of Odors
>>> Injection, the direct approach
Directly inject the odor into the females' bloodstream.

>>> Direct release into the environment
Release odors into the environment in an uncontrolled fashion (fish). 

>>> Secent Marking
Put scents in specific places.  scent mark places or yourself or other
animals.
  - Motor Structure: basic mechanism differs among animals, and is
    strongly affected by where the odor glands are.
  - Sexual Dimorphism: often males mark more than females.  sometimes
    they mark in different ways.
  - Regulation by gonadal steroids and seasons: scent marking is often 
    seasonal (manipulate photoperiod).  Marking is regulated by
    hormones (usu. sex hormones).
  - Odors and social status: scent marking is regulated by odors --
    try to cover up other animals' odors with your own.  scent marking 
    is also regulated by dominance -- dominant animals tend to scent
    mark more than subordinates.
  - Neural mechanisms of scent marking: in mongolian gerbils, there
    are specific places in the brain where you can inject hormones to
    cause scent marking.  Lesioning these areas reduces scent
    marking.  In hamsters, vasopressin injected in a particular area
    will elicit a lot of scent marking.  A vasopressin-blocker will
    reduce scent marking.  Dominant animals have more vasopressin
    fibers in this area of the brain.

> Getting the Message Across
>>> Substrate
Where you put the odor affects how long odors will last.  Sebum may
cause odors to evaporate more slowly.

>>> Conspicuousness: scratching and UV
Animals can make their marks more conspicuous -- pair visual cues like 
scratched trees with the mark etc.  Sometimes odors might be colored
(e.g., iguanas' odors are colored in the UV spectrum).  

>>> Correlation with environmental context
Sex attractant components of scent tend to evaporate quickly.
Territory marking components evaporate less quickly.

Where humidity is higher, animals produce heavier chemicals (to slow
down evaporation, since humidity increases evaporation rate).  

> Perception and Function
>> Information in the Scent
>>> Species
Animals prefer the odors of their own species.

>>> Sex and Reproductive Condition
Males are attracted to female odors, & females are attracted to male
odors.  Also, odors contain info about reproductive condition.  Odors
of reproductive animals are much more attractive.

>>> Social Status
Females prefer the odors of dominant males.  But if a female isn't in
reproductive condition, this preference disappears.

>>> Mate recognition
Prefer to interact with familiar mates rather than unfamiliar mates.
Prefer to investigate your mate's odor.

>>> Individual discrimination and recognition
Animals can learn to discriminate individuals' odors (via skinner box) 
very easily.  Animals recognize which odors from different glands go
to the same animal.

>> Mate Choice
Odors are usually involved in female mate choice.  Females tend to
prefer dominant males with high testosterone.

>>> Overmarking and resource holding
Males mark over other males' cues.  The better a male does this, the
more a female will prefer them.  If there is lots of M1's scent, but
it's marked over by a small amount of M2's scent, the female will
still prefer M2.

>>> Parasitism
Females avoid the urine of infected males. 

>>> Diet and choice
Females prefer the scents of males that eat more protein.

>>> Costs of scent marking/odor cues
Urine is detected by predators.  

>> Territorial Interactions: the scent-matching hypothesis
Dominant animals will expand their territory to subordinate animals'
territories.  Most marking happens at territory borders.  If males
detect that the odors of an animal don't match the odor of the
territory they're on, then they'll be more agressive.

>> Neural Mechanisms of Social Odor Recognition
>>> The Olfactory and Vomeronasal chemosensory systems
>>> Petrulis' Work

[11/08/99 09:05 AM] Socail Living and Altruism I: Kin
> Why live in a group?

  - Predation pressure
  - Food is difficult to find

In some guppies, sociality increases as predation pressure increases.

>> Predation Pressure
>>> Vigilance
The more animals in your group, the more animals can look out for
predators.  

Goshawks eat pigeons.  In an experiment, the more pigeons, the less
effective the goshawk was at getting any prey.  Also, the earlier the
pigeon sees the hawk, the more likely it will survive.  And as you
increase group size, the distance at which you see the hawk
increases.  The time spent looking for predators also decreases as
group size increases (presumably giving more time for foraging).
Also, scanning of a single pigeon is fairly predictable -- they just
scan every 10 seconds or so; but much harder to predict scanning times 
of a group of animals.

>>> Alarm Calls
Often, groups of animals have "sentinals" whose job is to look out for
predators.

Certain type of squirrel has 3 alarm calls -- snake, leopard, bird.
Other monkeys respond differently to each call.

When an arial predator is spotted, everyone makes call and everyone
runs.  In this context, the animal that gives the alarm call decreases 
their chances of getting caught.  This  type of call is not sensitive
to who's around you.  Explanation: make the call to create confusion
and confuse the predator.

>>>  Dilution Effect
The more animals that are around, the less likely it is that you'll be 
attacked -- there are many other targets for predators to choose
from.  

Scanning as group size increases.  If group size gets big enough (>60
or so), the amount of total scanning becomes less than an individual
would scan (everyone assumes that everyone else is scanning).  But
even though there's less scanning, the dilution effect makes it
beneficial -- it's unlikely that you'll be killed even if a predator
does come.

