Greg Detre
Sunday, 19 May, 2002
Gould (1981) described intelligence as the ability to deal flexibly or creatively with problems. Gardner (1983) placed emphasis on symbolic processing as the hallmark of intelligence. Byrne (1995) gives the following definition: �the term intelligence should be restricted to that quality of flexibility that allows individuals to find their own solutions to problems; genetical adaptations, by contrast, are fixed and inflexible, however well tuned to their environments they are.�
In
the famous case of Clever Hans (reported in Pfungst,
1965) many scientists believed that Hans� owner had succeeded in teaching a
horse to do simple mathematics. Careful control experiments eventually showed
that Hans was in fact responding to undetectable and subconscious cues from his
questioner and lost his ability when the questioners and observers did not know
the answers to the questions.
Gould (1994) gives the following account of how a
young chickadee might learn to open sunflower seeds having been attracted to
the seeds by other chickadees feeding, and perhaps finding a seed that is
already slightly open exposing the kernel inside:
�At length the seed he
has been hammering at gives way and he gains his reward. He selects another
seed and this time begins hammering from the direction that gained him access
before. Adjusting his grip and his pecks to a narrower range of angles, he
breaks into the shell a bit sooner.��
(Gould, 1994, p.44)
a simpler and more plausible explanation is that observing another animals behaviour simply facilitates the animals own trial and improvement learning
Tool use has been discovered in other animals, sea otters use stones to open shellfish, chimpanzees use twigs to fish for insects, even adapting the shape of the twigs to make them more suitable, and wooden hammers to open nuts. These behaviours are not innate and, certainly in the case of chimpanzees are learnt by young from their mothers (McGrew, 1998).
Von
Frisch (1967) discovered and decoded complex communication in bees. Honey bee
scouts return to their hives having discovered a new source of food. A
combination of her frantic behaviour, the smell of nectar and samples that the
scout regurgitates attract the other bees to the scout. � The ability to
perform and understand the dance is probably an innate ability for the bees and
it is therefore confusing to refer to it as language because of the similarity
this implies to our own language. And yet the dance is a form of language and
von Frisch is often described as the man who broke the code. But it is not
flexible and would not count as clever behaviour according to Byrne�s
definition.
Although we do have a vocabulary many times larger than the few different words that these monkeys the real distinguishing feature appears to be the grammatical forms that we are capable of. The same word can be used in different sentences to mean different things.
Much
research has concentrated on whether chimps were capable of learning grammar.
The chimp Kanzi (Savage-Rumbaugh et al., 1991) was able to understand
synthesised monotonic speech (eliminating inadvertent cueing) so well that he
could not only segment words and decode their phonemic component, but could
also comprehend the semantics and structure of quite unusual English sentences.
Kanzi was also reported to have produced two-word utterances such as �car
trailer.� These utterances were argued to be novel and not able to be conveyed
by single words, they were especially exciting as they often referred to
objects that were not present at the time. Their has been much criticism of
these results, primarily that there is not conclusive evidence that the chimp
has a clear understanding of the language and is not just learning by more
simple conditioning, especially since the extra excitement that is caused by
such utterances is likely to be of huge reward value to the chimp.
There have been attempts to investigate without using language that have looked at imitation, mirror guided body inspection (self-recognition), deception, role taking and perspective taking.
Povinelli et al. (1990) showed that
chimps appear to be able to model the visual perspectives of other due to their
ability to chose a container as indicated by a �knower� as opposed to a
�guesser ???
example of long-term knowledge gained from other animals: acquisition of a foraging technique in chimpanzees that the animal will keep for a lifetime (e.g. Manning, 1979).
A less complicated form of communication can be found in the group living mongoose, Surricatta surricatta. Usually there is at least one of the adults of the group that assumes a raised sentry position while the other group members forage for food. If there are no predators visible then the sentry will constantly produce the �watchman�s song� (Benekoff, 1997), if there is a predator present the sentry will give an alarm call and then run for a burrow with the other members of the group.
Some
theorists claim that there is a direct link between this form of information
transfer between generations and the way in which our own culture evolves.
Dawkins (1989) goes as far as to propose that there are units of culture, or
�memes,� that evolve in a way that is conceptually similar to the progress of
genes. Baldwin (1895) attempted to provide an evolutionary explanation for
behavioural change using biological conceptions. Baldwin devided imitation into
�organic imitation� and �conscious (i.e. mental) imitation.� Baldwin argued
that biological adaptation in a changing environment was an example of
biological or organic imitation. Morgan (1890, 1896, 1900) criticised Baldwin�s
all encompassing opinion and provided his own dissection of the phenomenon of
imitation from the perspective of the developing child.
Morgan
(1890) proposed three levels of imitation: the instinctive stage, the
intelligent stage and then a further reflective stage. Morgan used the
production of sound as a behaviour to study so that he could contrast behaviour
with species such as parrots and mocking birds. He claimed that at the
instinctive stage the infant�s sound production was automatic, the sound was a
stimulus to the motor system. An older child, Morgan claimed, would hear the
sound that they made, derive pleasure from the sound and be able to hear the
sound according to some �higher faculty.� Morgan thought that this was the
level of imitation that was reached by birds. What made the imitation of human unique
was that, although only gradually, they achieved a higher cognitive state in
which they were no longer subject to the chance occurrence of happy results but
were able to form deliberate and intentional schemes of behaviour. Could it be
that the development of conscious thought came from animals developing more and
more deliberate forms of imitation? This is certainly a very disputable claim
but it is worth considering the possible implication that it could have towards
the importance of this area of study.
Thorndike (1898) provided a clear definition of imitation �learning to act from seeing it done�
but I don't think that�s clear enough � it doesn't take into account behaviour that�s somehow/implicitly influenced by watching someone else as opposed to an actual attempt to reproduce it???
There is much anecdotal evidence of imitation by animals, especially animals living in captivity. Tayler and Saayman (1973) report the amusing story of a captive dolphin that, after repeatedly seeing a diver cleaning a viewing window, started cleaning the window with a seagull feather while emitting sounds almost identical to those of the diver�s air-demand valve and relaeasing a stream of bubbles from its blowhole. The reason the dolphin might imitate such behaviour are unclear, the most likely explanation is that it was receiving positive reinforcement from the humans that were watching it. A more subtle theory is that there is an advantage for the dolphin to experiment with behaviour that it observes in other creatures to see if it can gain any benefit from similar behaviour. Whatever the explanation the consequence of this report is clear, dolphins are capable of accurate imitation of unnatural behaviour.
Much work has been done with tame apes. Hayes and Hayes (1952) report training a chimpanzee to obey the command �Do this� with an attempt to copy their own behaviour. They further report that their chimpanzee appropriated a lipstick, stood on the washbasin and applied the cosmetic to her mouth. Although this seems like another amusing anecdote the implication is clear, whatever her motive, this chimpanzee was capable of replication complicated behaviour.
