I would strongly recommend reading Species Concepts in Biology for a history and explanation of the current system(s). In summary, there are multiple joints at which you can carve reality, and which joints you choose to carve depends on why you are carving reality in the first place.
As someone who studies viruses of bacteria, how are you supposed to apply the Biological Species Concept to asexual organisms anyways? Are there no bacterial “species”? But we need a way to categorize and talk about them, even if they’re not a species under the BSC. Even if this will group bacteria together whose genome lengths can vary by an entire order of magnitude.
Fundamentally, the Species Concept is the same problem. We need a way to categorize and talk about animals, so we find ways to categorize them based on how useful the categories are. Birds are feathered reptiles. Reptiles are non-feathered reptiles. We form these categories not because they carve reality at the phylogenetic joints but because these categories are useful. Fish are fish. Trees are trees.
(Viral species are even worse. Linear phylogenetic inheritance is thrown out the window. Everything is modular and recombines with each other.)
I don’t really follow the IUCN or like the focus on “biodiversity” for its own sake (just expose them to mutagens if you want more genetic diversity) (I also have philosophical issues with “conservation”—what, exactly, are you conserving and why is that time more worth conserving than any other time?) I’m probably informed by my study: in my opinion everything is a bag of genes, and genes constantly flow between them; species aren’t real and are just useful lines to draw in a continuum of gene flow, which constantly changes over time.
all human populations are biologically capable of mutual gene flow, so all humans would be the same species.
But can you verify it? Do you have experimental evidence that a human from location A can form fertile offspring with a human from remote location B? That’s why geographical isolation is used to define species; because they (mostly) don’t overlap, we don’t have evidence if they can or cannot interbreed. Only when we have evidence of hybridization can we start discussing whether to collapse them into the same species (ring species notwithstanding).
Also, I don’t know nearly enough about plant hybridization, but I’m pretty sure that plants can hybridize extremely easily and if we applied only an “interbreeding” species concept the categories would collapse into uselessness.
I have many thoughts here. First I would say that I totally agree that the species concept stops making any sense as a strict, discrete category the second horizontal gene transfer / true asexual reproduction enter the picture. I should probably have said this at the beginning of the post so that people who work in viruses and bacteria (as you seem to!) knew that my argument didn’t apply to organisms capable of that.
In my experience, however, microbiologists are too quick to generalize from their part of the field to all parts of it. It is 100% true and defensible that “everything is a bag of genes, and genes constantly flow between them; species aren’t real and are just useful lines to draw in a continuum of gene flow, which constantly changes over time” if you’re talking about bacteria. But genes do not constantly flow between humans and chimpanzees via HGT. When you’re talking about most of the animals that the IUCN cares about (for better or worse), you’re talking about sexually reproducing organisms that (imo, once they SPECIATE) stop sharing genes. You cannot treat them the same way you would a virus.
I also totally agree with this statement: “We form these categories not because they carve reality at the phylogenetic joints but because these categories are useful. Fish are fish. Trees are trees.” I don’t mean to propose that everyone stop using the words fish, or tree, even though they’re not monophyletic. But there is (and should be!) a different between technical scientific vocabulary and common usage. I don’t insist people call bananas “berries” because, in common usage, they’re not. But botanically speaking, we have a definition for “berry” that we apply consistently, no matter what the common usage is, and this definition makes bananas a berry. I believe that the technical definition of “species” should be similarly transparent and consistent. I agree that “there are multiple joints at which you can carve reality, and which joints you choose to carve depends on why you are carving reality in the first place.” But as I hope the essay above illustrates, the Biological Species Concept as it is usually formulated cannot hope to be consistently applied to relevant organisms (ie. sexually reproducing macroorganisms incapable of HGT etc)
In terms of this question asking about all human populations being able to reproduce: ”But can you verify it? Do you have experimental evidence that a human from location A can form fertile offspring with a human from remote location B?” There have been many many recorded instances of human populations that have been long isolated being perfectly capable of being able to reproduce with other populations. To this day, there has NEVER been a recorded instance of the opposite. We know that Homo Sapiens could produce fertile offspring with Neenderthals, MUCH more different from us than any modern humans.
It would be very strange to lend much credence to the idea that not all human populations could produce fertile offspring. The burden of proof would clearly be on the person trying to say otherwise.