Dilution effect is seen in groups of monarch butterflies, too.

Horses in a swampy area are assulted by blood-sucking insects.  The
larger the group the horse is in, the fewer insects are sucking their
blood -- if there are enough horses around, then there are fewer
insects/horse.   

>>> "Selfish Herd" effect: Bluegill Sunfish
Groups band together to try to avoid predation.  It's the animals on
the outside of the group that get eaten, so you should try to stay in
the middle of the group.

In sunfish, predators tend to attack single nests instead of
colonies.  Solitary nests can be attacked from any side, but
peripheral colony nests can only be attacked from one side. 

>>> Group Defence: Cultural Transmission of Mobbing
Groups can get together for defence.  When a predator comes, they will 
all attack & drive off predators.

Mobbing can also be culturally transmitted: let an observer bird see a 
bird attack another bird.  Then test the observer's response to the
bird they saw attacked.  They will react much more agressively.  This
could allow transmission of information (which animals should be
mobbed) across generations.

>> Obtaining Food
>>> Finding Good Food Sites: "Information center" Hypothesis
In weaverbirds, there are very sparce sites, each of which has lots of 
food.  A colony is an "information center:" you can follow fellow
animals to food sources & eat there.  There is experimental evidence
that weaverbirds can tell what other weaverbirds have been eating
well, and will follow those birds.

Cliff swallows live in colonies.  Individuals seem to follow
successful foragers.

>>> Foraging Together
Often foraging together is better than foraging alone.  Jacks hunt
anchovies in groups.  The group of jacks tries to break up the
anchovie group so they can attack individual anchovies.  Jacks can get
food faster and can get more food when they're in a group.  The first
animal to attack usually gets the most food.  But the first animal to
attack is the first animal to see the anchovies, so it evens out in
the long run..

Harris' Hawks attack rabbits in groups (rabbit is 2-3 times heavier
than the hawk).  The groups tend to be of related animals.  A variety
of group tactics.  E.g., take turns attacking, or having one chase the
rabbit into the others.  Larger group sizes signifigantly increase the
number of rabbits killed.

Gulls feed on fish.  Gulls in flocks of six each get more food than a
single gull would.  If there's one gull, the fish all run from it.
But if there are 6 gulls to run from, then one gull can catch a fish
as it's running away from another gull.

>>> Foraging/Predation Tradeoffs: House Sparrow Chirrups
?

>> Costs of Sociality
>>> Competition for food resources: starvation in fieldfares
If you're living in a group, then everyone in the group is competing
for the same food sources.  Children can starve.  

>>> Effort parasitized by other group members: male lions
Others in the group might not share.  In lions, females tend to do the 
killing; but once the prey is dead, the males tend to displace the
females. 

>>> Parasites: cliff swallows
The closer-packed a population, the more disease and parasites can
spread. 

> Does "Altruism" Exist in the Animal World - Kinship
>> Indirect selection and inclusive fitness: Hamilton's Rule
Often, if you're living socially, you're living with relatives.

Sometimes, you forgo reproducing to help your relatives reproduce.  We 
can explain this with "indirect selection."  Direct selection is
normal selection; i.e., you're more fit if you produce more viable
offspring.  Indirect selection is selection because of shared genes
with relatives.  You share 25% of your genes with a neice, so helping
your sister have 2 more children will increase your genome
representation in the population as much as having 1 child will.
Inclusive fitness = how much you increase your genome's representation 
in the population.

Hamilton's rule: kin selection iff rB-C > 0b

Where r is the degree of relatedness to non-kin that you help.  B is
the benefit (the additional number of offspring you caused).  C is the 
cost (the number of children you didn't have yourself because you were 
helping kin).  

Animals take how related they are into account in interacting.
Animals are less agressive to their relatives, more likely to
cooperate, less likely to trespass on relatives' territory.

>> Eusocial insects (see Alcock text)

>> Alarm Calling and Social Interactions in Belding's ground squirrels
Belding's squirrels produce calls in response to ground calls.  These
ground calls have a cost: a calling squirrel increases their chances
of getting eaten.  Primarily females give calls (females are the ones
who interact with their own relatives; males disperse from their
homes).  

>> Cannibalism and kin: Arizona Tiger Salamanders
Larvae can be either normal animals or cannibals.  The cannibals
prefer to eat less closely related animals over more closely related
animals.  

>> Kinship and Schooling: Toad Tadpoles
Toad tadpoles aggregate with their relatives -- help their relatives.

>> Kinship and outbreeding
>>> Toads
>>> Golden Hamsters
Female seem to be less sexually attracted to closely related animals. 

>> Kinship and Breeding: Male Lions
A coalition or males holds a pride.  In small coalitions, males tend
to be unrelated: they share the mating equally.  In large coalitions,
males tend to be brothers/kin: most males can't reproduce.  These
males get fitness benefit indirectly via their brothers.

[11/15/99 09:04 AM] Social Living and Altruism II: Kin & Non-Kin
> Cooperative Breeding and Kinship

Cooperative breeding is fairly common (hundreds of mammal species &
bird species).  Help your kin breed.  Usually decreases your ability
to breed.