Byrne and Byrne (1991) report that mountain gorillas learn to process herbs in specialised fashions that vary across local populations. They conclude that such behaviour must be transmitted through generations by imitation in the infants. McGrew (1998) report similar findings for tool use in chimpanzees. It is now almost universally accepted that imitation is important for the learning of behaviour in apes, but is imitation found in other animals?
One of the most investigated groups of animals is the Japanese macaques of Koshima Island. Kawai (1965) reported that potato-washing of a juvenile female, Imo, was soon adopted by most of the younger members of the group. There has been much subsequent questioning of this finding. Green (1975) noticed that only the members of the group that showed this behaviour were given potatoes by provisioner so that the behaviour could easily have been caused by simple reinforcement. Galef (1990) looked at the time scale over which the novel behaviour spread and found that the mean and median times for the acquisition of this trait were two years, surely imitation would allow behaviour to spread almost instantly through a population, as is suggested by the experimental studies of imitation.
Many of the early theories of imitation (e.g.
Darwin and Romanes, 1883) were developed using anecdotal evidence from
observational studies of behaviour. This was criticised at the time for being
unscientific. Thorndike (1898).developed a paradigm that was the first attempt
to experiment and has subsequently been little altered. Using a puzzle box
Thorndike compared the similarity of action pattern and the speed of escape in
na� cats and cats that had previously watched an experienced cat escaping.
Thorndike claimed that in the �impersonal records� of the time to escape there
was true proof that cats were not able to imitate behaviour. Such early work failed
to provide evidence, free from a subjective judgement of similarity of
behaviour, that any animals could truly imitate behaviour.
Fisher and Hinde (1949) proposed that the annoying habit that tits developed in urban areas of opening milk bottles and drinking the first few centimetres could have been developed independently in a few tits and then been acquired by the population through observational learning.� The theory was based on the observation that the behaviour seemed to start in one area and then spread outwards. Sherry and Galef� (1990) have shown experimentally with black-capped chickadees that na� birds that had observed a model opening cream tubs are more likely than non-observers to adopt the behaviour themselves. However this is far from proof that the observer watches the models behaviour and then replicates it. Other possible explanations include various forms of enhancement, in which the activity of the model draws the attention of the observer to the milk bottle - not an obvious source of food for a na� bird. Seeing the model be in the presence of the bottle and not come to any harm will reduce the observers fear of the bottle and make them more likely to approach the bottle. Spence (1937) was the first to propose stimulus enhancement, Thorpe (1963) proposed that there should also be local enhancement, where the observers were attracted to a place and not to a stimulus. In the case of tits and milk bottles the observer may learn the behaviour more quickly because its actions are directed towards the relevant area or object and this might facilitate by trial and error. Pecking and pulling at objects, the actions required to open a bottle, are by no means unusual behaviour for birds. If the actions were atypical of normal behaviour then it would be easier to argue that imitation was truly occurring. Furthermore it has been shown that tits learn to open milk bottles just as quickly if they are simply let in the presence of an open bottle. True imitation along with the other possible explanations of how this learning might occur now come under the general heading of social learning.
Traditionally
imitation has been thought of as the most cognitively demanding explanation of
social learning as the visual representation of the other animal�s behaviour
has to be interpreted and then translated into an effective motor plan. Perhaps
this is not actually as complicated as traditionally thought. Stimulus
enhancement involves being aware of a stimulus, understanding that it is
involved in the obtaining of the reward, and then being able to recognise
similar stimuli
Dawson and Foss (1965) developed the �two-action test� that distinguishes imitation from local enhancement.� The paradigm is that a na� animal will watch a model obtaining food by one of two action patterns. For the paradigm to be effective the behaviour must be unusual, and both actions must lead to the same reward. If the observer prefers to use the technique that was shown to it then this appears to be explicable only by imitation. Dawson and Foss experimented with budgerigars and found that their subjects did indeed use the techniques that they had observed. However, they only had five subjects and they themselves stated that their results could only be thought of as preliminary.
Fritz and Kotrschal (1999) used the two-action test to investigate social learning in ravens. They had food rewards in wooden boxes with sliding doors, which the ravens could open either by pulling red tags at the sides of the box or by using their beaks to lever the boxes open. Observers approached the boxes more readily than na� birds. Observers that had been shown the levering technique only used the levering technique. Observers that had been shown the pulling technique started using both techniques and then began to favour the levering technique. Levering was concluded to be easier. Fritz and Kotrschal did not try and claim that their experiment was conclusive proof for the existence of imitation. They point out that the reduction of fear associated with seeing the models interact with the flaps would be sufficient to explain the discovery that pulling the flaps led to the reward. The models did always gain the reward from on top of the box so it is surprising that the observers did not go straight to the top of the box. They also noticed that the observers always started to obtain rewards by scrounging from the boxes that the models had already opened, this might well be the mechanism that in the natural world leads the birds to learn new behaviours.
Voelkel and Huber (2000) used Dawson and Foss� paradigm to investigate social learning in marmosets. Rewards were hidden in plastic film canisters. The marmosets were divided into three groups. One group observed a demonstrator that was opening the canisters with its mouth, one group observed a demonstrator opening the canisters with its hand and there was a control group that had no demonstrator. All the animals then had two trials. In the first trial the lids were only half-closed, as they had been in the demonstration. In the second trial the lids were completely closed and the monkeys could only open them with their mouths. In the second trial none of the monkeys from the hand group opened the canisters with their mouths, and one monkey from the mouth group that had used its hands in the first trial immediately started to open the fully closed canisters with its mouth. Voelkel and Huber concluded that this must be imitation of behaviour.
Gallese Goldman (1998) report finding neurons in monkeys that respond to a particular behaviour whether it is performed by the subject or by a monkey that the subject is watching. Future advances in neuroscience may lead to an understanding about the comprehension of behaviour and may provide an answer to the questions that behavioural experiments have failed to conclusively answer.
Nicol and Pope (1996) define teaching as when an animal modifies its behaviour, at cost to itself, in such a way that a na� observer acquires the skill more efficiently than it might otherwise do.
Nicol and Pope went on to experiment with chickens
to attempt to discover if chickens teach feeding behaviour to their young.
Using coloured food they had chicks feeding in the sight of an adult bird that
thought the feed they were eating was normal or unpalatable. When the adult
thought the chicks were feeding on unpalatable food they would increase their
ground pecking and scratching behaviour and when the adult was given food they
would eat more. Although this is very different from our sense of the word
teaching it is interesting to note that chickens do alter their behaviour to
facilitate social learning, meaning that such knowledge transfer must be
important.
