Transcript Cooperation

Chap.09 Cooperation
鄭先祐 (Ayo) 教授
國立台南大學 環境與生態學院
生態科學與技術學系
環境生態研究所 + 生態旅遊研究所
Cooperation
The range of cooperative behaviors
 Helping in the birthing process (fruit bat)
 Social grooming (primates)
Paths to cooperation
 Path 1: reciprocity
 Path 2: byproduct mutualism
 Path 3: group selection
Coalitions
Phylogeny and cooperation
Interspecific mutualisms
2
Ayo 2010 Ethology
Cooperation
 The word cooperation typically refers to an
outcome from which two or more interacting
individuals each receives a net benefit from
their joint actions, despite the potential costs
they may have to pay for undertaking such
actions.
 例如：
 jointly hunting group
 Two male guppies (lower left and lower center of
photo_ inspect a pike cichlid predator. Guppies
cooperate during such risky endeavors. (Fig. 9.1)
3
Ayo 2010 Ethology
4
Ayo 2010 Ethology
The range of cooperative behaviors
Helping in the birthing process
(Rodriques fruit bat)
 Unrelated female “ helpers” assist
pregnant individuals in the birthing
process.
 During the birthing process, the helper
continues providing assistance by
grasping the wins of the pregnant
females, thereby providing both
protection and warmth, and
subsequently cleaning newborn pups
upon their emergence. (Fig. 9.2)
5
Ayo 2010 Ethology
6
Ayo 2010 Ethology
Social grooming
 Social grooming, or allogrooming (Fig. 9.3)
 Social grooming is “tension reduction” within
primate group.
 Primates are capable of exchanging one sort
of resource– for example, social grooming –
for another resource in what amounts to a
“biological marketplace”.
 Numerous experiments in primates have
examined whether individuals cooperate with
one another by exchanging social grooming
for aid during aggressive interactions.
7
Ayo 2010 Ethology
8
Ayo 2010 Ethology
Paths to cooperation (Fig. 9.4)
Path 1: reciprocity
 Reciprocal altruism
 Prisoner’s dilemma (table 9.1)
Path 2: byproduct mutualism
Path 3: group selection
Path 4: kin selection
9
Ayo 2010 Ethology
10
Ayo 2010 Ethology
11
Ayo 2010 Ethology
Prisoner’s dilemma
 If both suspects cooperate, they both receive
a payoff of R (reward, 1 year in jail), and if
they both defect, each one receives P
(punishment, 3 years in jail).
 If suspect 1 defects, but suspect 2 cooperates,
the former receives a payoff of T (Temptation
誘惑 of cheat payoff, 0 years in jail), and the
latter receives S (sucker’受騙者 payoff, 5 years
in hail).
 T > R > P > S (Table 9.1)
 0>1>3>5
12
Ayo 2010 Ethology
Evolutionarily stable strategies
 A individual’s success may depend on what
others are doing
 Evolutionarily stable strategy (ESS):
 the optimal strategy for an individual to follow
when the rewards (payoffs) depend on what
others are doing
 When adopted by most members of a population, this
strategy cannot be beaten by a different strategy:
 no other strategy confers more fitness benefits
13
Ayo 2010 Ethology
The favored strategy maximizes benefit
 A hypothetical population of fish-catching
birds has two strategies for getting dinner
 Catch your own fish or steal one from another bird
 Thievery is favored first: it minimizes its costs
and gets full benefits from the efforts of others
 As the number of bandits increases, so does the
chance of encountering another robber or a bird that
had its fish stolen
 Then, honesty becomes the best policy
 When hard-working birds become common, thievery
once again becomes profitable
14
Ayo 2010 Ethology
The prisoner’s dilemma game
 If a pair of individuals plays the prisoner’s
dilemma game just once, then on the one
play of the game, the only strategies possible
are “cooperate” or “defect”.
 In the iterated prisoner’s dilemma game,
however, more complex rules, including “ifthen” rules of the form “if the other individual
does X, then I will do Y” can be employed –
for example, “if she cooperates, I will
cooperate; otherwise I will defect”.
 Tit for tat (TFT) (一報還一報) = reciprocitybased strategy (Fig. 9.5)
15
Ayo 2010 Ethology
TFT strategy (Fig. 9.5)
1. “niceness” –never the first to defect,
cooperates as long as its partner
cooperates.
2. Swift “retaliation” –immediately
defeats on a defecting partner since it
copies its partner’s previous move and
so, if its partner defects, it defects,
3. To do what their partner did no the last
move.