>> Florida Scrub Jay

Live in very harsh environment.  Resources are patchy etc.  About half
of the nests have helpers.  Helpers are often the youngest jays (1-2
yrs old).

>>> Breeders benefit from helpers

Helpers assist in countering predation and in getting food.  64% of
helpers help both of their parents.  Most of the rest help one of
their parents.  A couple (~1%) help non-parents.

Helpers do help.  Usually a nest w/o helpers generates 1
offspring/year, and a nest w/ helpers generates 2 offspring/year.

Helpers bring in food, but total food brought to chick doesn't
increase -- the parents bring less.  This alleviates strain on
parents, and helps the parents survive.  Also, helpers do alarm calls
to protect vs. predators.

>>> To breed or to help?

On average, each helper produces 0.33 offspring/year.  In contrast, a
new breeding pair can produce 1.4 offspring/year.  So why not breed?
One reason is that leaving the nest early increases bird mortality.
Chance of successfully finding a nesting site is low.  Also, helping
might make scrub-jays better parents.

>>> Why stay and help: saturated habitat

Since the environment is so patchy and scarce, the jays have saturated 
all nest sites.  Thus, it's very hard to start a new nest.  If we give 
the jays lots of nesting sites, then there won't be any helpers.

>>> Helpers expand and inherit territories

Males can inheret their parents' breeding territory.  This is the
normal way of acquiring a territory.  Nests with larger groups of
helpers will increase the size of that territory.  If they increase
the size of the territory enough, then they can bud off.  So by
helping, you are increasing the size of the territory you might get,
and you're carving out your own territory.

As long as you're staying at home (to get territory), you might as
well help your siblings.

>> Black-Backed Jackals
Form long-term pair bonds with helpers.  Helpers are usually the
parents' offspring from previous years.  Helpers feed, groom, defend
vs. predators, and protect vs. other breeding groups.  Presence of
helpers increases number of pups surviving.  In general, each helper
adds about 1.5 pups/yr.  

>> White-Fronted Bee-Eaters
>>> Clans, Relatives, and Helpers
Live in clans -- groups of extended relatives.  Several pair-bonds can 
be formed within a single clan.  Each helper will help exactly one
breeding pair.  Helpers are more likely to help close kin.  Give them
the choice between helping a more or less closely related pair, & they 
will almost always help more closely related pair.

>>> In-laws
If you join the clan from the outside as a spouse of a member, you
very rarely help.  This makes sense, since there's no one around who's 
related to you.

>>> Disruption of Breeding Attempts
Often fathers will disrupt the breeding attempts of their sons, to
make them into helpers.  They can do this by disrupting courtship,
blocking nest sites, etc.  The younger the animal, the more likely it
is to be recruited.  

>> Do helpers really help? Experimental Evidence

Experimentally remove helpers from the nest.  Then the number of
offspring goes down.

>> Do Breeding Constraints Really Exist?
>>> Superb Fairy-Wren: A female!  My kingdom for a female!

Cooperative breeder, breeding pair + male helpers.  Remove the
breeding males from nearby territories.  Then most of the helpers will 
disperse.  But if you remove both male & female (so there's an empty
territory, but it has no females), then they won't dispurse.  

>>> Seychelles Warbler: Discriminating tests

Very small species (~300 members).  Cooperative breeders appear once
they start saturating their territories.  There are some good
territories and some bad territories.  If you remove breeding pair
from a bad territory, then helpers from good territories won't grab
them.  But helpers from a bad territory will.  Part of this comes from 
the fact that helpers on good territories will live much longer, so
they can afford to wait for good breeding sites.

> Mechanisms of Kin Recognition
>> Spatial Proximity
Anyone in your nest is probably either your sibling or your
offspring.  Parents will often reject offspring if they are outside
the nest, and will accept unrelated animals if they're in the nest.

>> Association in the Nest
Learn who your kin are because you grow up with them.  Remember who
you grew up with.  Put unrelated animals in nest.  Then everyone that
grows up with them will treat them as kin.  As soon as animals leave
the burrow, associating with someone doesn't make you treat them as
kin.  

>> Phenotype matching
>>> Belding's ground squirrels
You can recognize close relatives even if you didn't grow up with
them.  1 theory: animals share common uterine environment.  But in B's 
ground squirrels, they can tell full-sisters apart from half-sisters.
Another theory: match your own scent to other animals'.  Treat animals 
with similar scents as kin.

>>> Golden hamsters
If you're raised with non-relative A, then how will you treat A's
siblings?  If you treat A's siblings more like kin than random people, 
then you must be doing some type of phoenotype matching, since you've
never seen A's siblings before.  You do treat A's siblings more like
kin.  

> "Altruism" without relatedness -- Mutualism
Sometimes it pays for animals to cooperate, since they both gain
fitness from it.  

>> Male Lion coalitions
In small coalitions, the males are unrelated.  But they both get more
reproductive success then they would alone.  

