Wednesday, January 7, 2015

The complex evolution of personality

So, apparently even sea anemones have personalities.  The idea that non-human animals can be measurably, say, bolder or shyer than others of their species may or may not be a surprising idea to you, perhaps depending on how many cats, dogs, horses, laboratory mice you have known.  But, scientists who study animal behavior are currently focusing on animal personality in a big way.  The BBC Radio 4 program Discovery discussed this the other day, and to us, the discussion raised some unintended points.

Hermit crab; Wikipedia

Presenter Adam Hart interviewed behavioral scientists studying personality in animals as diverse as songbirds and sea anemones. All agreed that variation is the norm.  Daniel Nettle, at the University of Newcastle, described five different dimensions to human personality: extroversion, neuroticism, agreeableness, conscientiousness, openness to experience.  They aren't all found in non-human animals, he said, though some seem to be, and all seem to be present in chimps.

Personality variation in the great tit, a small songbird, has been studied by many people (e.g., here and here).  For example, Samantha Patrick from the University of Gloucestershire described catching birds in the wild and releasing them into a room furnished with artificial trees, which they hadn't seen before.  Each bird's behavior upon first seeing the room is recorded, and its 'exploration score' calculated, to determine where the bird sits on a boldness/shyness range.  Fast explorers are more aggressive, and more willing to take risks than slow explorers.  And, when birds are artificially selected for parental aggression or calm, heritability of such personality traits is consistently around 50% -- that is, 50% of the behavior seems to have a genetic source.  But behavioral plasticity has been found to be common in great tits as well.

Great tit; Wikipedia, photo by Lviatour

But it's not just vertebrates that are being studied.  Mark Briffa at Plymouth University studies boldness and shyness in hermit crabs. He disturbs them by lifting them out of the water and turning them upside-down, leaves them thus for five seconds, and then replaces them in the water, and measures the time it takes for them to return to normal.  He then gives them a boldness rating.

Hart asked Briffa whether boldness in hermit crabs is at all equivalent to human extroversion.  That is, whether understanding hermit crab behavior give us any insights into human behavior.  Briffa's answer was that humans share an ancient neurobiology with other animals, including hermit crabs, and looking at that will help "simplify the problem" of human behavior.  Maybe.

So, there are conscientious crabs, those that spend a lot of time investigating empty snail shells as they make the decision about which one to move into next.  And there are bold hermit crabs, who seem to choose with little consideration.  There are evolutionary trade-offs to each of these behaviors, Briffa said.  The conscientious crabs get better shells, but waste a lot of time looking, while the bold crabs get iffier shells but saved time.  Time for what, exactly, it wasn't clear.  Can't be making more hermit crabs, because personality traits are not linked with fitness, and Briffa was not the only interviewee who said this about the animal they study.  That is, there's no individual reproductive advantage to being bold or shy, conscientious or not.  If there were, of course, there'd be a lot less variation in personality because it would have been selected out of the population in favor of the personality trait that led to more offspring.

How did these traits evolve?
This means that these traits aren't here because they were favored by natural selection, at least not in the present if these studies are any indication.  They could be here just by chance for reasons of ecology or population structure of some kind in the species' pasts.  Or, it might mean that there's a lot more plasticity in personality than is being reported, or identified by current methods, and that in some way plasticity is genetically enabled.  And indeed this would be expected, given that adaptability is so widespread that we've called it a fundamental principle of life.  A brain that can sense its circumstances, evaluate them, and plan responses may be what has evolved, but different brains, even if they were developed from the same genotype, might make different decisions.

However Hart, and interviewees, asserted that natural selection has favored a variety of personality types. But if no personality type has more offspring, gradually out reproducing the others, this can't be. It's only possible if group selection, natural selection that can see, and thus choose traits that benefit a group, is at work.

So, great tits can be shy or bold.  If an entire flock is bold, they are often on the move and able to locate new food sources, but they ignore each other, and that's bad for the cohesion of the group.  Shy birds stay together, but they don't move to new food sources, and that's bad for the health of the group.  A mix of bold and shy birds is ideal; the bold birds ensure that the flock moves to new food sources, and the shy birds follow.  Group selection would have favored a flock that includes a mix of personality traits, for the benefit of the flock rather than only a single personality trait, with its serious flaws.

But, this is a controversial issue.  Even Darwin, who himself addressed behavior including that in humans and our closer relatives, was rather mixed on this point.  The kind of mixed-flock just described could be a case of complex balanced polymorphism if the traits are genetically determined.  Too many bolds, bad for the group, too many shy, bad for the group.  The bold/shy genotypes' fitness is a function of the population in which they occur.  That would be standard genetic theory.