In terms of your philosophical problems with conservation: “(I also have philosophical issues with “conservation”—what, exactly, are you conserving and why is that time more worth conserving than any other time?)” These are all worthy questions that bear consideration. But noting that a practitioner in any given field will be required to make contestable value judgements is not really an argument against that field. The same could be said of doctors, politicians, lawyers, journalists, etc.
But can you verify it? Do you have experimental evidence that a human from location A can form fertile offspring with a human from remote location B? That’s why geographical isolation is used to define species; because they (mostly) don’t overlap, we don’t have evidence if they can or cannot interbreed.
Do you think that we should categorize e.g. the Sentinelese people as a separate species due to lack of evidence of hybridization? That seems like a strange thing to do given that (I assume) we’re all pretty confident they can still interbreed.
The Sentinelese has been reproductively isolated for probably less than twenty generations, which can be sufficient for speciation (given strong selection pressures), but relatively unlikely. So I don’t think it’s necessary to assume that speciation has occurred.
(aside, estimates of Sentinelese isolation are full of wild speculation and misinformation. We have no genetic samples from North Sentinel Island, which means we have no evidence. All estimates are speculation, with an upper bound at a few thousand years due to shared ethnographic artifacts. Missing artifacts (metalworking, etc.) may or may not help establish a lower bound, but is confounded by natural resources available on the island. Also, going back more than tens of generations would imply a high amount of inbreeding given the limited carrying capacity of the island, which could theoretically be sustained with strong selection.)
--
Imagine a forest with one species of frog. Then a road is built through the forest, separating the forest in two. The day after the road is built, I discover that there are now two reproductively isolated groups of frogs! Are these now different species?
(under most species frameworks) No! That’s preposterous. The two populations have been isolated for less than a day. Sure, they may at some point in the future diverge (and now natural selection is acting on the two groups differently), but at this point the two species are genetically identical.
--
But now imagine you stumble across on species of frog in one forest, and a similar-looking frog in a different forest. The two population are reproductively isolated. Are they the same or different species?
Well, you might ask how long they’ve been reproductively isolated for, since if the forests were connected yesterday then we converge on the same scenario as above. A week, a month, a year, a decade, a century, a millennium, a million years. At some time of separation we cannot reject [no speciation] out of hand, or, rather, number of generations, since that’s what matters in (genetic) speciation. And in nature, we can generally assume that the groups have been isolated from each other for thousands of generations (unless the barriers are human-made, or with recorded history, and because our genetic divergence estimation tools don’t work well for times less than that).
Can they produce fertile offspring? Well, no, not in nature, since that’s what reproductive isolation means. Maybe you can put two of them in a box and test it out manually, but you don’t have the funding to do that. And genetics don’t help too much in determining hybridization potential.
(an aside, what counts as producing fertile offspring? If they can produce fertile offspring via IVF, but never crossbreed if you cage them together because they don’t have the same mating calls or compatible mating apparatus or look funny to each other or sleep at different times of the day, does that count as producing fertile offspring? this is why “natural populations that are reproductively isolated” is added to the definition)
So you have two groups of frogs, isolated from each other for thousands of generations. Evolution is acting upon them as two different groups. But they look suspiciously similar. Do you call them different species?
I could go either way, and lean towards labelling them as different species until we have evidence to the contrary. Otherwise one of the forests might be bulldozed before we can gather the evidence.
Hi eniteris! I appreciate your long and detailed responses. Clearly you’ve thought a lot about this topic!
To respond to this point:
”Imagine a forest with one species of frog. Then a road is built through the forest, separating the forest in two. The day after the road is built, I discover that there are now two reproductively isolated groups of frogs! Are these now different species? … No! That’s preposterous.”
If you think this is preposterous because the two populations are not genetically distinct enough, then you are not using reproductive isolation as your definition of species, you are using genetic distinctness. Genetic distinctness COMES from reproductive isolation, but you need to find a definition that you can apply consistently. If you separate two populations so they are reproductively isolated, but you don’t split them into separate species because you actually have a different, more important criteria that they don’t yet satisfy, then reproductive isolation is upstream of your true definition.