>> The Dwarf Mongoose and the Hornbill
Mutualism exists between species.  Dwarf mongoose and hornbill forage
together.  Both eat insects.  Mongoose will go into anthills to eat,
and hornbills will eat the insects that scatter.  So hornbills get
more food.  Mongooses gain predator protection -- hornbills make alarm 
calls.  Mongooses do less alarm calling when hornbills are around.

>> Cooperative breeding and non-kin: Pied Kingfisher
You can sometimes find non-related helpers.  Primary helpers
(offspring from previous years) and secondary helpers (unrelated).
Primary helpers will chase off secondary helpers unless they're having 
trouble feeding their young.  Why do secondary helpers help?  Well,
first, they help less than primary helpers.  Second, if a female
looses their mate, it's easier for you to pair up with her.  Also,
secondary helpers survive longer than birds that just hang out.

> "Altruism" without relatedness -- Reciprocity

Very rare form of cooperation.  If cost to you is much less than
benefit to other guy, and other guy is likely to reciprecate, then
reciprocity can happen.  The problem is that this system is open to
cheating: you can get helped by someone else and not recipricate.

In order for reciprocity to be evalutionarily stable:
  1. Animals can't get away with cheating
  2. Probability of encountering the animal you're being reciprical
     with again must be high.

>> Impala grooming
Animals groom each other.  The amount of grooming you do to them is
almost identical to the amount of grooming that they do to you.
Grooming helps control parasites -- animals can't sufficiently protect 
themselves against it.

>> Coalition Grooming in Vervet Monkeys

>> Vampire Bats and Blood Regurgitation
Vampire bats suck blood from large animals.  Individuals often fail to 
get any food in a given night.  When they do, they will beg for food
from their neighbors, and their neighbors will give them food.

>>> Must recognize cheats
Bats that receive blood tend to be more likely to give blood in the
future.  

>>> Long-term interactions
Bats interact with each other for years.

>>> Benefits of receiving aid must outweigh costs of donating
1 or 2 days without feeding can cause mortality.  But if you have a
lot of blood, and donate a little, it doesn't cost you much.  But it
helps them a lot.  

[11/17/99 09:04 AM] Sacpe and Navigation I: The Why of Space
(lecture by Seth Ramus)

> Deciding Where to Live
What factors might animals take into account to choose their habitats?
  - Resources
     - Food
     - Water
     - Shelter
  - Mate Availability
  - Low Predation

Animals are optimizers: they choose where to live based on factors
that maximize their potential reproductive success.

>>  Ideal Free Distribution (Fretwell)
Dictates that animals will choose the best environment for their
fitness.  But this model makes some assumptions:
  - Animals are "free" to go where they will do best -- no exclusion. 
  - the animals have complete information about the availability of
    resources.  
  - Every animal ends up with about the same fitness

Asserts that each animal will move into the environment that maximizes 
its fitness, taking into account how the animals currently in an
environment effect how good it is.  E.g., if E1 has 10 food units, and 
E2 has 4 food units, the first 2 animals to arrive will move into E1.
The next animal will move into E2.  The next will move into E1.. Etc.
An example of ideal free distribution is people choosing the shortest
line in a grocery store.  Milinksi did an experiment to test this with 
fish.

>> Despotic Distribution
Dictates that the first ("best") animal to claim a resource patch
defends it to the exclusion of all others.  The next animal takes the
next best resource and defends it, etc., until all resources are used
up.  The remaining animals are "floaters."

Krebs studies great tits, which show despotic distribution.

>> Ideal Free Distribution with Unequal Competitors
This is closer to reality.  Some animals are better competitors than
others, and they chose the best places to live.

One way to account for differences in how well animals compete is to
rate each animal with "competitive units" -- better competitors will
have more competitive units.  Then use competitive units to do ideal
free distribution, assuming that each animal will take up an amount of
resource proportional to the number of competitive units they have.
For example, consider a super market.  Just counting up the number of
people in each line isn't a good indicator of how long the line will
take -- you need to take into account how much resources each person
will take (i.e., how much they're buying).  People with lots of items
have more competitive units than people with just a few.  In the fish
example, a fish that eats twice as much as another fish will have
twice the number of competitive units.

Unfortunately, this makes the same basic predictions as ideal free
distributions -- since the big and little animals are usually
distributed randomly anyway, it's most likely that we'll just get the
same result as ideal free distribution.  The chance that we'll get
something signifigantly different from ideal free distribution (i.e.,
have lots of big fish in one region and lots of little ones in
another) is low.

>> Ideal Free Distribution with Resource Defence
In reality, most examples include elements of both resource defence
and ideal free distribution, resulting in unequal distribution of
resources.  

The example of aphids.  Aphids try to find places to reproduce and to
induce a gall.  The best places are (1) the base of leaves, and (2)
big leaves.  The first aphids go for the largest unoccupied leaves.
When all leaves are full, aphids will take places further out on
already-inhabited leaves.  Look at how well 1st stem mothers, 2nd stem 
mothers, and 3rd stem mothers, do on average.  They all do about the
same.  This makes sense, because if 1st stem mothers did better on
average, we'd expect some 2nd stem mothers to move to unoccupied small 
leaves\ldots  But also, aphids will fight for the position at the base of
the leaves.

