The first is for fun. See the miniature lasso that he used? It is easy to tie one out of a long blade of grass and catch lizards that way. The trick is that they've been adapted to ignore grass as a non-threat. So when they're about to be captured by a piece of grass, they literally can't see the grass. You have to remain far enough away, but you can bonk the lizard on the nose, pull your grass back, readjust and try, try again until you get it. It's a great trick to show kids!
The second is that no mechanism is needed to revert a temporarily selected trait. Francis Galton was the first to document regression to the mean in inheritance. The child of two tall parents tends to be not as tall. The child of two short parents tend to be not as short.
The reason this happens is that the recessives that influence the one parent's trait are often different from the recessives for the other parent. Therefore the child winds up expressing fewer recessives than either parent.
What this means is that if you select for a given complex trait in a given generation, those who survive will have that trait. But unless you also select for it in the next generation, the population will mostly revert back on its own. It is only regular selection, sustained over time, which causes the selected for trait to breed true.
> The reason this happens is that the recessives that influence the one parent's trait are often different from the recessives for the other parent. Therefore the child winds up expressing fewer recessives than either parent.
That's wrong. Regression to the mean has nothing to do with recessives (or dominance, or epistasis), and exists with purely additive effects (and everywhere in statistics, for that matter, anywhere that a correlation is not perfect). It is simply because the phenotype has r < 1 with the genotype (ie. the trait is not 100% genetically determined), so when you select the 'high' organisms, they are lucky on both genotype and 'environmental' variables. The environmental variable goes away, however, and only the higher genotype variable persists. Since that was only a part of why the parents were 'high', the highness is not as high.
> What this means is that if you select for a given complex trait in a given generation, those who survive will have that trait. But unless you also select for it in the next generation, the population will mostly revert back on its own. It is only regular selection, sustained over time, which causes the selected for trait to breed true.
This is also wrong. The effect of the selection is permanent and has already 'bred true' and will persist indefinitely unless undone by some other force. It's a ratchet, not a rubber band.
From the Stroud et al 2023 paper reviewed in the article:
Species’ phenotypic characteristics often remain unchanged over long stretches of geological time. Stabilizing selection—in which fitness is highest for intermediate phenotypes and lowest for the extremes—has been widely invoked as responsible for this pattern.
But, say Stroud et al:
However, even with an explosion of microevolutionary field studies over the past four decades, evidence for persistent stabilizing selection driving long-term stasis is lacking.
Stround & co's answer:
Here, by directly measuring natural selection in the wild, we identified a complex community-wide fitness surface in which four Anolis lizard species each occupy a distinct fitness peak close to their mean phenotype. The presence of local fitness optima within species, and fitness valleys between species, presents a barrier to adaptive evolutionary change and acts to maintain species differences through time.
Indeed. As I said, "The effect of the selection is permanent and has already 'bred true' and will persist indefinitely unless undone by some other force."
I gotta say, I don't quite grok your statement. Evolutionary biologists say selection effects are high in the very short-term, but very low in the long term - many creatures are basically the same for millions of years. That's the 'paradox of stability'.
I don't understand how your statement explains the long-term stability part. From the Quanta article:
With enough time, even the tiniest tugs should yield a measurable shift in an organism’s observable characteristics. If the beak-size changes the Grants observed [over a few years] continued over millennia, back-of-the-envelope calculations predicted some extreme phenomena, Pennell said. “You’d expect finches that were, like, 40 kilograms. This just makes no sense.”
How does your theory not produce 40 kg finches? And in fairness, I sure don't have a theory other than what I read, so it's asked in the spirit of curioisty.
I don't understand your confusion. I thought the Quanta article did a perfectly reasonable job of explaining the difference between stabilizing selection for a single fixed, permanent, unaltered optimal phenotype, and having simply a bit of selection randomly different each year which nets out to no change.
Lots of random changes average out to nothing, but each change was still a change. If you ratchet something a little bit one time, then ratchet it the other way randomly a bit, etc, it is still ratcheting each time, but it's not going to produce 40 consecutive clockwise spins.
I guess my interpretation is that the biologists say it does not average out and we would end up with 40 kg finches, and that another mechanism is needed to explain it.
They used to think that mechanism was stabilizing selection, "fitness is highest for intermediate phenotypes and lowest for the extremes", but the evidence doesn't support it.