If you want a scientific definition, it needs to be consistently applied. If you consistently apply the BSC, then those two frogs would be considered different species. Since I agree that is preposterous, we need to frame our definition more precisely than “potentially interbreeding populations that are reproductively isolated.”
I would strongly recommend reading Species Concepts in Biology for a history and explanation of the current system(s). In summary, there are multiple joints at which you can carve reality, and which joints you choose to carve depends on why you are carving reality in the first place.
As someone who studies viruses of bacteria, how are you supposed to apply the Biological Species Concept to asexual organisms anyways? Are there no bacterial “species”? But we need a way to categorize and talk about them, even if they’re not a species under the BSC. Even if this will group bacteria together whose genome lengths can vary by an entire order of magnitude.
Fundamentally, the Species Concept is the same problem. We need a way to categorize and talk about animals, so we find ways to categorize them based on how useful the categories are. Birds are feathered reptiles. Reptiles are non-feathered reptiles. We form these categories not because they carve reality at the phylogenetic joints but because these categories are useful. Fish are fish. Trees are trees.
(Viral species are even worse. Linear phylogenetic inheritance is thrown out the window. Everything is modular and recombines with each other.)
I don’t really follow the IUCN or like the focus on “biodiversity” for its own sake (just expose them to mutagens if you want more genetic diversity) (I also have philosophical issues with “conservation”—what, exactly, are you conserving and why is that time more worth conserving than any other time?) I’m probably informed by my study: in my opinion everything is a bag of genes, and genes constantly flow between them; species aren’t real and are just useful lines to draw in a continuum of gene flow, which constantly changes over time.
But can you verify it? Do you have experimental evidence that a human from location A can form fertile offspring with a human from remote location B? That’s why geographical isolation is used to define species; because they (mostly) don’t overlap, we don’t have evidence if they can or cannot interbreed. Only when we have evidence of hybridization can we start discussing whether to collapse them into the same species (ring species notwithstanding).
Also, I don’t know nearly enough about plant hybridization, but I’m pretty sure that plants can hybridize extremely easily and if we applied only an “interbreeding” species concept the categories would collapse into uselessness.
Hey Eniteris!
I have many thoughts here. First I would say that I totally agree that the species concept stops making any sense as a strict, discrete category the second horizontal gene transfer / true asexual reproduction enter the picture. I should probably have said this at the beginning of the post so that people who work in viruses and bacteria (as you seem to!) knew that my argument didn’t apply to organisms capable of that.
In my experience, however, microbiologists are too quick to generalize from their part of the field to all parts of it. It is 100% true and defensible that “everything is a bag of genes, and genes constantly flow between them; species aren’t real and are just useful lines to draw in a continuum of gene flow, which constantly changes over time” if you’re talking about bacteria. But genes do not constantly flow between humans and chimpanzees via HGT. When you’re talking about most of the animals that the IUCN cares about (for better or worse), you’re talking about sexually reproducing organisms that (imo, once they SPECIATE) stop sharing genes. You cannot treat them the same way you would a virus.
I also totally agree with this statement: “We form these categories not because they carve reality at the phylogenetic joints but because these categories are useful. Fish are fish. Trees are trees.” I don’t mean to propose that everyone stop using the words fish, or tree, even though they’re not monophyletic. But there is (and should be!) a different between technical scientific vocabulary and common usage. I don’t insist people call bananas “berries” because, in common usage, they’re not. But botanically speaking, we have a definition for “berry” that we apply consistently, no matter what the common usage is, and this definition makes bananas a berry. I believe that the technical definition of “species” should be similarly transparent and consistent. I agree that “there are multiple joints at which you can carve reality, and which joints you choose to carve depends on why you are carving reality in the first place.” But as I hope the essay above illustrates, the Biological Species Concept as it is usually formulated cannot hope to be consistently applied to relevant organisms (ie. sexually reproducing macroorganisms incapable of HGT etc)
In terms of this question asking about all human populations being able to reproduce:
”But can you verify it? Do you have experimental evidence that a human from location A can form fertile offspring with a human from remote location B?”
There have been many many recorded instances of human populations that have been long isolated being perfectly capable of being able to reproduce with other populations. To this day, there has NEVER been a recorded instance of the opposite. We know that Homo Sapiens could produce fertile offspring with Neenderthals, MUCH more different from us than any modern humans.