>> Economic Defendibility (Brown)
When the benefits of defending a territory outweigh the costs, then
you should defend the territory.

>>> Golden-Winged Sunbird (Gill & Wolf)
Sunbirds defend patches of flowers, and exclude other sunbirds from
the patch.  Can we show that defending the patch benefits the sunbirds 
more than it hurts them?  If you defend a patch, the amount of nectar
in each flower increases, since no one else is eating it.  Observe the 
animals in the wild, and figure out how much time they spend
defending, eating, foraging, etc.  Then figure out how expensive each
activity is, in terms of energy.
   - Foraging: 1000 cal/h
   - Resting: 400 cal/h
   - Defense: 3000 cal/h

In undefended territory, foraging takes about 4 hours.  In defended
territory, foraging takes about 2.7 hours.  This saves the bird
(1000 cal/h)(1.3 h) - (400 cal/h)(1.3h) (since it's resting instead of 
foraging).  So the bird saves 780 calories.  But how much time needs
to be spent for territorial defence?  The average sunbird spends about 
(.28 hours) defending, so defence costs 728 calories.  Since 780>728,
the bird is gaining more than it's losing, so it makes sense to defend 
the region.

> Deciding When to Move
Territories can have costs.  Defending a territory can decrease your
time available to mate, to find food, etc.  Also, you can get hurt
defending a territory.  So sometimes it is / becomes a bad idea to
defend a territory.  And sometimes it makes sense to move on and try
to find a new territory.  Three examples of times when animals choose
to dispurse or migrate:
  - Natal Dispersal.  When animals are first born, they choose to
    dispurse.  This may have a lot to do with mates.  If you don't
    dispurse, you'll probably mate with your siblings (inbreeding).
  - If a territory turns out not to be good, then dispurse.  E.g.,
    redwing blackbirds will usually set up in the same territories
    after migration.  But if a blackbird fails to mate well one year,
    they will tend to choose a new territory the next year.
  - Migration, to maintain resources due to expected changes in the
    environment (e.g., seasonal changes).  Cost of changing
    environment must be lower than benefit of moving.  Some animals
    conditionally migrate.

> Identifying Places in an Environment
An animal must be able to identify specific places in their
environment -- use local cues and other information.  You have to know 
where the food, shelter, etc., of your territory are.  Also, you have
to know where the borders are, etc.  It's important for defence and
for locating resources.

[11/15/99 10:32 AM] Space and Navigation II: The HOW of space.
(lecture by Seth Ramus)

The problem of spacial perception is extremely compelex, and animals
perceive space in every sensory modality possible.  By contrast,
spacial representation is more likely to be independant of sensory
modality.  It would be inefficient to have one visual map and a
separate olfactory map.

All navigation basically requires:
  - a map (a representation of the relationship between places stored
    in an animals brain)
  - a compass (external cues as to the direction or position of the
    animal in external space)

> Large Scale Navigation
navigation over large spaces (e.g., migration) doesn't require
learning.  Doesn't rely on landmarks, but usually on celestial and
magnetic compasses.

Sun timer: watch where the sun is at different times of day, to figure 
out what direction you're going.

Magnetic compass.

Use the polarization of sunlight to figure out directions, since
sunlight polarization changes over the day.  Especially true of
insects.  E.g., some ants will wander around looking for food, and
then will go streight home.  How does it tell where home is?  It
figures it out from its path so far.  If we pick up an ant and move it 
over by 10 feet, then it will search for its home at a 10-foot
displacement from where its real home is.  If we use a polarizing
filter, we can fool the ants.  

Bees also use the sun as a compass.  See alcock

Ants do path integration.  Path integration is one solution to
navigation without landmarks: simply keep track of your "coordinates"
relative to your home, so that when you want to go home, you know what 
direction and how far to go.

Large scale navigation is succeptible to errors: if you make a
miscalculation, or something unexpected happens (you get picked up),
you can get very confused.  Usually use landmarks to refine
navigation. 

> Small Scale Navigation (Local/Landmark)

Use landmark-based navigation for smaller scale navigation.  One
hypothesis is that we make a cognitive map. This is a representation
of the world around you that has landmarks and their relation to each
other (direction, distance, etc).  This means: (1) that you always
know where you are; and (2) if someone moves you, you'll still know
where you are..  

Landmark system must be plastic, since landmarks change over time.

Animals can take advantage of landmarks: place nests near large
landmarks.   An experiment to see if animals use landmarks: place
distinctive landmarks around a wasps nest, let them get used to it,
and then offset all of the landmarks by a fixed distance.  Then the
wasp will look for the nest where it should be relative to the
landmarks. 

An experiment for cognitive maps: let a rat learn a maze.  Then teach
it to go to a goal.  Then see if, when we block off passages, it knows 
enough about the maze to take the most efficient path.

Bees maintain a cognitive map.  Do bees have an understanding of the
space around them?  Expose a bee to a waggle dance, then put them
somewhere random, and they'll still go to the food.. So they're not
just going in the direction they're told.  Another experiment:
artificially make bees signal that there's food in the middle of a
lake..  Stick bees on a feeder in the middle of the lake, and see if
they can recruit other bees to come eat.  They can't: other bees don't 
believe them.