Group selection is coming back into fashion, being seen these days by all sorts of theoretical modelers and empirical investigators.  It has a checkered history.  The issue is that individuals shed their genes or not, and a genotype that engenders a particular behavior that is good for the group is fine -- so long as it's even better for the individuals with the genotype.  The reason this is contentious is that those strong Darwinians who are unhappy with any sort of resistance to pure individual selection argue that a genotype that favors the group cannot proliferate if that is at the relative expense of the individual with the genotype.  Otherwise, the group may do fine, but the genotype cannot become more relatively common over time.  At least not within the group.

The issues are quite mixed, and anthropomorphizing evolutionary modelers often say that an altruistic group-favoring gene variant will either be outcompeted, or that in effect it enables 'cheaters', without the good-guy genotype, to succeed at the good-guys' expense.  So the altruism-conferring variant loses out in the end.  If a group grows at the relative expense of other groups of the same species, the good-guy variants might overall increase in frequency (since the group without good guys disappears), but eventually, from a strongly deterministic Darwinian view of natural selection, the good-guy gene will get driven out.

How much does the work described above help explain behavior and the evolution of behavior in humans, where sociocultural factors clearly play a larger part in behavior than in non-humans?  The assumption, presumably, is that what's being explored here is the genetic aspect of behavior.  But, if personality is only 50% heritable in great tits, who don't share the extent of cultural influence on behavior that humans have, then it's hard to accept that we're getting at something that can be explained primarily by hard-wiring in humans.  At the very least, chance is playing a role comparable in strength to selection in determining group and individual success.  Of course, anthropomorphizing is difficult to avoid and it's difficult to tell when it's justified or not; indeed, cultural evolution has been described even in great tits, with the spread of learned behavior across a wide area.

Darwin wrote quite a lot about behavior, including altruism, aesthetics, personality and so on in Descent of Man, where he tried to show continuity between humans and other species.  His specific agenda, and even explicitly stated, was to displace religious creationism as an explanation for animal (and plant) diversity.  So he wanted humans to have traits that other animals have.  Several chapters deal with his ideas of these sorts of behavioral and communal sharing, including explanations of altruism.  He stressed behavioral continuity also in his Expression of the Emotions in Man and Animals.

Darwin was hand-waving much of the time when he did this.  And while he talked of selection, his main point was continuity and descent from common ancestry.  This is very different from observing what other species do, assessing when or how or if a trait is actually genetic in a way simple enough for selection to screen it at the gene level, and determining if, in fact, variants of the trait affect fitness.  Short term observations and risk of things like anthropomorphizing, and the likelihood that behavioral patterns for individuals may vary during their lives or in different circumstances, make the area difficult to study definitively in evolutionary terms.

But one thing is definitively clear:  animals do behave in variable ways, and that's fascinating enough.


Holly Dunsworth said...

Interesting how it's called "personality" in non-persons. Why not individuality? I guess that wouldn't bring to mind just behavior/attitude.

With the birds in the room of new trees... did they use a new room for each bird? Because wouldn't the scent of previous birds under testing influence how birds that came after them explored the trees? This is mostly rhetorical since I could just look into it myself and don't expect you to know the answer without doing the same. I'm sharing because I always wonder about residual odors with these bird (and other animal behavior) experiments. They're not being controlled for as far as my amateur eyes/nose can tell.

And, also, I'd like to offer that it's not necessarily "cultural" if it's learned behavior. Many wouldn't consider learned behavior to be cultural unless it's due to shared knowledge via language, restricting "culture" to humans and our ancestors with the capacity. So, for personality/individuality to be explained in crabs beyond the simplistic it's-hard-wired-in-the-genes story... one could say that environment (including other crabs) is conditioning particular crabs with particular genomes in particular ways resulting in particularities.

Ken Weiss said...

Good points, Holly. But so politically non-correct--denying poor birds and crabs their 'personhood'! You must not have been around in the '60s and '70s when such an attitude would simply not have been tolerated.....

Anne Buchanan said...

Holly, yes, many questions. How repeatable are the results, not over minutes or days, but weeks or months or years, depending on species lifespan? Cross sectional testing is surely less likely to pick up plasticity. Even if repeated several times in a given experiment, this isn't getting at possible effects of experience. Some do report plasticity, so it's not clear to me that these personality (bird/crabality) scores are all that meaningful.

The point seems to be to throw human behavior in the same mix as all other behaviors, and look for universals. A la Darwin. I suspect those universals would be genetics, but didn't get deep enough into the literature to determine that, as my issue here was with the idea that personality is naturally selected, even though there seem to be no fitness benefits of personality.

And, lots of semantic issues. What's 'intelligence', 'culture', instinct vs learned behavior, etc.?

Holly Dunsworth said...

Oh wow. I hadn't even thought of the plasticity problem with research design and analysis. Wow. Oh, I know! Watch how the hermit crabs vote while they're menstruating compared to how they vote while they're attractive. ;)

Anne Buchanan said...

Yes, or in the presence of hermit crabs with deep voices?

Ken Weiss said...