It would be very strange to lend much credence to the idea that not all human populations could produce fertile offspring. The burden of proof would clearly be on the person trying to say otherwise.
In terms of your philosophical problems with conservation: “(I also have philosophical issues with “conservation”—what, exactly, are you conserving and why is that time more worth conserving than any other time?)” These are all worthy questions that bear consideration. But noting that a practitioner in any given field will be required to make contestable value judgements is not really an argument against that field. The same could be said of doctors, politicians, lawyers, journalists, etc.
Do you think that we should categorize e.g. the Sentinelese people as a separate species due to lack of evidence of hybridization? That seems like a strange thing to do given that (I assume) we’re all pretty confident they can still interbreed.
The Sentinelese has been reproductively isolated for probably less than twenty generations, which can be sufficient for speciation (given strong selection pressures), but relatively unlikely. So I don’t think it’s necessary to assume that speciation has occurred.
(aside, estimates of Sentinelese isolation are full of wild speculation and misinformation. We have no genetic samples from North Sentinel Island, which means we have no evidence. All estimates are speculation, with an upper bound at a few thousand years due to shared ethnographic artifacts. Missing artifacts (metalworking, etc.) may or may not help establish a lower bound, but is confounded by natural resources available on the island. Also, going back more than tens of generations would imply a high amount of inbreeding given the limited carrying capacity of the island, which could theoretically be sustained with strong selection.)
--
Imagine a forest with one species of frog. Then a road is built through the forest, separating the forest in two. The day after the road is built, I discover that there are now two reproductively isolated groups of frogs! Are these now different species?
(under most species frameworks) No! That’s preposterous. The two populations have been isolated for less than a day. Sure, they may at some point in the future diverge (and now natural selection is acting on the two groups differently), but at this point the two species are genetically identical.
--
But now imagine you stumble across on species of frog in one forest, and a similar-looking frog in a different forest. The two population are reproductively isolated. Are they the same or different species?
Well, you might ask how long they’ve been reproductively isolated for, since if the forests were connected yesterday then we converge on the same scenario as above. A week, a month, a year, a decade, a century, a millennium, a million years. At some time of separation we cannot reject [no speciation] out of hand, or, rather, number of generations, since that’s what matters in (genetic) speciation. And in nature, we can generally assume that the groups have been isolated from each other for thousands of generations (unless the barriers are human-made, or with recorded history, and because our genetic divergence estimation tools don’t work well for times less than that).
Can they produce fertile offspring? Well, no, not in nature, since that’s what reproductive isolation means. Maybe you can put two of them in a box and test it out manually, but you don’t have the funding to do that. And genetics don’t help too much in determining hybridization potential.
(an aside, what counts as producing fertile offspring? If they can produce fertile offspring via IVF, but never crossbreed if you cage them together because they don’t have the same mating calls or compatible mating apparatus or look funny to each other or sleep at different times of the day, does that count as producing fertile offspring? this is why “natural populations that are reproductively isolated” is added to the definition)
So you have two groups of frogs, isolated from each other for thousands of generations. Evolution is acting upon them as two different groups. But they look suspiciously similar. Do you call them different species?
I could go either way, and lean towards labelling them as different species until we have evidence to the contrary. Otherwise one of the forests might be bulldozed before we can gather the evidence.
Hi eniteris! I appreciate your long and detailed responses. Clearly you’ve thought a lot about this topic!
To respond to this point:
”Imagine a forest with one species of frog. Then a road is built through the forest, separating the forest in two. The day after the road is built, I discover that there are now two reproductively isolated groups of frogs! Are these now different species? … No! That’s preposterous.”
If you think this is preposterous because the two populations are not genetically distinct enough, then you are not using reproductive isolation as your definition of species, you are using genetic distinctness. Genetic distinctness COMES from reproductive isolation, but you need to find a definition that you can apply consistently. If you separate two populations so they are reproductively isolated, but you don’t split them into separate species because you actually have a different, more important criteria that they don’t yet satisfy, then reproductive isolation is upstream of your true definition.
If you want a scientific definition, it needs to be consistently applied. If you consistently apply the BSC, then those two frogs would be considered different species. Since I agree that is preposterous, we need to frame our definition more precisely than “potentially interbreeding populations that are reproductively isolated.”