16
Ayo 2010 Ethology
17
Ayo 2010 Ethology
Predator inspection and TFT in guppies
T > R > P > S (Table 9.2)
Is T > R ?
Inspectors are more likely to get eaten
the closer they approach a predator
(Fig. 9.6)
 So it is more dangerous to be leading an
inspection than lagging behind.
 Inspectors transfer the information that
they receive during an inspection, so that
any fish lagging behind would still receive
the benefits associated with inspection.
 Fig. 9.7 information transfer in minnows.
18
Ayo 2010 Ethology
19
Ayo 2010 Ethology
20
Ayo 2010 Ethology
Fig. 9.7 information transfer in minnows
Some fish (transmitters) were allowed
to inspect, while others (receivers)
could see the transmitters, but not the
predators.
As the predator approached the
transmitters, the latter fed less often
(A), as did the receivers (B), suggesting
that receivers were getting information
about danger from the transmitters.
21
Ayo 2010 Ethology
22
Ayo 2010 Ethology
Predator inspection and TFT in guppies
Is R > P?
 If P is greater than R, it would not pay for
any individual to inspect, and the
phenomenon of inspection would be rare
and maladaptive when it occurred.
 Fig. 9.8 inspection behavior in the wild
Is P > S?
 Evidence from a number of experiments
indicates that single fish suffer very high
rates of predation, suggesting that P>S.
23
Ayo 2010 Ethology
Predator inspection and TFT in guppies
Precisely measuring T, R, P, and S is
difficult.
But it seems meets the prisoner’s
dilemma requirement that T> R > P > S
The dynamic nature of inspection
behavior in guppies and sticklebacks
supports the idea that inspectors do , in
fact, use the TFT strategy when
inspecting potential predators.
TFT = nice + retaliatory + forgiving
24
Ayo 2010 Ethology
Reciprocity and food sharing in vampire bats
A typical group of vampire bats (Fig.
9.10) is composed largely of females,
with a low average coefficient of
relatedness (between 0.02 and 0.11).
Females in a nest of vampire bats
regurgitate blood meals to other bats
that have failed to obtain food in the
recent past. (Fig. 9.11)
25
Ayo 2010 Ethology
26
Ayo 2010 Ethology
27
Ayo 2010 Ethology
Reciprocity and food sharing in vampire bats
 When examined which individuals were
involved in food sharing, it was indeed the
case that, despite the fact that the average
relatedness in groups was low, genetic
relatives were still more likely to swap blood
meals than were other individuals (Fig. 9.12)
 Index of opportunity for reciprocity
1. The probability of future interaction between
group members (TFT model)
2. Blood meals provide a huge, potentially lifesaving benefit for recipients, while the cost of
giving up some blood may not be as great to the
donor.
3. Vampire bats are able to recognize one another
28
Ayo 2010 Ethology
Fig.9.12 vampire bat blood meals
 (A) all possible association patterns that could
be found between the recipient bat and others
in the nest.
 (B) the actual association patterns found
between donors and recipients.
 (C) the genetic relatedness between recipients
and all other roost members
 (D) the actual relatedness between a recipient
and a donor.
 Bats were much more likely to regurgitate a
meal to close kin and to those with whom they
associated more often. Bats were capable of
keeping track of who fed them in the past and
who didn’t.
29
Ayo 2010 Ethology
30
Ayo 2010 Ethology
31
Ayo 2010 Ethology
32
Ayo 2010 Ethology
33
Ayo 2010 Ethology
Neurobiological and endocrinological
underpinnings of human reciprocity
Neuroeconomics: a collaborative
research effort between economists and
neurobiologists who specialize in brain
science.
 Experiments in neuroeconomics typically
involve subjects who are making some
economic decision.
 儀器：
 fMRI (magnetic resonance imaging)
 PET (position emission tomography)
34
Ayo 2010 Ethology
Table 9.3 the monetary prisoner’s dilemma game, The payoff
matrix for the game played by women who were either
cooperating or cheating (defecting), in an economic cooperation
experiment.
35
Ayo 2010 Ethology
36
Ayo 2010 Ethology
(B) The fMRI scans showed that, when both subjects
cooperated, brain areas associated with reward
processing – OFC, rACC, the anterovetal striatum, and
ACC)– were activated.
