Always a step ahead.

March 1, 2011 11:17 AM

Why are arthropods the only animals with more than four legs?

Mammals, lizards, amphibians, birds.. all two or four legs. Is there absolutely no evolutionary benefit to having more legs if you're over a certain size or weight? Why would this be the case?

posted by curious nu to science & nature (41 answers total)

Why are arthropods the only animals with more than four legs?

The common ancestor of mammals, lizards, amphibians and birds had four legs.
posted by Threeway Handshake at 11:25 AM on March 1, 2011 [2 favorites]

IANAEB, but I suspect the answer may have to do with the vertebrate/invertebrate distinction and that may be a good place to continue your research.
posted by norm at 11:26 AM on March 1, 2011 [1 favorite]

I just happen to be reading a book on this very subject at the moment! (Life, by Richard Fortey. It rules.)

From what I understand, it's as simple as this: mammals, lizards, amphibians, and birds all have a common ancestor in some sort of fishlike animal that made its way out of the sea. By sheer chance this animal had four limbs and five toes. But arthropods branched off way, way before that. Think trilobites. They were probably the only land animals for a while there.
posted by showbiz_liz at 11:27 AM on March 1, 2011 [2 favorites]

Is there absolutely no evolutionary benefit to having more legs if you're over a certain size or weight?

Evolution does not result in advancement of every possible advantageous trait. Only those traits that happen to come into being by way of mutation at just the right time and place in order to have a reproductive advantage get to advance. There's no intelligent mechanism that says "hey, I bet this doohickey would be useful to have as part of an organism."

By sheer chance this animal had four limbs and five toes.

This.
posted by The World Famous at 11:29 AM on March 1, 2011 [5 favorites]

Sure, I get the common ancestor bit, but I'm wondering if there's any structural reason why, after millions and millions of years, we haven't wound up with any 4+ legg'ed mammals, say. Why NOT an eight-legged horse? Does that require such a huge shift in the skeletal structure that any mutation is more likely to die out than be successful? Is it a brain/CNS issue?
posted by curious nu at 11:32 AM on March 1, 2011

why, after millions and millions of years, we haven't wound up with any 4+ legg'ed mammals, say. Why NOT an eight-legged horse?

Because horses with extra mutant stump leg thinggies that might eventually be able to evolve into extra legs after millions of years do not have a reproductive advantage. The girl horses don't want to have sex with the boy horses who have weird extra legs. Evolution of a trait requires that organisms with that trait be more successful than others at reproducing. Why would having the beginnings of an extra set of legs make it more likely for a horse to have sex?
posted by The World Famous at 11:36 AM on March 1, 2011 [3 favorites]

Yes. Look up "hoc genes". Basically the 4 limbs with an increasing number of bones as you go out is pretty fundamental to vertebrae. Losing the functionality of links (kangaroo t-rex) is easy enough but sprouting a whole new one would require some deeply complicated protein refolding very early on.

If, for example, echnoderms ever come out of ocean we can safely expect them to skitter rather than walk about since they have radial symmetry.
posted by fshgrl at 11:39 AM on March 1, 2011

Argh. "hoc genes" should be "hox genes".

I need to evolve smaller thumbs.
posted by fshgrl at 11:40 AM on March 1, 2011 [4 favorites]

Given that mammals with the properties you want do exist, the question you have to ask is, what is the advantage that outweighs the cost of having an extra leg? If the cost is major, it may die off before reproduction. Only if the cost is minor would one expect the mutation to survive in the population, and wander into an improved configuration.
posted by pwnguin at 11:44 AM on March 1, 2011

So, in a science fiction example, separate colonizations of land on Pandora is "implied" (insofar as the writers thought of this) because the Na'vi are humanoid but their horses are hexapod?
posted by chengjih at 11:48 AM on March 1, 2011

As it's been pointed out, the basic 4-limbed structure is defined in the spine and core region, and adding limbs to the basic earthling vertebrate structure without major structural changes (see polymelia, thanks pwnguin for the link) does not make a more functional animal. Okay, so what about changing the basic structure? The time to have made your 8-legged horse would have been not "after millions and millions of years", but in fact way back at the beginning. All the years since the basic shape was established are kind of moot.