> Representations of Space

An oversimplification:
# Receptor \to processor1 \to processor2 \to \ldots \to high-order rep.

The later processors combine modalities.  2 separate pathways for
vision: the what pathway and the where pathway.

Cognitive map probably resides in the hippocampus.  Hippocampus
involved in memory.  It's an association cortex that stores
relationships between landmarks.  Lesions of hippocampus cause
navigation deficits.  

2 interrsting cells that have been found in the hippocampus: head
direction cells fire when the animal faces a certain direction
(direction depends on environment).  place cells fire when the animal
is in a specific part of space.

Tape: place cells of a rat running around a maze.

[11/29/99 09:00 AM] Agression and Territoriality

> Territoriality
>> Territory vs. Range
A territory is a fixed patch of resources that an animal defends.
Ranges aren't defended.

Animals must "decide" how large of a territory to have, whether or not 
to even have a territory, etc.

>> Economics of Territory Defence
It takes energy to maintain & defend a territory.  Animals use the
costs/benefits to "decide" what to do about territoriality.

>>> Costs of testosterone and aggression
  - spiny lizards: males defend territories.  try experimentally
    increasing the testosterone levels of some lizards.  those lizards 
    have a substantial increase in mortality.  (thus, there's a direct 
    cost associated with testosterone).  testosterone-increased
    lizards are more agressive, and more active.  this gives less time 
    to forage for food: these animals tend to starve.  also, they run
    about more, so are more conspicuous to predators.  So there is a 
    TRADE-OFF with testosterone: agression/activity for chance to
    live.  (Experiments show that it's probably starving that kills
    them, not predation.)
  - red grouse: males form large territories containing several female 
    nests.  Implant testosterone in some males, and their territories
    grow.  Reproductive success is very high (almost tripple), but
    they all died over the winter, because they had spent less time
    feeding.  
  - dark-eyed juncos: monogomous species.  Try implanting
    testosterone.  Then they sing a lot more.  But they spend less
    time feeding their young.

>>> Defence and resources
Sometimes it makes sense to decide whether to have a territory
depending on the value of the territory and the cost of defending it.
So if costs are high, or territory value is low, don't defend it.  If
territory is too big to be useful, defend a smaller territory.  Larger 
territories are harder to defend, expose you to more predators.

  - Maximizing gain (Rufous hummingbird): defend an optimal territory
    size.  Animal figures out over the course of about a week how
    large of a territory to defend to maximize net gain from
    territory.  
  - To defend or not to defend (Pied wigtails): defend patches of land 
    that border rivers, to eat bugs that wash up.  After you eat bugs
    from the river bank, it takes about 40 min for the bugs to come
    back.  Intruders can come eat food before you can get to it.  So
    it's good to defend a territory.  Occasionally, a wigtail will
    allow a satellite male on the territory.  They eat out of sync, so 
    they each get 1/2 the food.  But satellite male helps to defend
    the territory.  Especially like to have satellite when there's
    lots of food, or when there's lots of intruders.

> Mechanisms of Agression and Dominance: Hormones
Agression often correlates very well with hormones (e.g.,
testosterone).  

>> Seasonal variation in testosterone secretion and agression
Seasonal variation in hormones correlate well with agression.

>>> Red Deer of Rhum
Breeds during a short period of time.  Levels of testosterone are
maximum around breeding time.  This is also the time when the deer are 
most agressive.  Castrated males have smaller antlers, etc.  Also,
castrating an animal drops their dominance status.  Testosterone shots 
increse an animal's dominance.  (Part of this comes from antlers --
remove a deer's antlers, and its dominance drops)

>>> Badges of status: Harris's sparrow
Sets up territories in breeding area and dominance territory in
wintering flocks.  You can tell who's dominant by how many black
feathers are on the chest: the more black feathers, the more
dominant.  Testosterone correlates well with black-feather-area during 
the summer.  But not during the winter.  Try dying a dominant white.
Then other animals will think he's subordinate, but he'll keep
responding agressively.  Or try dyiing a subordinate black.  Then they 
can't respond to challenges.

>>> Not always correlated: Dusky wood rat
Defend fairly stable breeding areas.  Castrated animals aren't less
agressive.  Even though there's a correlation between agression and
testosterone, it doesn't seem to be causal.

>> Puberty, testosterone and agression
Testosterone increases and agression appears at the same time:
puberty.  Inject pre-puberty animals with testosterone, and they'll
become agressive.

>>> Why increase testosterone at puberty?
Animals disperse at puberty.  You need to go find a territory, etc.,
and fight more.  Giving testosterone early can cause early dispersal.

>>> Dispersal in Belding's ground squirrels
Dispersal is sexually dimorphic: females tend to stay at home.  Also,
males explore more.  Provisioned males (males with more food) will
disperse earlier.  But there's no correlation between testosterone
levels and dispersal (males and females have the same testosterone
levels when they disperse).  But it's possible that, even though
there's no activational difference from testosterone, there might be
an organizational difference..  Provide testosterone to mothers, and
their daughters will disperse much further.