Boesch (1991) spent ten years observing the chimpanzees of the Tai National Park in the Ivory Coast. He claimed to have found three ways in which mothers help their infants to learn to use tools, specifically how to crack nuts using a hammer and an anvil. (1) Stimulating, mothers leave hammers or nuts near an anvil when the infant starts to show an interest in nuts (normally adults carry the hammer with them while foraging and crack all the nuts they find). (2) Facilitating, mothers will give good hammers to infants at a cost to themselves. (3) Active teaching, in the ten years that Boesch observed the chimpanzees he claimed to have seen two instances in which a mother was seen to correct the technique of her infant. With so few occurrences of this behaviour it cannot really be taken as significant evidence, it could easily just have been a misinterpretation of a chance action.
probability of sharing a certain gene for any relative. It is simply .5 times each independent link (meiotic event) between you and a kinsman going through a common ancestor
In the 1970s Paul Sherman did a series of field tests of some predictions from kin selection. He watched Beldings Ground Squirrels in the Rocky mountains.
These squirrels give alarm calls when coyotes or eagles are near. Sherman and students showed that alarm call givers were likely to be killed in 4% of all predator encounters whereas non callers were killed in only 2% of encounters. Thus giving a call is an altruistic act.
The following things are true about ground squirrels.
1) they live in small overlapping ranges with females and their babies sharing some burrows and males living with a social group for some time period.
2) females stay near the place they were born, whereas males migrate far away when they become adult
3) males move from one group to another almost every year.
4) therefore females tend to live near close kin (sisters, mothers, aunts, etc) whereas males are unrelated to most squirrels in their group.
Sherman predicted that mainly females should give alarm calls if the behavior was due to kin selection. He was right. Females give calls at a much higher rate thaan their proportional representation in the population. Sherman also showed that the females who had very close kin nearby were more likely to call than those who had only distant kin living near.
But in Black -tailed prarie dogs where males also live near relatives, both sexes give alarm calls depending on how many relatives they have nearby
two kin recognition mechanisms. 1) proximity 2) phenotypic matching
Arnold, K., �Kin
recognition in rainbowfish ( Melanotaenia eachamensis): sex, sibs and shoaling�
in Behav Ecol Sociobiol (2000) 48:385�391
see rainbowfish.pdf
Living
with relatives can be beneficial to individuals via the evolution of kin-directed
altruism, but this is tempered by the increased risk of inbreeding. Therefore,
in social species, the ability to recognise relatives can be highly
advantageous. This study focuses on kin discrimination in the Lake Eacham
rainbowfish, Melanotaenia eachamensis, an endangered freshwater species
from north-east Queensland, Australia. First, I examined kin recognition
abilities when a combination of both chemical and visual recognition cues was
available. When given a choice of shoaling with same-sex groups, females spent
significantly longer with full-sibs rather than half-sibs, full-sibs rather
than non-relatives and half-sibs rather than non-relatives. Males spent
significantly longer shoaling with full-brothers versus halfbrothers, but
showed no other shoalmate preferences. Second, in the presence of only chemical
cues, females did not discriminate among groups of different levels of
relatedness, but males showed a non-significant tendency to associate with
full-sibs rather than non-relatives. Male shoaling behaviour seemed to be more
influenced by factors other than relatedness, e.g. intra-sexual competition.
Finally, I found that the shoaling preferences of females changed when exposed
to groups of males. Females preferred to associate with non-relatives rather
than half-brothers and non-relatives rather than fullbrothers. There was no
significant difference in the time spent with half-brothers versus
full-brothers. Taken together, my results suggest that females have very good
kin recognition abilities. They prefer to shoal with female relatives but avoid
male relatives, and so are able to balance the benefits of nepotism and the
costs of incest.
two main reasons why it is beneficial for animals to be able to recognise kin. Firstly helping relatives to raise offspring will increase the genetic representation of the helper in the genepool, this important concept is called inclusive fitness. It would therefore be important for animals to have a mechanism for distinguishing between relatives and non-relatives in order to ensure that the costs they incur by helping have the greatest benefit to themselves. But not all considerations of inclusive fitness involve helping behaviour, a male mouse will kill young that it discovers not to be related to, it is obviously therefore important that the mouse can tell whether or not they are related to the young. Secondly, kin recognition is important for mate selection. Inbreeding increases the number of loci that are homozygous and therefore it is advantageous to be able to not breed with close relatives.
There is extensive empirical evidence for kin specific care, especially in birds. Emlen (1990) reports that 45% of helping in the white-fronted bee-eater is directed to rearing full siblings, and that on average r=0.33 between helpers and nestlings. It has been reported that when faced with a choice of potential recipient nests that both the white-fronted bee-eater (Emlen and Wrege, 1988) and the Galapagos mockingbird (Curry, 1988) will preferentially help the breeding pair to whom the helper is most closely related. Only the Mexican jay (Brown and Brown, 1990) is not reported to show kin favouritism. The indirect fitness benefit to the helper will be reduced when the offspring are only half-siblings and theory would therefore predict that they should reduce their contribution to raising the offspring, evidence for this is not conclusive but the effect has been found in white-fronted bee-eaters (Emlen, 1997).
scepticism as to whether cooperative behaviour is actually not just selfish behaviour. Heinsohn and Legge (1999) suggest the following possible direct reasons that might explain cooperative behaviour (as opposed to the indirect benefits associated with inclusive fitness): increased social prestige, the payment of rent in return for being allowed to live on a desirable territory, direct access to parentage, enhancement of group and territorial quality, formation of alliances to aid in competitive situations, and enhancement of skills to aid in later reproduction. They cite interesting observational work by Zahavi (1974, 1995) in which he showed that Arabian babblers would pretend to feed the young in the nest and if there were no other adults near-bye would then eat the food themselves, suggesting that there must be benefits in being thought of as a good helper or the animals would not waste energy pretending to feed the young. In his review of helping behaviour Cockburn (1998) showed that where helping is male biased there are usually no improvements in offspring numbers, but where helpers are mixed or female there is always a positive effect. This would suggest that it might not be important for male birds to be able to recognise kin as the benefits are always selfish, there is as yet not enough empirical evidence to suggest if this is the case.
Hamiltons
equation is elegant and conceptually important but includes no reference to the possible fitness benefits of helping at
the nest and should therefore be considered too simple to be a true
representation of what is actually occurring in the real world. It is also
worth considering if it would ever be possible to obtain a meaningful real
value for either b or c.
As well as the above observational evidence it has been shown in the Caspian tern that although the bird will adopt young substituted for their own they will always feed their own young if given a choice between their own young and alien young in nest scrapes on either side of the original nest scrape (Shugart, 1987). Beecher (1991) reports his own experiment of 1981 in which a single rough-winged swallow chick was added to a bank-swallow brood the foster chick was typically rejected indicating that the bank-swallow parents were capable of individual recognition. However rough-winged swallows would invariably accept a bank-swallow chick, so the ability is not universal. In the same experiment Beecher also showed that the birds were able to recognise their own broods when they were swapped with an adjoining brood of the other species. The parents would continue to feed their own brood at the new location.