As far as I can tell, the article omits the mechanism. Stroud's paper describes it thusly, though to me it doesn't seem so different from stabilizing selection, other than maybe the variations aren't smoothed out as immediately (from the abstract; in the body Stroud describes other theories as valid, but I haven't had time to sort it out):
The presence of local fitness optima within species, and fitness valleys between species, presents a barrier to adaptive evolutionary change and acts to maintain species differences through time. However, instead of continuously operating stabilizing selection, we found that species were maintained on these peaks by the combination of many independent periods among which selection fluctuated in form, strength, direction, or existence and in which stabilizing selection rarely occurred.
I'm getting the impression that the significance of Stroud's paper might be less the mechanism, and more that stability occurs on 'milli-scales' (my term), the short period of Stroud's data, in contrast to high variability on micro-scales (as shown in the Grants' finch research) and the well-known stability on mega-scales (millions of years). The article returns many times to the value of Stroud's long-term, high-quality data collection.
Because he had followed four species for three generations, he was able to show that a long-term pattern of stasis could emerge from such short-term fluctuating selection.
But I don't see how that differs from the Grants's finch data, which was over more time and was at least two generations, and I think more. Also they say:
Fully resolving the paradox will require scientists to study time spans between macro- and microevolution, Porto said — on the scale of tens, hundreds or thousands of years. They need to find a sweet spot that is long enough to allow both change and stasis to emerge, he said, although at the moment biologists don’t have a long enough data set to draw from.
Anyway, I am happy to keep talking, but I need to read something else today - something that I get paid to read! :)
height is controlled by thousands of genes each of which provides a small effect, and the relationships between those genes are non-trivial. i don't think concepts from mendelian inheritence (like recessive) are super-meaningful in cases like this.
Btw, doesn't Mendelian inheritance only really work for species with two copies of each chromosome?
Lots of plants have many more copies. I guess Mendel got lucky that he investigated plants that were simple enough. (And from what I've read, his research assistants actually cheated him: the numbers he got were too close to the theoretical averages, the variances were much smaller than math would predict. His assistants made the numbers come out like the boss expected them.)
This plays slightly into the slow climate change vs fast climate change issue. If climate changes slowly enough, this allows enough generations to build up the new traits needed for survival. Too fast, nothing lasts to evolve at all.
That makes me think of how--IIRC--amphibian genomes contain coding for a lot of different proteins that have about the same function but are stable/effective at different temperatures, especially during early development in the egg. (In contrast, mammals invest in genes that ensure the egg's external temperature is reliable.)
(See also: Orthologous proteins, Heat shock proteins.)
That'll work the cruelty right out of them and set them up for a promising life post nuclear apocalypse. Lizards will bioaccumulate less fallout from the food chain than larger animals so they'll be a safer source of protein.
I find myself thinking of dog breeds: Over the last ~150 years, selective breeding has led to an explosion of phenotypes, but there's a limit to how much humans can take credit--or perhaps blame--for all the results.
Much of what we've done is to evoke particular blends and combinations of a great many individual genes that already existed throughout in the species.
P.S.: IANADogGeneticist so these numbers are totally made-up, but imagine an island population of 1,000 "average dogs" or mutts that look the same. They collectively bear 20 distinct genes that promote Long Noses, and 15 distinct genes that promote Short Noses, and due to what chromosomes they're on, each dog might have a random ~4 from the Long Nose set and ~3 from the Short Nose set. On average you'll keep getting similar looking individuals, until some sinister aliens arrive and begin... The Pug Project, leading to a bunch of poor lapdogs with 8-10 Short Nose genes and 0-1 Long Nose ones.
> I find myself thinking of dog breeds: Over the last ~150 years, selective breeding has led to an explosion of phenotypes, but there's a limit to how much humans can take credit--or perhaps blame---for all the results.
Dogs as a species wouldn't exist if humans didn't selectively breed a now extinct lineage of wolves, which to me means that humanity is to blame for all of it.
It is really not. All we have are two popular hypotheses, which may even be true at the same time. It is clear the self-domestication pathway would not have worked without humans allowing it to happen. This quickly devolves into semantics. It is however not up for debate that humans have been breeding dogs/wolves for millennia.
> resulting in the subspecies Canis lupus familiaris.