And what about the crab or (sensually and suggestively titled, 'crab dip') in bars and restaurants. They are probably their for hook-up purposes.

DG said...

Hi AB,
Nice post.I hope you have the time for some questions.

“A brain that can sense its circumstances, evaluate them, and plan responses may be what has evolved, but different brains, even if they were developed from the same genotype, might make different decisions.”
Different decisions mean that some decisions will be better than others. Better decisions should mean more descendants. Increasing the complexity of the circumstances should increase the value of making better decisions.

“ the good-guy variants might overall increase in frequency (since the group without good guys disappears), but eventually, from a strongly deterministic Darwinian view of natural selection, the good-guy gene will get driven out.”
This is what I cannot understand. Let’s say the groups with 10% good guys deselect the groups with only 1% good guys, wouldn’t the groups stay at 10% good guys until something acted on the groups to change that.


Anne Buchanan said...

Thanks, DG. No, the idea there is that the brains that make better decisions are doing it by chance. They are all equal, so there's nothing about making those decisions that natural selection can see or choose. Even if the brains making better decisions are making decisions that affect survival, or even fitness, because even if so, what survives is everything that comes along with that brain, but not better-decision-making-powers. That was just the luck of the draw, so there's nothing to differentially evolve.

Question #2 later...

Ken Weiss said...

Response to DG's second point:
The answers are ‘ceteris paribus’, that is, greatly oversimplified and perhaps even fanciful population genetics theory, that has rather little specific support empirically but makes sense in principle.

First, a new group-selected allele arises by mutation and has frequency 1/2N in a diploid population of N individuals. If it helps the group as a whole to expand, which is what ‘group selection’ means, then on average its frequency relative to other existing alleles at the relevant gene doesn’t change—if its group exterminates other groups of the same species. Indeed, if the group expands by incorporating parts of otherwise rival groups of the same species, the index variant’s frequency is actually diluted—decreases! Otherwise, as a neutral variant, its likelihood of becoming fixed is 1/2N, that is, it’s almost doomed to extinction as any new neutral variant is, just because its fate is determined by chance alone. If its presence allows ‘cheater’ variants to arise and expand under the group-selection allele’s shadow then again the index variant decreases in frequency.

The real controversy about group selection has been that the allele has to be altruistic—has to act for the good of the group to its own detriment. This is where Hamilton’s Rule comes into play: if the self-sacrifice helps others with the same allele to increase by more than the inverse of the kinship between them, relative to the index allele’s loss due to its altruism, then that allele can still increase in frequency. It’s helping the group only in appearance, but really it’s helping its relatives. Of course, this doesn’t work in the beginning since at that point there are no other copies of the allele in any relatives. And Hamilton’s Rule has not really proven very apt in empirical studies—it’s a mathematical principle. And complex traits are rarely going to be that single-gene based.

So, the whole group selection issue is far more complex than the very oversimplified, usually theoretically driven arguments seen about it. Now, of course, it’s fashionable. In reality, I think the whole argument is rather inane. One can construct all sorts of scenarios for how socially relevant genotypes increase in frequency (since, clearly, socially cooperative groups and species do exist, and yet to the extent this has a genetic basis it must be consistent with successful reproduction).

I hope this rather long attempt to simplify what isn’t simple is helpful in response to your question!

Holly Dunsworth said...

"We present four experiments showing that subtle cues to social status (i.e., prestige and reputation in the eyes of others) modulate prosocial orientation. The experiments found that individuals who experienced low status showed more communal and prosocial behavior, and endorsed more egalitarian life goals and values compared with those who experienced high status." Article in current PNAS: Social status modulates prosocial behavior and egalitarianism in preschool children and adults

DG said...

Thanks Ken.

I promised myself that I would not be confounded by the true if conditions are true argument again. It was only last month that you pointed that out to me. I am embarrassed to say how many times over the years that I have tripped myself by not recognizing it. I went back to AB’s post and there it is, “from a strongly deterministic Darwinian view.”

HD. I only read the abstract.
Did they mix the groups or were they kept all boys, all girls, etc.? It seems to me that it would be very important to control for that.

Also, did they vary the status (let individuals move up and down) to see if the behavior followed the status or the person?

Does that mean that serfs cooperate and kings do as they please?

DG said...

“Many wouldn't consider learned behavior to be cultural unless it's due to shared knowledge via language, restricting "culture" to humans and our ancestors with the capacity.”
Those experiments that show very young children readily showing an adult how to open a box or door when the adult feigns inability are said to show altruism. I wonder about that because maybe they just want be imitated, that is, they want to teach.
Do chimps “teach” or do they just learn by watching?

Ken Weiss said...

The world of life sciences is populated by professionals who routinely accept the if-true to mean is-true, who knowingly elide the complexities and subtleties, and who accept story-telling. It's more convenient than thinking about what's really going on and how elusive that can be.

Sush said...

Thanks, Nice post.