There is clear
evidence that inbreeding is detrimental. Keane (1990) showed that inbreeding in
white-footed mice leads to smaller litter sizes and lower offspring weight at
weaning. In great tits inbreeding causes more hatching failure (Noordwijk and
Scharloo, 1981) and higher nestling mortality (Greenwood, Harvey and Perrins,
1978). This does not mean that every animal should try and find a mate as dissimilar
to itself as possible, at an extreme perspective this could lead to animals
breeding with the wrong species; but too much outbreeding could break up
co-adapted gene complexes or cause the loss of adaptations that were
specifically useful to a local environment (Shields, 1984). These complications
give a clue to how specialised kin recognition has to be if it is going to be
possible to achieve what Bateson (1978) labelled �optimal outbreeding.� Bateson
(1982), himself, found that quail prefer the company of first cousins to
brothers or third cousins and that birds mated to first cousins breed rather
earlier (Bateson, 1988).
Kin recognition is
currently explained at four different, but not exclusive, levels
histocompatibility complex (MHC), a hugely variable gene that is thought to have at least 60 alleles. The loci produce cell surface molecules that differ between individuals in order to provide a defence against pathogens that might try to mimic such characteristics. The metabolic waste produced by these molecules contributes to urine odour and it is thought that it is this odour to which mice are sensitive (the odour may be expressed as sweat in humans).
In mice the MHC genes have been shown to alter behaviour in three contexts. It has been shown that mice prefer to mate with individuals differing from themselves at this group of loci (Yamazaki et al., 1976). Females are more likely to nest communally with females that are the same at the MHC (kin selection), a pregnant female will often abort a litter or engage a �pregnancy lock� if she smells a strange male (Lenington, 1994). This is clear evidence of individual recognition and a corresponding active change in behaviour. More information will be discovered about the genetic influences of kin recognition as genetic understanding and technology improves.
Porter, Wurick and Pankey (1978) showed that spiny mice prefer to associate with siblings rather than non-siblings but that the discrimination was based on the mice an individual had been housed with and was therefore altered by familiarity
forming
a broad view of what they are like, this may be based on parents, siblings or
self, and other individuals are compared to this standard. Individuals that show
low discrepancy are then treated as kin. To separate this from a familiarity
system animals should be able to distinguish between unfamiliar kin and
unfamiliar non-kin. Sibling mice are less aggressive to one another than are
non-siblings even if they have never encountered one another before (Kareem and
Barnard, 1986). Many different characteristics could be used to distinguish kin
from non-kin as long as it is more similar in kin than non-kin.
Bateson's�(1982) quail appear to use visual cues, McGregor (1989) claims that
vocal signals learned from relatives are important, although evidence is not
conclusive.
in species that do not disperse very far an animal�s neighbours are likely to be its kin.
There is evidence that the situation will change behaviour and that this is linked to the probable, if unproven, relationship between an adult and young. Male rodents often kill pups they encounter but are less likely to do this after a period of living with and mating with a female (e.g. Elwood and Ostermeyer, 1984).
debate as to whether spatially based recognition and familiarity count as kin recognition at all. Grafen (1990) defines recognition as �recognition by genetic similarity detection.� Although his opinions are perhaps a little extreme and should not mean that more general forms of kin recognition are ignored it is worth considering that the only system currently known that can provide a true prediction of r is human language. This has serious consequences when considering models such as Hamilton�s.
a widow bird flying low over the vegetation displaying the length of its tail to attract mates
sensory modalities:
patterned flowers attract insects, many mammals scent-mark their territory, sperm whales emit low frequency sounds that can travel for 100 miles, chimpanzees touch and kiss each others hands in a gesture of reconciliation, and the male electric fish seduces his mate with electircal impulses
Zahavi (1991) defines a signal as �a character the adaptive significance of which, to the signaller, is to provide information which may change the behaviour of other individuals.� He continues by stating that other characteristics may provide information (e.g. size) but that the observation of these non-signal characteristics is no different to the observation of inanimate object. Zahavi�s definition is clever because it encompasses many of the different areas of signalling. These include mimicry, deception and the simple languages that are reported in cooperating animals. Burkhardt (1970; cited in Manning and Stamp Dawkins, 1998) illustrates the active requirement of the sender in signal communication with the example of a snake that uses ants� scent trails, that the ants use themselves to find their way back to their nest, to feed from the colony. Although the snake is using the ants� trails no one would argue that the ants were signalling to the snake. A benefit to the sender must therefore be an important requirement for a behaviour to be classified as a signalling behaviour.
What are the essential properties
of an effective signal? The signal must be clearly perceivable by the receiver.
It must be on average reliable, if a signal were too exploitable by cheats
there would be a huge selection pressure on the receiver to ignore such
signals. There is also a selection pressure on the sender that the signal
should be as energetically cheap as possible while still being effective.
There is an important distinction
between competitive and non-competitive signals. I define a non-competitive
signal as one in which no advantage could be gained from dishonesty, such as
the dance of the honey-bees, if there is no advantage to be gained then it is
highly unlikely that there would be any deception. It is important to be
sceptical about whether any situation is truly without risk from deception.
Vervet monkeys of the Amboseli National Park in Kenya have at least three
different alarm calls for their different predators, leopards, snakes and
eagles, the other members of the group have appropriate avoidance behaviour
(Cheney and Seyfarth, 1990). This seems like a non-competitive signal and
probably is, but it is possible that an individual could put a rival at risk by
giving the wrong alarm call. It is unlikely that vervet monkeys could be
capable of such devious behaviour but it is a good illustration of how hard
completely non-competitive behaviour is too find.
Moving on to competitive deception
it is interesting to consider whether the animal is capable of the deception.
There have been many reports of deception amongst primates. Lawick-Goodall
(1971) reports that a chimp, �Figan,� would lead away the rest of his large
group and then return ten minutes later to get his share of the bananas.
Initially this was thought to be a simple coincidence but Figan performed this
behaviour repeatedly and other researchers have since reported similar
distracting behaviour. Whiten and Byrne (1988) give an example off a baboon
being chased aggressively by another baboon �suddenly adopting the alert
posture and horizon watching normally shown when an important entity like a
neighbouring group or predator had been spotted,� the chaser stopped to look
and then never resumed the chase. Whiten and Byrne refer to this type of behaviour
as �tactical deception.� They propose that to qualify as tactical deception the
behaviour must come from the usual behavioural repertoire, be used rarely in
the wrong context and frequently in the right context and must cause another
individual to misinterpret and give an advantage to the actor. Although the
evidence is by nature anecdotal and relies upon observation, with all the
problems of over interpretation, there is now such a weight of evidence it must
be concluded that many primates are capable of simple, farce like deception for
their own advantage. This is a clear example of an animal signal that is not
honest.
If there are clear advantages to be
gained from such deception what prevents it from taking over as the normal form
of behaviour? It is worth considering some of the problems of this behaviour.
Most primates live in complex social groups and, as in our own society, it is
important not to get a reputation as a liar both because of the problem of then
not being trusted when communication is important (as in �Crying wolf�) and
because the trick would no longer work preventing further advantage being
gleaned. There could also be more serious consequences for having a reputation
as an individual not to cooperate with, thus Whiten and Byrne conclude that the
frequencies at which such behaviours occur must be low, or, as in the case of
Figan, the chances of getting caught must be small.