(But, fshgrl, will smaller thumbs help you have kids? If not, no dice.)
posted by aimedwander at 11:52 AM on March 1, 2011

Due to common descent, the way novel features are typically added through evolution is by gradual modification rather than wholesale spontaneous introduction. It just so happened that the common ancestor of tetrapods had four limbs. But to sprout a new pair of limbs is not a trivial thing to do, requiring mutations in a number of genes at the same time--typically when that happens, the organism has been exposed to some external agent, and dies during development. Even if you get the mutations, there's no guarantee that 1) you can find a mate, and 2) that those mutations are all dominant and can be readily passed on to your progeny.
posted by reformedjerk at 11:52 AM on March 1, 2011

Or the Navi lost a couple limbs along the way, yeah. This isn't as simple as sprouting an extra limb and everyone being all "thats awesome!" and then you have a bunch of kids with extra limbs and they take over the world. Embryonic development has a huge number of controls on it: feedback loops, physical restraints due to molecular shape, on-off switches in sections of dna. It's pretty oriented towards producing an organism very similar to the parent.

Sexual selection is different but can only act upon variation that already exists within the population. Extra limbs are not going to be a common variation in vertebrates because its such an old structure.
posted by fshgrl at 11:56 AM on March 1, 2011 [2 favorites]

The circulatory system is one of the most expensive things about being a mammal, especially on land where we have to be able to pump blood out to all of it, and get the blood to come back for re-pumping, against gravity. Water mammals can get big because they don't have to deal with the whole gravity thing, whereas land mammals are stuck at elephant for that reason.
More limbs, more circulation, more risk, lower odds of succeeding at life well enough to pass on the trait.
posted by L'Estrange Fruit at 11:59 AM on March 1, 2011 [2 favorites]

While I appreciate the historical accident/natural selection doesn't work like that arguments, I don't think it's completely crazy to consider it. Snakes, for instance, sometimes grow an extra section in the middle and aren't scorned as mutants. Presumably because it isn't visible, but the point is that "do this, but again" isn't that huge a genetic leap as you might think. Actual human beings are born with extra fingers and toes all the time, for instance.

So, simply from an engineering POV: Large animals are either bipeds (for some good reason, because you need a lot of extra sensors and processing) or quadrupeds (otherwise). Quadrupeds can leave 3 feet on the ground while stepping, which is stable. If you have 3 feet exactly, you have to be a biped while stepping, forcing you to basically be a biped with an extra leg, which is dumb.

Now let's say you want to move faster than one step at a time. Like galloping in horses. If you are a lame-o robot, you do this by having, say, 6 legs and lifting 3 at a time. Except the whole premise is I want to be moving fast. Nimble is the opposite of stable. You don't want 3 feet on the floor while you are running, you'd like to have as many in the air as possible (perhaps leaving one down for control). This is in fact how horses gallop.

So I think 4 represents a good balance of stability when you need it vs instability when you need that.
posted by DU at 12:01 PM on March 1, 2011

Water mammals can get big because they don't have to deal with the whole gravity thing...

I think you'll find that gravity works underwater.
posted by DU at 12:02 PM on March 1, 2011 [2 favorites]

Animals will have the minimum features necessary to do the job, because everything has an energy cost. A six legged horse would have two extra legs to feed. It would not be able to run faster or with more endurance. It might have a slight advantage that it could lose one or two legs and remain stable, but it would make more evolutionary sense -- it would be cheaper -- to make four sturdy legs -- and in the wild, horses' legs are quite sturdy.

The jump to two legs is fairly difficult, since balance is so hard. (Balancing on two legs is a feat that computers are only now coming to grips with.) But once you're down to two legs you have a major jump in efficiency. Human beings expose quite a bit less skin to the sun, and are among the most efficient runners in the world -- not the fastest, but using the least energy per mile. That's why human beings can run antelopes to exhaustion. Kangaroos are also ridiculously efficient. It's just hard to get from four legs to two bouncy legs.