Put the wrong two zoologists in the same room and if you're unlucky they'll bore you all day debating whether domesticated dogs are a distinct taxon. Just say species without specifying and you'll sidestep the issue. If you want to call them Canis lupus familiaris rather than Canis familiaris, then sure! I guess that's why I meant!
It quickly devolves into you being wrong that humans get all the blame for it. That's the point and the rest is deflection and strawmen like "It is however not up for debate that humans have been breeding dogs/wolves for millennia" -- duh.
I long ago encountered a suggestion that the great behavioral malleability of historical dog breeding resulted from taking a common ancestral hunting sequence, and tweaking dials on the steps. So for fudged illustration, given a sequence of <chase-it-down, bite-hard-on-neck, shake-until-dead, carry-back, bury-it>, turning the "bite" step down, and the "shake" and "bury" steps way down, gets you a retriever breed. This malleability then misled people about the broader flexibility and efficacy of selective breeding. Like someone going "Look what I did today tweaking a configuration file! ... Programming is so easy!". I thought the paper might make for a fun web interactive (but IIRC had difficulty refinding it).
Selective breeding has some huge and obvious morality issues around it, but it is fun to think about possibilities. Hyper intelligent dogs? 10 foot tall humans?
James T. Stroud, et. al. "Fluctuating selection maintains distinct species phenotypes in an ecological community in the wild". Proceeding of the National Academy of Sciences. 09 Oct 2023
Evolution couldn't work quickly if there wasn't a lot of small variance from generation to generation. It's exogenous fitness testing that maintains stasis, not lack of variance.
A long article for what ought not to have been a paradox, but great to have a solid study.
After reading the article it sounds absurd when you first think about it but leaves you with how else could it possible work after seeing the evidence.
I am no evolutionary biologist but one might look at rapid adaptation as an evolutionary strategy in itself. I have always felt humans did this by using intelligence. Does anyone know of other mammals that can live in such a large range of environments without physically adapting?
Humans have physically adapted though. Skin and eye color, subcutaneous fat, underwater vision, high-altitude adaptations, resting body temperature, nose shape and nostril size, etc. We just haven't had populations separated long enough to be unable to breed or be considered separate species.
There are certainly other ways it could work. The variability that Stroud was surprised to see might not have occurred. Or it might have resulted in persistent change rather than stability.
Variability varies among species, and high and low variability can both be seen as evolutionary strategies that are suitable for different ecologies.
Humans using intelligence is not biological (genetic) evolution, it's mimetic evolution. But we employ other sorts of physical but non-biological mechanisms to adapt, like clothing and housing. Dogs accompany us and can survive in many environments by living in our homes and depending on us for food. Also lice, bedbugs, etc.
I don't get point of your question, but why limit it to mammals? Spiders, bacteria, and viruses are everywhere with few environment-specific adaptations.
Maybe there are some aspects of genetics that are like dials and can be rapidly tuned, as opposed to more radical changes that would significantly alter the body plan and perhaps are subject to more gradual changes.
I'm with Hendry, the guy who said the paradox is an illusion. I figured it was rapid change within a stable envelope within a few sentences of starting the article, but I waited for the other shoe to drop, because there must be something trickier if people have spent so much time on it ... It turns out the other shoe is "but now we have data about the details". Which is good, but nowhere near as dramatic as they make it out to be. You can see the outline of the solution just by thinking about the question for a second.
Yeah, I said the data was a good thing. But your critique circularly assumes there was a paradox to begin with. It's like being mystified that an object oscillating on a spring stays close to the spring's anchor point. A question as to mechanism, sure, but not a paradox.
The first is for fun. See the miniature lasso that he used? It is easy to tie one out of a long blade of grass and catch lizards that way. The trick is that they've been adapted to ignore grass as a non-threat. So when they're about to be captured by a piece of grass, they literally can't see the grass. You have to remain far enough away, but you can bonk the lizard on the nose, pull your grass back, readjust and try, try again until you get it. It's a great trick to show kids!
The second is that no mechanism is needed to revert a temporarily selected trait. Francis Galton was the first to document regression to the mean in inheritance. The child of two tall parents tends to be not as tall. The child of two short parents tend to be not as short.
The reason this happens is that the recessives that influence the one parent's trait are often different from the recessives for the other parent. Therefore the child winds up expressing fewer recessives than either parent.