There are examples of
inter-specific deception, although it does not fit Whiten and Byrne�s
definition, away from the primates. Ground nesting birds, such as plovers, have
been observed luring a predator away from their nest site by feigning a broken
wing. The bird will then fly away once it is far enough away from the nest
site. Although this is not part of the usual repertoire of the bird, it is
trying to give an incorrect message to the predator and can perhaps be
classified as a dishonest signal.
Consideration of tactical deception
in primates and other animals is interesting, and has important implications
for debates into the intelligence of our closest evolutionary cousins, but most
deception has much less resemblance to the non-scientific use of the word
deception.
Mimicry is an interesting form of
deception that involves dishonest signals. Harmless milk snakes look like the
poisonous corral snake and are therefore avoided by other animals. Many birds
will avoid the palatable hover fly thinking it to be unpalatable, like the wasp
that it resembles. Not all mimics are avoiding predators, an angler fish�s lure
allows it to eat fish that come to investigate, female fireflies of the genus
Photurus can flash the pattern of the smaller species Photinus, attracting the
smaller males which they then kill and eat. Although these behaviours and
markings are not consciously intended to deceive and have obviously evolved in
a standard fashion they are still signals according to most definitions and are
indeed dishonest signals. Like all deception it is important that the deception
is not too common within the poulation as otherwise there would be too great a
selection advantage for an animal to develop other strategies. In the case of
the hover fly they depend on most birds choosing not to eat yellow and black
striped flying insects, if all insects had this colouring then there would be
no point in the hover fly investing in such markings.��
Zahavi (1975, 1987) argued that
reliable animal signalling, although his arguments do not refer to signals such
as those previously described in this essay, could only occur if the signal was
a true indicator of the underlying quality of the signaller. This means that
the signal has to be costly (i.e. fitness reducing) to give. He uses, as
examples, signals such as loud calls or costly displays that use a lot of
energy to produce and can only be given by individuals that are capable of
surviving in spite of the cost of giving such signals. Modelling experiments by
Grafen (1990) produced an ESS that at equilibrium had costly signals being a
true guide to signaller quality. Maynard-Smith (1994), however, proposes an
evolutionary stable model in which the signals are cost free.��
In predator-prey relationships
there is much evidence to suggest that the prey will give costly signals to
indicate that they are healthy individuals that will be difficult to catch even
thought the signals appear to hamper the chance of the animal getting away.
When chased by a group of wild dogs gazelles and springbok will jump high into
the air in a behaviour called stotting that would appear to draw attention to
themselves, and certainly reduces the distance that the animal puts between
itself and the dogs. Fitzgibbon and Fanshawe (1988) found that stotting is done
more by animals in a good condition and the rate and height at which the animal
leaps appears to be a true indication of their condition. The wild dogs appear
to discriminate between gazelles stotting at different rates and will
concentrate their efforts on the animals that do so the least. Stotting is a
costly behaviour that appears to be an honest measure of the strength of the
gazelle. Cresswell (1994) investigated the apparently suicidal behaviour of
some skylarks that sing while being chased by merlins. Curiously skylarks that
give a continuous song while being pursued are less likely to be caught.
Cresswell found that males in good condition continuously sang, he concluded
that by avoiding the singing skylarks the merlins could actually find the
weaker birds.
One of the most interesting areas
of signalling concerns sexual selection. Darwin (1874) suggested that as well
as natural selection the evolutionary paths of animals were controlled by
sexual selection: �the males have achieved their present structure, not from
being better fitted to survive in the struggle for existence, but from having gained
an advantage over other males, and from having transmitted this advantage to
their male offspring alone.� Although these ideas have not been universally
acclaimed many researchers believe that some mechanism of sexual selection is
accountable for many of the strange attributes that some males have, such as
the large tails of peacocks and widow birds. Zahavi believes that these
attributes are costly signals that give a true guide to underlying fitness, the
idea is that if the male is capable of investing so many resources in his
displays or surviving with an apparent handicap then he must have good genes
and is therefore a suitable mate. Fisher (1930) narrowed sexual selection to
consider only those processes that produced extravagance, ignoring abilities
such as fighting that do not need a distinct evolutionary mechanism. Zahavi
(1991) actually proposes that we should stop using the term sexual selection
and instead proposes that evolution is the result of natural selection and
signal selection.
For sexual selection there are two
different classes at which signals work. Signals are only really important if
the father plays little or no role in raising the offspring. The important
difference is whether the female has the choice as to who she mates with. If
the males fight each other for control of the females then it is abilities such
as strength and stamina that are important and the signalling systems involved
are concerned with communicating this fighting ability, Zahavi (1991) refers to
this as �war propaganda.� If it is the female that chooses then it becomes more
debatable as to what the signals are trying to convey.
The most reported case of
war-propaganda is the aggressive displays of roaring seen in red deer as
studied by Clutton-Brock and Albon (1979). During the breeding season a stag
will devote most of his attention to defending a group of females or trying to
get control of a group of females by challenging another stag. Normally when a
challenger approaches a resident stag he will not immediately start fighting
but will begin by giving a small number of roars, the owner replies by roaring
at a slightly higher rate and this roaring contest will oscillate until the
stags are roaring at a high rate. If the challenger is out roared then he will normally
retreat without fighting. It is usually only if the challenger can either roar
at the same rate or out roar the owner that a fight will ensue. It appears that
the challenger trusts the owner�s roaring ability as a measure of his fighting
ability and is obviously not keen to get involved in costly fights that he will
probably lose. Clutton-Brock and Albon claim that roaring is an honest signal
because the muscles and stamina required for sustained roaring are the same as
those involved in fighting and therefore this is an honest signal that cannot
be faked. Zahavi would argue that what is important is that the signal is
costly, this is true but it is more important that the signal uses the same
properties as fighting does. Are there any cases when the aggressive signal is
not a true guide to fighting ability? Cases when the bark truly is worse than
the bite. There probably are but they are unlikely to be as influential as true
signals.
In many species the male will use a
signal to attract females, and sometimes females use signals to attract males,
as in fireflies. What are the mechanisms at work here? Fisher (1930) proposed
that super-normal male attributes could be explained by females choosing males
that they found attractive, there daughters would also find similar properties
attractive and there would then be a selection pressure to develop more and
more extreme attributes in order to appear attractive. This theory is much
simpler than the apparent paradox of the good genes theory � that by showing that
they can survive with a disadvantage they in fact have good genes, when they
would really be much better off without the disadvantage. If Fisher is correct
then the signals conveyed by male attributes are more involved with attention
seeking and there is therefore no an argument as to whether they are honest
signals.