37
Ayo 2010 Ethology
Oxytocin (OT)
Oxytocin is a hormone that has been
associated with numerous affinitive
behaviors like pair bonding and parental
care in nonhumans.
Because there are dense accumulations
of OT receptors in the amygdala of the
human brain, a region associated with
social behavior.
 OT would also play a role in affinitive
interactions in humans.
 Fig. 9.14 the trust game and punishment.
 Fig. 9.15 Oxytocin and trust
38
Ayo 2010 Ethology
Fig. 9.14 the trust game and punishment
Two subjects played the trust game
while one of them (player A) was
hooked up to a PET scanner that
monitored his brain activity.
The caudate nucleus, which is part of
the dorsal striatum of the brain –
depicted in yellow – was very active
when player A punished player B for
failing to return some of the money that
A had provided to B.
39
Ayo 2010 Ethology
40
Ayo 2010 Ethology
The level of oxytocin
was higher when
subjects believed
money was sent to
them voluntarily
(versus sent as a
function of a random
draw)
41
Ayo 2010 Ethology
Path 2: byproduct mutualism
An individual would incur an immediate
cost or penalty if it did not act
cooperatively
 Such that the immediate net benefit of
cooperating would outweigh that of
cheating.
Fig. 9.16 byproduct mutualism.
 Here, neither person gains by failing to
move a stone that neither can budge alone.
42
Ayo 2010 Ethology
43
Ayo 2010 Ethology
Skinnerian blue jays and byproduct mutualism
 The study of blue jay cooperation
 To use Skinner boxes
 Fig. 9.17 byproduct mutualism and blue jays.
 Three pairs of blue jays were tested in a threestage experiment: Stage 1 = prisoner’s
dilemma, Stage 2 = byproduct mutualism, and
Stage 3= prisoner’s dilemma.
 Jays cooperated when the payoff matrix
matched byproduct mutualism, but not when it
matched the prisoner’s dilemma.
44
Ayo 2010 Ethology
Skinnerian blue jays
 A pair of blue jays, each of whom could peck
one of two keys – a cooperate key or a defect
key.
 After the birds made their decisions, they were
given a certain amount of food. The amount of
food they obtained depended on what action
they took, what action the other bird took, and
which of two different payoff matrices the
researchers had set up. (Table 9.4)
 P matrix: prisoner’s dilemma matrix
 M matrix: byproduct mutualism matrix
45
Ayo 2010 Ethology
P matrix
46
M matrix
P matrix
Ayo 2010 Ethology
47
Ayo 2010 Ethology
48
Ayo 2010 Ethology
House sparrow food calls
 House sparrow produce a unique “chirrup” call
when they come upon a food resource.
 These calls appear to attract other birds to a
newly discovered bounty, and as such chirrup
calls may be regarded as some type of
cooperation.
 Chirrup call rates were higher when the food
resource was divisible (Fig. 9.18)
 Chirrup call were associated with larger food
items– that is, those that were too big to
remove from the experimental area.
49
Ayo 2010 Ethology
50
Ayo 2010 Ethology
51
Ayo 2010 Ethology
Path 3: group selection
Within-group selection
 Selection against cooperators and altruists.
 Selfish types – those who do not
cooperate– are always favored by within
group selection
Between-group selection
 Favors cooperation
52
Ayo 2010 Ethology
Group selection in ants
 Cooperative colony foundation has been well
studied in the desert seed harvester ant
Messor pergandei.
 Between-group selection: adult ants are very
territorial, and “brood raiding” – wherein
brood captured by ants from nearby colonies
are raised within the victorious nests, and
colonies that lose their brood in such
interactions die.
 Within groups, all co-founding queens in a
nest assist in excavating their living quarters,
and each produce approximately the same
number of offspring. (Table 9.5)
53
Ayo 2010 Ethology
54
Ayo 2010 Ethology
Group selection in ants
 Until workers emerge, queens within a nest do
not fight, and no dominance hierarchy exists.
 Fig.9.19 from cooperation to aggression
 Co-founding queens are cooperative during
worker production, with very little queen-queen
aggression during this phase of colony
development. Once workers are producedknown as “ worker eclosion”, which is when
they emerge from eggs, however, queen-queen
aggression in nests escalates, as does the
queen death rate,
55
Ayo 2010 Ethology
56
Ayo 2010 Ethology
Acromyrmex versicolor
 Many nests are founded by multiple queens, no
dominance hierarchy exists among queens, all
queens produce workers, and brood raiding
among starting nests appears to be common.