In fact the question isn't really why mammals don't have six legs. If there was an evolutionary advantage, they could have developed them at some point. The question is why spiders need eight legs and insects need six. Why can't ants make do with four legs?
posted by musofire at 12:03 PM on March 1, 2011

2 (bilateral symmetry) times 2 (beginning and end) is four. At the most basic structural level, we're mirrors input end to output end, and side to side. Some lower vertebrates have nerve clusters at the base of the tail to balance the brain. Shoulder blades and pelvises are mirror images of each other, as are legs and arms (pre-specialization). It's the simplest configuration using only a mirror function, that is also stable (vs. a tripod).
posted by notsnot at 12:08 PM on March 1, 2011 [1 favorite]

Water mammals can get big because they don't have to deal with the whole gravity thing...

I think you'll find that gravity works underwater.

Yes yes, it was a poorly stated oversimplification of the principle involved. And now I can't remember how I should have stated it, but the principle is still that mammals in water can be bigger because it is less effort to pump the blood around all that mass.
posted by L'Estrange Fruit at 12:09 PM on March 1, 2011

the point is that "do this, but again" isn't that huge a genetic leap as you might think. Actual human beings are born with extra fingers and toes all the time, for instance.

On an embryonic level growing an extra finger or toe is absolutey trivial compared to sprouting an entire extra limb. Sprouting an entire functional limb that also has a functional attachment to the rest of the body is another thing again.

It's the difference between finding a penny on the ground and winning the CA lottery. Two weeks in a row. With the same numbers.
posted by fshgrl at 12:22 PM on March 1, 2011

The question is why spiders need eight legs and insects need six. Why can't ants make do with four legs?

I don't know if you're just putting that out there hypothetically, but I believe this question is heavily explored in robotics R&D. A big problem is getting the machines to travel on anything other than totally flat, smooth terrain without falling over or getting stuck. Insects are frequently used as a model because of their ability to get around on challenging terrain with a minimum of intelligence. Even four-legged locomotion requires that significant resources be dedicated to maintaining balance. Six or more legs makes it much less of an issue.

I believe spiders make use of their extra legs in web building and for wrapping up their prey. Those back legs are handy for reaching their spinnerets while keeping six legs steady.
posted by keep it under cover at 1:02 PM on March 1, 2011

If you study the history of life one thing that becomes evident is that it is vastly easier, genetically and evolutionarily speaking, to change the size and shape of an existing structure than it is to create an entirely new one, even if it's a copy of an existing structure.

That's why whales still have hind leg bones. There are two femurs, two fibulas, and two tibias inside of whales. But they're smaller than yours, and they don't stick out of the surface of the whale. Evolution has reduced them to the point where they're irrelevant, and that's good enough. There is no longer any evolutionary advantage to reducing them even further.

It is indeed possible for things to vanish entirely, and among land vertebrates the best example of that is in the number of digits on each limb. Bird forelimbs (the wings) have five, but their feet only have four (I'm pretty sure). Horses are down to a single one. (I believe there are two others which are vestigial.)

But adding structures is exceedingly improbable, and there really aren't any important ones I can think of in the history of land-based vertebrate life. (For instance, the legendary Panda's Thumb wasn't a structural addition; it was a dramatic repurposing of an existing structure.)
posted by Chocolate Pickle at 1:19 PM on March 1, 2011

Water mammals can get big because they don't have to deal with the whole gravity thing...

I think you'll find that gravity works underwater.

Yes yes, it was a poorly stated oversimplification of the principle involved. And now I can't remember how I should have stated it, but the principle is still that mammals in water can be bigger because it is less effort to pump the blood around all that mass.