There are simpler cases. Male frogs
use sound to attract females. In the Tungara frog a low-frequency �chuck�
attracts more females (reported by Guilford and Stamp Dawkins, 1993). Only large
males are capable of low-frequency chucks and it is advantageous for a female
to find a large male as they are able to fertilise more eggs. This may indeed
be a costly signal but as with the roaring of the red deer I would argue that
being costly is not as important as the fact that it is unavoidably linked with
relevant information, and so cannot be faked.
Animal signalling is an immense
area of research. If signals occur in a cooperative situation when there is no
advantage to be gained from dishonesty then there will be no dishonesty.
Dishonesty is found in that live in advanced social systems, although only
reported in primates it may be found in other animals. Animals trick animals
from other species, either in defence of themselves or in trying to catch prey.
Dishonesty is only possible at low frequencies or there is too great a pressure
to evolve methods to avoid it.The only truly reliable signal are those that are
linked to the ability which is being signalled. Consequently they have to be
costly but I argue that the link is more important.
see http://www.bio.unc.edu/faculty/pfennig/lab/pfennig_files/kinrecoginization.htm
Griffin & West (see �griffin west.pdf�)
kin selection has been very successful (e.g. in explaining sterility in eusocial insects (Hamilton 1964) or avoidance of cannibalism in salamanders (Pfennig 1999))
when considering the evolution of a supposedly altruistic behaviour, a high relatedness between interacting individuals is usually taken as evidence for a primary role of indirect fitness even though it is not, in itself, sufficient evidence for the role of kin selection
Conclusions:
There can be considerable direct fitness benefits to supposedly altruistic traits. These can be more important than indirect fitness benefits even when interactants are close relatives. Empirical work is required to quantify the relative importance of indirect and direct fitness benefits of behaviours.
Multiple mechanisms and their interactions working simultaneously might be crucial in many cases. For example, a subordinate meerkat that helps raise the offspring of another individual's offspring might be increasing: (a) its own survival; (b) its chance of attaining dominance in a group; and (c) the number of helpers it will have if it obtains dominance in that group (let alone any indirect fitness benefits) (Clutton-Brock T.H. et al. (2001)). Importantly, the relative importance of different mechanisms might vary with sex and status (e.g. natal versus immigrant), as well as between species. Empirical work is required to determine the relative importance of different mechanisms, how they interact and how they vary both within and across species
Competition between relatives reduces kin selection for altruism. Extreme cases, such as fighting in fig wasps, show how this can even totally negate kin selection for altruism towards relatives. The next crucial step is to estimate the importance of competition between relatives in other systems where it is likely to have more intermediate effects, ranging from cooperatively breeding vertebrates to facultatively social insects to malaria parasites to bacteria.
K.
M. Passek and J. C. Gillingham [1999, Copeia (3):831-835] investigated
kin-recognition abilities in hatchling American alligators (Alligator
mississippiensis) in three separate experiments: association preference,
olfactory cue detection, and physical contact tests by examining the
relationship between animals that chose to group together. Hatchling alligators
preferred to group with other hatchling alligators over remaining solitary or
seeking cover near vegetation, but these hatchlings did not necessarily have to
be kin. Alligators also did not show a preference for water-borne chemosensory
cues of kin over those of non-kin. Alligators spent equal amounts of time with
both kin and non-kin, familiar and nonfamiliar animals. These results are
discussed in terms of alligator maternal behavior and ecology.
Female lions risk their own lives to defend another lion�s cubs, vampire bats give up their own hard earned food to less fortunate unrelated foragers, a mongoose will give up searching for food in order to guard the rest of its group � why?
Dugatkin (1997) gives the following definition of cooperation:
�Cooperation is an outcome that despite potential relative costs to the individual is �good� in some appropriate sense for members of the group, and whose achievement requires collective action.�
Darwin worried that if all creatures were involved in a race to reproduce as much as possible then there would be no cooperation between animals � eventually concluded that it could be explained as giving a selection advantage to a group rather than an individual, this theory has been called the �group selection theory.�
Grinnel, Packer and Pusey (1995) suggest that there are three possible routes through which cooperation may develop:
1. kin selection (Hamilton 1964) is the theory that the inclusive fitness effects of cooperation can outweigh the temptation to defect because the payoff for cooperation exceeds that of defection.
2. reciprocity (Trivers, 1971; Axelrod and Hamilton, 1981) a theory that involves looking at repeated interactions with reference to previous behaviour, reciprocate cooperation and exclude defection
3. mutualism (Maynard Smith, 1983; Lima, 1989) when it is considered that there is no temptation to defect because the payoff from cooperating is always higher no matter what the behaviour of the opponent. The first two theories involve conditional cooperation, mutualism is unconditional because the individual will always cooperate.
Dawkins (1989)
describes the reasoning in whether to defect or co-operate: If A defects then
the best thing for B to do is to defect too. If A co-operates then it is still
more advantageous for B to defect. So on a single game a rational player has no
choice but to defect, although they know that their opponent will do exactly
the same thing, if only they would both cooperate then they would both be
better off.
The conditions for IPD (iterated) are that no
information may be exchanged, participants are allowed to �score keep� and the
participants must interact a large but unknown number of times. Dawkins
explains why there must be an undetermined number of games: If the players know
how many rounds there are going to be, say 100, then the final game will be
equivalent to a single game so both players will defect. If they know that they
are always going to defect on the 100th game then the 99th
game effectively becomes the last game so again both players have to defect and
so on until both players defect always.
Rappaport.
TFT starts by playing C and then copies the previous move of its opponent.
According to Axelrod its success is because it is a �nice� (it is never the
first to defect), �forgiving� (its memory lasts only one turn), �not envious�
(as long as it succeds itself it does not mind how sucessful its opponent is)
and yet still has the threat of punishment for defectors. There are, however,
still criticisms of TFT. It has been described as brittle, a small mistake
between two tit-for-tatters can lead to generations of failed cooperation. A
population of tit-for-tatters can be invaded by a an all C strategy and this
will then leave the population open to further invasion by an all D strategy
which would cause long term damage.
plus it starts out playing nice, and so could lose
to a tit-for-tat that starts out defecting
Nowak and Sigmund (1993) win-stay, lose-shift (Pavlov)
if:
both players cooperate
or Pavlov defects and its opponent cooperates
then Pavlov keeps the same strategy (both considered a win for Pavlov)
if:
the result is one of the two possible other results
then the decision is shifted.
The extra success found in Pavlov is due to the fact
that it is able to correct mistakes and can exploit an all C strategy if it
discovers no retaliation, however it is susceptible to permanent cheats???
it is able to correct mistakes and can exploit an
all C strategy if it discovers no retaliation, however it is susceptible to
permanent cheats. Nowak et al. claim that TFT is only successful in a
deterministic cyber world that is free from the errors found in nature and that
a Pavlov like strategy is more likely to account for the way more sophisticated
animals behave.
The simulated success of TFT is used as evidence for the
reciprocal altruism theory of cooperation, but is there any real empirical evidence that TFT
is found in nature?