>>> Dispersal in primates: effects of status
Primarily the males disperse.  Pubescent males are forced out of the
group by the dominant male.  Dispersing males can either try to
directly take over a new group, or to slowly edge their way into
groups.  Sons of high status females show more play, less stress, less 
fight-or-flight, more exploration, etc..  They tend to be the ones to
directly take over a group.  The sons of low status females tend to
try to sneak into a group.  This has something to do with hormones,
although I'm not sure what -- maybe the sons of high status females
have more testosterone?

>> Agressive behavior is often sexually-dimorphic
>>> Male mouse agression: activational vs. organizational effects
Mice use odor cues to decide whether to attack mice.  Male mice will
only attack mice whose urine has testosterone (they won't attack
females or castrated males).  Being exposed to a female can increase
agressive behavior.  Where do these male/female differences come from?
Agression comes from the organizational effects of testosterone: if
there is testosterone circulating early, they will become agressive;
otherwise, they won't.  You're more agressive if you grow up between 2 
males than if you grow up between 2 females.

>>> Social dominance in beagles
These dogs form linear dominance hierarchies.  You can masculinize
females by giving them testosterone early.  Between a male and a
masculinized female, they'll fight for the bone; but females will
always let males take the bone\ldots  

>>> Rough-and-tumble play in rhesus macaques
Males prefer agressive, rambunctios play.   If you masculinize
females, they will engage in more rough-and-tumble play.  Thus, this
seems to be an organizational effect, not an activational one.

>>> Female can be more agressive - the spotted hyeana
Social animal.  Females are socially dominant to males.  They're more
agressive.  Females have pseudopenises.  Fairly small differences
\between males and females in testosterone.  But there's another
androgen, that females have more of.  Inject animals with that
androgen, and they get more agressive, etc.

>> Agression can increase testosterone levels
The "challenge hypothesis:"
>>> Male red-wing blackbirds: T and LH
>>> Seasonal correlation between agression and T
>>> T important for initial interactions only
>>> Increased T: Having a territory or acting agressive?
>>> Simulated intrusions and T levels
>>> T-implants: effects on nearby males
>>> Non-T factors important in some species
>>> Correlation between challenge response and mating system
>>> Costs related to chronic T levels?

[12/01/99 09:07 AM] Animal Intelligence and Cognition
!!You can turn in papers on monday (in class)!!

Problems with trying to judge "intelligence:" 
  - what is it?  
  - how do we measure it?  

> Scala Naturae

Aristotle proposed a Scala Naturae: a hierarchical classification of
living organisms.  Scala Naturae assumes:
  - "intelligence" varies along a linear scale

But evolution gave us an evolutionary tree.. 

Differences between S.N. and tree view:

#  Phylogenetic Scale       Phylogenetic Tree
#  ------------------       -----------------
#   hierarchical(?)           non-ranked
#   unilinear                 linear
#   less complexity           sequence indep. of complexity 
#       \to more complexity
#   complexity \uparrow with time    complexity independant of time 

>> Silly Ideas to Try to Preserve Scala Natura

One proposal: see how far back they split off (from our heritage): the 
more like us they are, the more intelligent.

Another proposal: Try ranking intelligence based on the size of
brain.  Or on ratio of brain size to body volume.  Or on ratio of
brain size to skin surface area.

> Measuring Intelligence

>> Ability to Adapt
Barnett proposed that intelligence is the ability to adapt (i.e.,
change behavior adaptively in response to a chaning environment).  So
measure intelligence through learning experiments.  But different
animals can adapt in different ways..  A variety of studies based on
this view gave confusing results.  

(spider joke: take off all its legs & its deaf)

How do you decide which questions to ask?  The questions asked have a
large effect on what answers you get.  The garcia effect (bait
shyness): radiation can cause you to develop a food aversion.  Bait
shyness is single-trial learning, and can last a lifetime.  But other
types of negative feedback will need many more trials to learn
association..  There are some types of learning that brains are
specialized to do..  Can teach rats to rear up on hind legs to avoid
shock, or to press bar to get food, but can't teach them to rear to
get food.

It's very hard to find experiments that are eqivalant between
species. 

>> Information Processing
Intelligence = ability to take info in, process it, and use it.  This
model has turned into the modern field of compartive cognition.
Learning set: learning non-specific facts through a series of trials
(e.g., learning what the basic task is).  

> Intelligence as a Justification
If someone is more intelligent, we shouldn't use that as a
justification for subjugation, etc.. 

Problems with standard iq test: correlates well with socio-economic
status; cultural biases; test conditions; etc.  

If we -could- find a good metric, then there's no reason to believe
that different populations would score the same.

But at any rate, between-group variances are much smaller than
within-group variances.  

> Why Study Intelligence?
Try to learn more about how we deal with our environment, and about
how our intelligent behavior works.  Comparitive cognition: try to
figure out how cognition works in a variety of animals, so we can
understand cognition better.  

H.M.: lost declarative memory, preserved skill memory etc. 