What this means is that if you select for a given complex trait in a given generation, those who survive will have that trait. But unless you also select for it in the next generation, the population will mostly revert back on its own. It is only regular selection, sustained over time, which causes the selected for trait to breed true.
That's wrong. Regression to the mean has nothing to do with recessives (or dominance, or epistasis), and exists with purely additive effects (and everywhere in statistics, for that matter, anywhere that a correlation is not perfect). It is simply because the phenotype has r < 1 with the genotype (ie. the trait is not 100% genetically determined), so when you select the 'high' organisms, they are lucky on both genotype and 'environmental' variables. The environmental variable goes away, however, and only the higher genotype variable persists. Since that was only a part of why the parents were 'high', the highness is not as high.
> What this means is that if you select for a given complex trait in a given generation, those who survive will have that trait. But unless you also select for it in the next generation, the population will mostly revert back on its own. It is only regular selection, sustained over time, which causes the selected for trait to breed true.
This is also wrong. The effect of the selection is permanent and has already 'bred true' and will persist indefinitely unless undone by some other force. It's a ratchet, not a rubber band.
Species’ phenotypic characteristics often remain unchanged over long stretches of geological time. Stabilizing selection—in which fitness is highest for intermediate phenotypes and lowest for the extremes—has been widely invoked as responsible for this pattern.
But, say Stroud et al:
However, even with an explosion of microevolutionary field studies over the past four decades, evidence for persistent stabilizing selection driving long-term stasis is lacking.
Stround & co's answer:
Here, by directly measuring natural selection in the wild, we identified a complex community-wide fitness surface in which four Anolis lizard species each occupy a distinct fitness peak close to their mean phenotype. The presence of local fitness optima within species, and fitness valleys between species, presents a barrier to adaptive evolutionary change and acts to maintain species differences through time.
I don't understand how your statement explains the long-term stability part. From the Quanta article:
With enough time, even the tiniest tugs should yield a measurable shift in an organism’s observable characteristics. If the beak-size changes the Grants observed [over a few years] continued over millennia, back-of-the-envelope calculations predicted some extreme phenomena, Pennell said. “You’d expect finches that were, like, 40 kilograms. This just makes no sense.”
How does your theory not produce 40 kg finches? And in fairness, I sure don't have a theory other than what I read, so it's asked in the spirit of curioisty.
Lots of random changes average out to nothing, but each change was still a change. If you ratchet something a little bit one time, then ratchet it the other way randomly a bit, etc, it is still ratcheting each time, but it's not going to produce 40 consecutive clockwise spins.
They used to think that mechanism was stabilizing selection, "fitness is highest for intermediate phenotypes and lowest for the extremes", but the evidence doesn't support it.
As far as I can tell, the article omits the mechanism. Stroud's paper describes it thusly, though to me it doesn't seem so different from stabilizing selection, other than maybe the variations aren't smoothed out as immediately (from the abstract; in the body Stroud describes other theories as valid, but I haven't had time to sort it out):
The presence of local fitness optima within species, and fitness valleys between species, presents a barrier to adaptive evolutionary change and acts to maintain species differences through time. However, instead of continuously operating stabilizing selection, we found that species were maintained on these peaks by the combination of many independent periods among which selection fluctuated in form, strength, direction, or existence and in which stabilizing selection rarely occurred.
I'm getting the impression that the significance of Stroud's paper might be less the mechanism, and more that stability occurs on 'milli-scales' (my term), the short period of Stroud's data, in contrast to high variability on micro-scales (as shown in the Grants' finch research) and the well-known stability on mega-scales (millions of years). The article returns many times to the value of Stroud's long-term, high-quality data collection.
Because he had followed four species for three generations, he was able to show that a long-term pattern of stasis could emerge from such short-term fluctuating selection.
But I don't see how that differs from the Grants's finch data, which was over more time and was at least two generations, and I think more. Also they say:
Fully resolving the paradox will require scientists to study time spans between macro- and microevolution, Porto said — on the scale of tens, hundreds or thousands of years. They need to find a sweet spot that is long enough to allow both change and stasis to emerge, he said, although at the moment biologists don’t have a long enough data set to draw from.