Wilkinson (1984) investigated food sharing amongst vampire bats. The bats feed on blood during the night. Not all bats will find food every night but those that have found food are likely to have a surplus of blood. Wilkinson observed 110 instances of regurgitations, where a bat that had fed shared its blood with another bat. It is often the younger and less experienced bats that are unsuccessful, and 77 of the regurgitations were mothers feeding their children. But this did still leave some cases where the sharing was between unrelated bats. By examining the body weights of the bats Wilkinson calculated that donation of blood gave more benefit to the recipient (in terms of time till starvation) than was lost by the donor. To prove that the bats could identify each other (a vital part of any reciprocal system) Wilkinson captured a selection of bats from two different colonies a great distance apart. He kept them in a cage together and each night removed one bat, selected at random, before feeding the others. The starved bat was then returned to the cage. 12 out of the 13 regurgitations observed by Wilkinson were between members of the same original population. Bats that had previously received blood were more likely to donate themselves. This seems to be an example of reciprocal altruism.
Adults of the group living species of mongoose, surricatta surricatta, take turns in guarding the group, taking an elevated position and watching for predators. When on guard the animal will make a noise known as the �watchman�s song� which is thought to reassure the other members of the group and will give an alarm call if a predator is sighted. They were thought to be at greater risk from predation as well as giving up forraging time and using up there own energy giving the watchman�s song. This has traditionally been thought of as reciprocal altruism. Benekoff (1997) proposes that it may in fact be beneficial for a near satiated animal to be on guard. Clutton-Brock et al. (1999) found that animals on their own would often adopt this sentinel like behaviour. By feeding animals with eggs they represented extra foraging success and found that this led to animals being more likely to go on guard and that there was no regular rota of guarding duty. There was no apparent risk of predation to the animal on guard. So the only part of the behaviour that seems to only be explained by reciprocal altruism is the alarm call and the watchman�s song, and neither of these behaviours is likely to be much of a disadvantage to a specific individual.
One of the most commonly cited examples of cooperation amongst animals is the lion. They rear cubs, defend territories and hunt together. Legge (1996) describes them as �an emblem of cooperation.� Related females live together in prides, coalitions of 2 or 3, often unrelated males, fight to maintain residencies (Packer et al., 1991) and therefore offer a valuable opportunity to study models of cooperation. Larger coalitions have longer residencies and therefore greater reproductive success (Bygott et al. 1979, Packer et al., 1988). However individual reproductive success becomes increasingly skewed with increasing coalition size. Grinnel et al. (1995) used playback experiments to simulate intrusions into a pride�s territory. The males did not appear to condition their response to either the relatedness or behaviour of their companions, even when other members of the coalition lagged behind. The shared defending of the pride was traditionally thought to be reciprocity. Grinnel et al. suggest that it is actually more likely to be a case of mutualism. A coalition is only likely to establish one residency in a lion�s lifetime and they are therefore going to try and defend it, no matter what the other lions do as it is always in their in their interest to maintain the residency, although it would be even better if the other members of the coalition did the fighting instead.
Heinsohn and Packer (1995) performed similar
experiments using playback but were observing the behaviour of the females.
They found that some individuals consistently led the approach to the speaker
in spite of their companions holding back. They found that when these �lead�
females were paired with a �laggard� the leader would still approach the
speaker but would go slower and would continually check the progress of the
other lion. So the cooperation cannot be maintained by a TFT or a Pavlov
strategy. (???)
The contribution that theoretical modelling has
made to behavioural ecology is enormous. Most ideas start as theoretical and
therefore modelling is very important. However the above examples emphasise the
great diversity of individual behaviour and the inadequacy of current theory to
provide a single coherent explanation. IPD allows no information to be
exchanged between players and is not a good model for complex interactions when
large numbers of animals are involved because it does not allow for payoffs to
animals that are not involved in the game. The studies of lions have shown that
it is important to try and escape from explaining all behaviour in terms of
strategies for a limited game and that future research centres on the empirical
study of behaviour.
see C:\WINDOWS\Desktop\Lecture 14 Notes.htm
Two categories of
signals:
Assessment signals are correlated with fighting ability and
"resource holding potential." This type of signal is generally
considered "honest" because it would be difficult or impossible to
cheat (if in poorer condition could not exhibit signal of animal in better
condition). EXAMPLES include horns and antlers, peacock tails, roaring in red
deer
Conventional signals (e.g., badges) may or may not be honest depending
on the response of the receiver. Not costly to maintain nor used directly in
fighting. EXAMPLES include red winged blackbird epaulets and black badge on
throat of sparrow. In Harris' sparrows, dominance rank is closely related to
extent of black badge on throat. Most fighting occurs between males of roughly
equal sized badges so if display a signal of dominance they are more likely to
be challenged by similar individuals. If artificially enlarge badge, suffer
higher rates of attacks by conspecifics.
Why have both types of signals? Assessment signals should be used if
cost of assessment is low and costs of escalated fighting are high.
Conventional signals should replace or evolve instead of assessment if costs of
giving and receiving signals are low compared to those for assessment.
Causation - Tinbergen had in mind primarily the physiological
mechanisms present
Survival value - the effects
of behaviour rather than the causes, in helping an animal to survive. Also
labelled �function�.
Ontogeny - Tinbergen�s addition to Huxley. The
development of the behaviour in the animal, with a strong awareness of the
nature/nurture distinction.
Evolution - �the elucidation of the course evolution
must be assumed to have taken, and the unravelling of its dynamics� (p.428).
The term phylogeny has also been used, although sometimes in different senses
from Tinbergen�s meaning.
Mayr (1963) for example stresses the dichotomy between
functional and evolutionary biology, using the terms �proximate cause� and
�how� questions in relation to functional biology and �ultimate cause� and
�why� questions in relation to evolutionary biology. This of itself suggests a division between
questions of causation and ontogeny on the one hand, and survival value and
evolution on the other. Further partitions, for example by individual vs.
population, or immediate cause vs. antecedent origin, have also been made
(especially by Hailman). Dewsbury�s sees these classifications as being
unnecessary at best, and dangerously misleading at worst by introducing
artificial divisions in what should be a continuum, and suggesting a hierarchy
where all parts should have equal regard. In particular, the ultimate/proximate
distinction can have the misleading implication that some causes are �merely�
the immediately observable mechanisms, while others are the more important,
�ultimate� explanations.
Dewsbury�s proposed reformulation:
organisms will be able to recognise kin for one of two reasons. Either an ability to recognise kin is selectively advantageous, or else the ability has developed as a by-product of some other system - i.e. species, group or individual recognition systems
e.g. Grafen (1990) argues convincingly that in the majority of cases where kin recognition has been reported, the ability to do so has arisen as a result of some other function and so should not be termed kin recognition as such; indeed, in many of these cases he argues that the demonstrated ability to recognise kin is not selectively advantageous or even of any functional value at all.