I think you are referring to the density difference between air and water: a whale is more or less the same density as the surrounding water, so gravity (or buoyancy, if you want to call it that way) does no put mechanical load on the animal. On land, increasing the volume (and thus mass) of a body requires an immense increase in mechanical support (i.e. super-strong bones) once you get bigger than e.g. an elephant.

With respect to blood circulation: this is still under debate. Keep in mind, however, that the potential energy that you recover from the "falling" blood is more or less balanced by the increase in potential energy of the rising blood (it is a closed loop!). Many people oversimplify it as a "syphon", but that is not a closed loop. There is an effect due to vessel distensibility and partially collapsed veins, but that's probably too much detail.
posted by swordfishtrombones at 1:39 PM on March 1, 2011 [2 favorites]

I believe spiders make use of their extra legs in web building and for wrapping up their prey. Those back legs are handy for reaching their spinnerets while keeping six legs steady.

FYI, keep it under cover: spiders' ancestors had eight legs long before they had web spinning.
posted by IAmBroom at 3:10 PM on March 1, 2011

The question is why spiders need eight legs and insects need six. Why can't ants make do with four legs?

They do. There spiders that have six legs - they use the front two as pseudo-antennae, in order to camoflage themselves to ants.

There are insects with four legs, such as the praying mantis, which evolved it's front two legs into pincers, the same way birds evolved their front legs into wings, and we evolved ours into hands.

Beyond that, any insect with a different number of legs is called something else, insects have six legs by definition, though there are no shortage of insect-like animals with a different number of legs.

So it seems that changing the use and design of limbs happens frequently, but you have to go much further back up the evolutionary tree to reach the points where organisms shifted their basic spinal configurations or arrangement of limbs. But likewise, it still happens. Just on much broader timescales.
posted by -harlequin- at 4:36 PM on March 1, 2011

I recommend you read Endless Forms Most Beautiful: The New Science of Evo Devo and the Making of the Animal Kingdom by Sean Carroll.
posted by neuron at 4:42 PM on March 1, 2011 [1 favorite]

So, after reading Wikipedia for a bit and skeeving myself out looking at pictures of shrimp (how can people eat those things? SO CREEPY.), I have a couple thoughts. Apparently, in decopods (like lobsters), each body segment can bear one pair of appendages - a lobster has 19 body segments, 5 in the head (antennae, mandibles), 8 in the thorax (claws, walking legs), and 6 in the abdomen (swimmerets, tail/anus). Best I can tell, the two limbs per body segment is true for other arthopods (centipedes/millipedes, insects, spiders). Centipedies can be hatched with only 6 legs and gain more during molts.

Presumably, vertebrates (which is most of what we're talking about here, right? Because the only other things with limbs have radial symmetry, I think) have lost the ability to increase or decrease the number of body segments. My assumption is that this has to do with forming a closed circulatory system - it seems a lot easier to increase body segments when there's just a big hole in the middle for a heart. Also, we don't molt - dissolving the entire rigid skeletal system allows a lot more flexibility (no pun intended) with changing your body plan.

So, I guess I'm saying we don't have more limbs because we don't moult. Without moulting I don't think the genetics to modify your body plan so drastically exist. But I don't *know* any of this. I'm just logicking from Wikipedia so...yeah. And while this has been a lovely distraction to think about, I'd best go back to my far less complex E. coli now...
posted by maryr at 4:47 PM on March 1, 2011

Another thing to consider is that additional limbs would change the gait of an animal which could have far reaching consequences for overall physiology of it. This could make it much more difficult for extra limbs to be selectively advantageous.
posted by Midnight Rambler at 5:34 PM on March 1, 2011

maryr -- millipedes have four legs per body segment. Centipedes and other myriapoda have only two.

I find it interesting that caterpillars have only six legs, just like other insects, and that their numerous extra "feet" are pseudopod "prolegs".
posted by autopilot at 5:35 PM on March 1, 2011

autopilot - according to Wikipedia, "each segment that has two pairs of legs is a result of two single segments fused together as one." Makes sense from a genetics (i.e. Hox genes) standpoint. But I know nothing specific about it and y'know, Wikipedia is my source.
posted by maryr at 5:37 PM on March 1, 2011

Reptiles and mammals do, however, have tails; some can hold onto things with their prehensile tails. A relatively easy way for evolution to provide the function of an extra limb without the effort of producing a six-legged animal.
posted by bad grammar at 5:48 PM on March 1, 2011 [1 favorite]

I am an evolutionary biologist.