 Queens forage after colony foundation.
 As a result of increased predation and
parasitization, foraging is a dangerous activity
for a queen.
 Foraging involves bringing back materials that
increase the productivity of the nest’s fungus
garden, which is the food source for the colony.
57
Ayo 2010 Ethology
58
Ayo 2010 Ethology
A single queen taking on the dangerous role of forager for
everyone in the nest. All indications are that reproduction
within nests is equal between foragers and nonforagers.
59
Ayo 2010 Ethology
Coalitions
 Dyadic interactions, two individuals interact
 Polyadic interaction, interactions that involve
more than two individuals.
 One example of polyadic interactions involving
cooperation is coalition(聯合) behavior, which
is typically defined as a cooperative action
taken by at least two other individuals or
groups against another individual or group.
 When coalitions exist for long periods of time,
they are often referred to as alliances (同盟).
60
Ayo 2010 Ethology
(A) Three male dolphins swim together, forming a
long-term coalition (or alliance)
(B) Pairs of male chimps often form coalitions to act against
larger, more dominant, individuals.
61
Ayo 2010 Ethology
Coalitions in baboons
Male reproductive coalitions in baboons
(Papio anubis)
Fig.9.22 baboon coalitions.
 A male baboon (middle) involved in an
aggressive interaction (with male on left)
will often solicit (請求) others to aid him by
turning his head in the direction of a
potential coalition partner (male on right).
62
Ayo 2010 Ethology
63
Ayo 2010 Ethology
Alliances and “herding” behavior in cetaceans
 Bottlenose dolphins
 “first-order” alliances in dolphins are
composed of pairs or tros of males acting in a
coordinated fashion to keep females by their
side, presumably for the purpose of mating.
 Different first-order alliances also join together
in “second-order” superalliances and
aggressively attach and steal females from
other alliances.
 The complex social interactions inherent in
dolphin superalliances, may explain the
evolution of large brain size in dolphins.
64
Ayo 2010 Ethology
Phylogeny and cooperation
 Phylogeny and cooperative breeding in birds
 166 species of cooperatively breeding
passerine birds in ninety-seven genera
 The distribution of cooperative breeding
species in nature differed significantly from
the random distributions generated by
computer simulations, with some genera
having more than the expected number of
cooperatively breeding species, and others
less than the expected number (Fig. 9.23)
 Phylogeny and cooperative in social
spiders
65
Ayo 2010 Ethology
66
Ayo 2010 Ethology
Phylogeny and cooperative in social spiders
social spider species: 23 out of about
39,000 species
In these species, individuals build very
large communal webs, jointly maintain
these webs, cooperatively hunt for prey,
and cooperate in raising brood born in
their colonies (Fig. 9.24)
 With respect to foraging, this sort of
cooperation allows spiders to obtain
more and larger prey.
67
Ayo 2010 Ethology
68
Ayo 2010 Ethology
Phylogeny and cooperative in social spiders
Phylogenetic analysis found that
sociality had evolved either 18 or 19
different times in spiders.
 This (18 or 19) is a remarkably high
number of evolutionary origins for
cooperation.
Social spider life may be “evolutionary
dead ends” (high rates of extinction)
 Inbreeding and skewed sex ratios
(females dramatically outnumbering
males, sometimes in a 10:1 ratio)
69
Ayo 2010 Ethology
Interspecific mutualisms
Ants and butterflies – mutualism with
communication?
In numerous species of butterflies and
ants, a mutualistic relationship has
developed in which butterfly pupae and
larvae produce a sugary secretion that
ants readily consume, and ants protect
the larvae from predators such as
certain species of wasps and flies. (Fig.
9.25)
70
Ayo 2010 Ethology
71
Ayo 2010 Ethology
Imperial blue butterfly and ants
The benefits to both parties in this
mutualism are enormous.
Butterfly larvae are much less likely to
survive when ants are experimentally
removed from their environment (Fig.
9.26).
Stridulating attracts ants (Fig. 9.27)
 Muted pupae, the experimenters applied
nail polish to its stridulatiory organs.
 Stridulating pupae attracted more ants
than muted pupae.
Stridulate 發尖銳的摩擦聲(尤指昆蟲如蟋蟀所發)
72
Ayo 2010 Ethology
73
Ayo 2010 Ethology
74
Ayo 2010 Ethology
75
Ayo 2010 Ethology
問題與討論
Ayo NUTN website:
http://myweb.nutn.edu.tw/~hycheng/