What cannot be disputed is the widespread ability of many animals to differentiate between kin and non-kin
separate question whether animals are doing so through learned familiarity or through self-referent matching to some genetically-influenced signal
Porter et al.
(1985) - humans: Mothers & offspring matches at greater than chances
levels. Could not reliably match husbands and wives. Capable of recognising
close biological kin through olfactory cues alone.
Manning et al. (1992) - house mice (Mus musculus domesticus) - communal nests, nurse other�s pups indiscriminately; kinship theory - should form nests with relatives to minimise exploitation and increase inclusive fitness? Females prefer communal nesting partners that share allelic forms of mhc genes. Grafen�s criticisms (1992).
Eklund (1997), in a study on mating preferences in wild house mice, concluded that fostering had little or no effect on MHC-nased mate preferences, and no evidence suggested that the MHC was used to avoid inbreeding; �inbreeding avoidance does not seem a strong mechanism for maintaining MHC diversity.�
Mechanism:�.
Visual (predisposition to know what you look like - self-referent?! Or else
familiarity with early group members. Non-chemical cue systems tend to involve
familiarity/learning), smell, secondary cues (burrows, e.g. bank swallows, then
auditory when they leave), something more direct (somatic cell incompatibility,
Botryllus Schlosseri - ascidian
(primitive chordate) - Grosberg & Quinn 1986)�. Grafen: must be released by
genes; visual cues can�t do that (?). Genetic component in smells does come
from the immune system. Phenotype will not express genotype unambiguously�.must
be highly polymorphic�.immune loci preadapted - challenge of pathogens,
self/non-self�.Crampton and Hurst[10] comment�tissue grafts in non-colonials�
(not vertebrates, echinoderms and a few other obile, non-colonial mobile
invertebrates)
Blaustein suggests
four possible mechanisms for recognition[1][11]:
based on spatial distance
based on familiarity, prior association
based on phenotype matching
based on action of recognition alleles
Functional requirement: differential behaviour w.r.t. relatedness. Cue, detection system, using system.
Grafen draws a clear
distinction between the different questions:
o
Is an ability
to differentiate between kin and non-kin present?
o
Is this
ability used by the organism for the purpose of distinguishing between kin and
non-kin?
o
Does such use
have a selective advantage?
o
Was this
selective advantage responsible for its origins and use?
It is important to recognise that there is a continuum from recognition of individual differences to recognition of kin, of members of different but conspecific populations, and on to recognition of specific differences involving reproductive isolating mechanisms. Thus principles applicable at one level are likely to be relevant at others (Fletcher & Michener 1987)
clear that many organisms can differentiate kin from non-kin, that this is achieved through a variety of mechanisms (learned behaviour, a combination of smell-sense and sensory adaptation, somatic cell incompatibility etc.), and that it many cases this can be of selective benefit both through increasing inclusive fitness and by reducing the risks of inbreeding.
Why? Improving
inclusive fitness - genuine altruism, avoiding excessive inbreeding but
maintaining slight relatedness. rb-c > 0. Trade-off between inbreeding
costs and loss of shared genes.
Can they? Apparently
yes. Can apparently discriminate - e.g. B. Schlosseri, sweat bees, tadpoles,
hamsters�. Humans obviously
How? Blaustein�s
list:
Remember: are these
cases on kin recognition or is that
aspect incidental?
see
��\dolphins\Brains, Behaviour and Intelligence in CetaceansBR (Whales, Dolphins
and Porpoises).htm�
Table
1. Approximate brain weights and body weights of some
mammals,
in
order of brain weight.
Species������������� Brain Weight�� Body Weight
�������������������� (approx.)����� (approx.)
�������������������� grams��������� tonnes
sperm
whale (male)���� 7,820� ������37.00
African
elephant������ 7,500������� 5.00
fin
whale������������� 6,930������� 90.00
killer
whale���������� 5,620������� 6.00
bottlenose
dolphin���� 1,600������� 0.17
human����������������� 1,500������� 0.07
cow������������������� 500��������� 0.6
But as the species with the biggest brains also tend to be the ones with the
biggest bodies, it might be that large animals just need larger brains to
control and maintain their larger bodies. Even when we talk of
"intelligence" in a general way, we mean something more than the sum
of body control systems. A simple way to make allowance for different body
weights is to express brain weight as a percentage of body weight (Table 2).
Table
2. Approximate brain weights as a percentage of
approximate
body weights of some mammals.
Species������������� Brain Weight as % of Body Weight
human��������������� 2.10
bottlenose
dolphin�� 0.94
African
elephant���� 0.15
killer
whale�������� 0.09
cow����������������� 0.08
sperm
whale (male)�� 0.02
fin
whale����������� 0.01
Macphail (1982) describes the experiments and species (chimpanzee, gorilla, bottlenose dolphin, California sea lion, pigeon) concerned, and argues that such performances to date are better described as ordering responses sequentially for reward, rather than as real steps on the road to language. He also puts forward an interesting interpretation of the human capacity for problem solving, which is quite beyond the capacity of any non-human. If humans solve problems, directly or indirectly, with the aid of language, the superiority of humans in problem solving might simply reflect the possession of language, and the capacity for language, in turn might be a species-specific specialisation, independent of general "intelligence".
Dolphin brains are relatively large, but again there are reasons for questioning the assumption that brain size is related to "intelligence". Crick and Mitchison's (1983) theory of the function of dream sleep may provide an alternative explanation for such anomalously large brains. They propose that rapid-eye-movement sleep (REM or dream or paradoxical sleep) acts to remove undesirable interactions in networks of cells in the cerebral cortex. They call this process, which is the opposite of learning, but different from forgetting, "reverse learning". Animals which cannot use this system need another way to avoid overloading the neural network, for example by having bigger brains. The spiny anteater and dolphins are the only mammals so far tested which do not have REM sleep (Allison, Van Twyner and Goff, 1972; Mukhametov, 1984) - and they also have disproportionally large brains. So, following this line of reasoning, dolphins and spiny anteaters would have to have big brains because they cannot dream.
Friendliness and helpfulness towards people are often discussed, but are we flattering ourselves in believing that the animals really "intended" to help? For perhaps obvious reasons we hear less of unhelpful behaviour, but there are well- documented cases. Many species of wild animals have been tamed or habituated to humans. Sometimes such animals become a danger to themselves or to people. Even tamed wild dolphins can become a considerable nuisance (for example setting boats adrift by pulIing up anchors) and sometimes dangerous. Instancesof "friendly" dolphins attacking swimmers (apparently unprovoked) are well documented, as are instances of swimmers being pushed out to sea, "abducted" or prevented from re-entering boats and other craft (e.g. Lockyer,1990).
Gaskin (1982) has concluded that there is abundant evidence that cetaceans communicate information about "what", "where" and "who". There is no substantive evidence that they transmit information about "when", "how" or "why". So with respect to Kipling's (1902) "six honest serving men" of learning and intellect, cetaceans appear to be three servants short.
kin selection vs recognition
are there questions just about why/examples, rather than how???