[12/06/99 09:06 AM] Evolutionary Psychology and Human Ethology

> Caveats about Evolutionary Study of Human Behavior
Our psychological traits have evolved.  blah...

>> Evolutionary does not = "genetic"
You can never ascribe a purely genetic cause for a behavior.  The fact 
that you can't find genes for something doesn't mean its not
evolutionary.  Also, evolution can occur through non-genetic
pathways.. 

>> Evolutionary does not = unmodifiable
Evolved behaviors tend to be very flexible.. Don't expect anything
less flexible for humans.

>> Evolutionary does not = optimally designed
Human evolved behavior need not be completely optimal.  In fact, we're 
in such a new environment that many of our traits are probably
maladaptive.  E.g., people like to eat fatty foods.  There's often
costs associated with different adaptations.  Have to balance costs
and benefits.  Adaptations should be seen as "good enough."

>> Evolutionary does not = conscious strategies
There's no reason to think our adaptive choices should be controled by 
conscious mechanisms.  When we talk about strategies, we mean that a
particular adaptation has had strong selection pressure on it to make
it more prevalant... 

> What is Evolutionary Psych?  What are its Premises
The study of evolved psychological mechanisms for behavior in humans.

>> Adaptations for the problem of survival & reproduction
Almost all adaptations are specifically designed to (eventually) help
us survive and reproduce.

>> Domain-specific psychological adaptatoins
Each adaptation seems to depend on a fairly small range of
information.  Posit a number of domain-specific strategies, that get
used in the appropriate contexts..

>> Variable output of evolved psychological strategies
When there are strategies/mechanisms, there are a number of ways that
humans can responds.. E.g., there may be some module that causes
jealousy.  It might cause a number of different behaviors, changes in
state, etc., such as hitting someone, feeling angry, etc.

> Human Sexual Strategies and Behavior: Females
>> Long-term mating strategies

>>> The return of sexual selection
First, consider long-term bonds (marriage, etc).  Females can produce
children much slower than males.. Makes them a scarce commodity, so
they can be choosy. (operational sex ratio is relatively high?)

>>> Preferences for males with resources
Since females can be choosy, they should pick males with high
resources (more children).  They tend to prefer men with more
resources: 
  - women prefer mates with good financial prospects (more than men)
  - women prefer mates with high social status (more than men)
  - women prefer older men: status and income tend to increase with
    age. 
  - structural powerlessness claims that, through social conditioning, 
    and sex differencess in access to power, the only way that women
    can get power is through men.  So they try to mate with more
    powerful males, to get access to power.  It's emperically refutted 
    to some degree.  E.g., it predicts that women with access to
    power should act like men, but they don't.  Also, it gives us no
    explanation for the power differences..  Other experiments help
    discredit structural powerlessness.
  - women prefer physical features of dominance: height.  Taller men
    are more likely to be hired, more likely to be promoted, etc.
    Also, more likely to be good leaders, etc.

>>> Preferences for males with good health & "genes"
Mating with healthy males has direct benefits (you won't get sick) and 
indirect benefits (your children will have better genes).  Healthiness 
indicates a (potentially heritable) ability to defend vs parasites.
Try to pick symmetric mates..  Fluctuating asymmetry = absolute amount 
of asymmetry.  Maintaining symmetry indicates that you are resiliant
to environmental difficulties.  Facial symmetry correlates well with
both physical and emotional health.  Females tend to pick symmetric
males.  

>>> Actual behavioral choice of males?
Most of this is self-reported.  Look at actual behavior.
  - Female response to male personal ads.  Females respond more to
    older, wealthier, taller men.
  - Who do you marry?  Usually older men.  Also, prettier women tend
    to marry higher-status men.
  - Male resources matter: Kipsigis are a tribe.  Males own cattle.
    Females must marry to get access to resources.  Often arranged
    marriages.  The more land a woman has, the higher her lifetime
    reproductive success will be.

>> Short-Term Mating Strategies
Women do have short-term mating strategies.  Extra-marrital affairs
exist in every culture studied.

>>> EPC's and copulatory orgasms
Copulatory orgasm may be involved in human sperm competition.  After
copulation, females discharge some portion of the sperm (about 35%).
Female copulatory orgasm decreases the amount of sperm discharged (by
a small abount -- about 5%).  Also, most EPCs occur during ovulation.
Women may be timing EPCs to maximize the chance of reproductive
success.  

>>> Costs and beneffits of short-term mating
Males show a strong dis-preference for long-term-relatoinships with
females with a large sexual history or a large number of EPCs.

Females may use EPCs for: 
  - mate switching (try to get a new mate, but don't throw away your
    old one until you're sure)
  - gene selection

>> Physiological mechanisms of female sexual behavior
>>> Lack of stereotyped postures
>>> Hidden ovulation
No obvious indicators to males about when females are ovulating.

>>> Sex and the menstrual cycle
Sexual behavior doesn't seem to change much over menstrual cycle. 

>>> Direct manipulations of steroid hormones and sexual behavior
Direct manipulations of some steroids (esp. androgens) increase sexual
desire and sexual behavior.  (Inject historectomized females with
hormones.)