Anyway, I am happy to keep talking, but I need to read something else today - something that I get paid to read! :)
Lots of plants have many more copies. I guess Mendel got lucky that he investigated plants that were simple enough. (And from what I've read, his research assistants actually cheated him: the numbers he got were too close to the theoretical averages, the variances were much smaller than math would predict. His assistants made the numbers come out like the boss expected them.)
(See also: Orthologous proteins, Heat shock proteins.)
Kids are cruel :( showing them how to hunt lizards is likely to lead to their torture and/or death
"You kill it, you eat it!"
That'll work the cruelty right out of them and set them up for a promising life post nuclear apocalypse. Lizards will bioaccumulate less fallout from the food chain than larger animals so they'll be a safer source of protein.
Much of what we've done is to evoke particular blends and combinations of a great many individual genes that already existed throughout in the species.
P.S.: IANADogGeneticist so these numbers are totally made-up, but imagine an island population of 1,000 "average dogs" or mutts that look the same. They collectively bear 20 distinct genes that promote Long Noses, and 15 distinct genes that promote Short Noses, and due to what chromosomes they're on, each dog might have a random ~4 from the Long Nose set and ~3 from the Short Nose set. On average you'll keep getting similar looking individuals, until some sinister aliens arrive and begin... The Pug Project, leading to a bunch of poor lapdogs with 8-10 Short Nose genes and 0-1 Long Nose ones.
Dogs as a species wouldn't exist if humans didn't selectively breed a now extinct lineage of wolves, which to me means that humanity is to blame for all of it.
It is really not. All we have are two popular hypotheses, which may even be true at the same time. It is clear the self-domestication pathway would not have worked without humans allowing it to happen. This quickly devolves into semantics. It is however not up for debate that humans have been breeding dogs/wolves for millennia.
> resulting in the subspecies Canis lupus familiaris.
Put the wrong two zoologists in the same room and if you're unlucky they'll bore you all day debating whether domesticated dogs are a distinct taxon. Just say species without specifying and you'll sidestep the issue. If you want to call them Canis lupus familiaris rather than Canis familiaris, then sure! I guess that's why I meant!
It quickly devolves into you being wrong that humans get all the blame for it. That's the point and the rest is deflection and strawmen like "It is however not up for debate that humans have been breeding dogs/wolves for millennia" -- duh.
I long ago encountered a suggestion that the great behavioral malleability of historical dog breeding resulted from taking a common ancestral hunting sequence, and tweaking dials on the steps. So for fudged illustration, given a sequence of <chase-it-down, bite-hard-on-neck, shake-until-dead, carry-back, bury-it>, turning the "bite" step down, and the "shake" and "bury" steps way down, gets you a retriever breed. This malleability then misled people about the broader flexibility and efficacy of selective breeding. Like someone going "Look what I did today tweaking a configuration file! ... Programming is so easy!". I thought the paper might make for a fun web interactive (but IIRC had difficulty refinding it).
Most of our science seems compatible with the idea that you can back-cross mammals, after all we do that in mice all the time.
James T. Stroud, et. al. "Fluctuating selection maintains distinct species phenotypes in an ecological community in the wild". Proceeding of the National Academy of Sciences. 09 Oct 2023
https://www.pnas.org/doi/10.1073/pnas.2222071120
Edit: also here:
https://www.jameststroud.com/uploads/2/6/1/3/26134722/stroud...
A long article for what ought not to have been a paradox, but great to have a solid study.
I am no evolutionary biologist but one might look at rapid adaptation as an evolutionary strategy in itself. I have always felt humans did this by using intelligence. Does anyone know of other mammals that can live in such a large range of environments without physically adapting?
Variability varies among species, and high and low variability can both be seen as evolutionary strategies that are suitable for different ecologies.
Humans using intelligence is not biological (genetic) evolution, it's mimetic evolution. But we employ other sorts of physical but non-biological mechanisms to adapt, like clothing and housing. Dogs accompany us and can survive in many environments by living in our homes and depending on us for food. Also lice, bedbugs, etc.
I don't get point of your question, but why limit it to mammals? Spiders, bacteria, and viruses are everywhere with few environment-specific adaptations.
> Why does natural selection appear to happen slowly on long timescales and quickly on short ones?
Compare this to Amara's Law: "We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run."
Just seems interesting. YMMV.
I wonder if Michael Levin's work ties into this.