In short, in my opinion, your initial commenters got it right -- you're looking at a common constraint of the Tetrapoda (the four legged animals -- the descendants of the early amphibians.)

I think your question is up-side down. The issue is not "why are the arthropods the only animals with more than four legs"; the issue is "why are the tetrapods constrained to four legs?" I'm using leg to mean things like wings and arms and whatnot as well. The arthropods represent the majority of the animal kingdom by species diversity and biomass. Arthopods are the norm. Tetrapods are the exception.

The arthropods are so diverse because they're segmented. Sometimes the segments have fused, and so the segmentation is not obvious, but, essentially, every arthropod is a series of repeated segments, each of which has a ganglion and two limbs. The limbs on each segment have been specialized for eating, walking, flying, sensing, copulation, etc. And by duplication of the master control genes (the hox genes referred to above), segments can be duplicated, and, over evolutionary time, the new segments can differentiate. So, arthropods can duplicate and differentiate body segments by fairly well understood genomic mechanisms.

The tetrapods don't have this duplication mechanism. While our bodies show traces of ancestral segmentation, our bodies are much more integrated. The genomic flexibility just isn't there. (I'd be curious to hear more about DU's assertions about segment duplication in snakes.)

You can't make an argument about whether or not a trait will spread in a population if the trait never arises in the first place. And genuinely functional extra limbs do not seem to arise in tetrapods at the rate necessary to allow any putative advantages to spread in a population. (Even beneficial adaptations are very likely to be eliminated early on by genetic drift when they arise spontaneously.)

tl, dr: Evolution requires variation. It doesn't matter how adaptive something would be if it represents too great a leap in genetic space.

Incidentally, there used to be centipedes six feet long, and dragonflies the size of raptors. At least one argument is that arthropods size is constrained by atmospheric oxygen, which was higher when things like that roamed the earth. So maybe the question isn't "does size affect limb number", but "do limb number and size both correlate with a third factor, mechanism of oxygen exchange?"
posted by endless_forms at 9:27 PM on March 1, 2011 [3 favorites]

Because the only other things with limbs have radial symmetry, I think

No, arthropods have bilateral symmetry like vertebrates.
posted by fshgrl at 10:19 PM on March 1, 2011 [1 favorite]

fshgrl, I think it depends whether you consider a starfish to have legs -- five of them.
posted by Chocolate Pickle at 2:15 AM on March 2, 2011

fshgirl - I'm sorry, I meant other than arthropods, vertebrates and starfish/jellyfish were the only things I could thing of with limbs, the latter having radial symmetry and the former being what I assumed we were comparing arthropods to.

endless_forms - thanks for the explaination! I'm glad to see that thinking of it from a molecular biology standpoint (i.e. focusing on the hox genes) was sort of getting me toward the right answer. =)
posted by maryr at 6:38 AM on March 2, 2011

Starfish aren't arthropods. They're echinoderms. Arthropods have a chitin exoskeleten and bilateral symmetry, echinoderms have calcified mesodermal spines and radial symmetry. (/ evolution major).
posted by fshgrl at 1:48 PM on March 2, 2011 [1 favorite]

And there are other phyla with limbs; annelids for example. I don't think starfish technically have "limbs" but for the purpose of this discussion (useful thingy with which to move around of an undetermined number) they probably count.
posted by fshgrl at 1:52 PM on March 2, 2011

I also got a little stuck on "what's a limb". Cephalapods have arms. As fshgrl alludes to, some annelids (and also some molluscs) have parapods. True jellyfish have tentacles (the comb jellies also have tentacles, but they have bilateral symmetry.) Is a limb a weight-bearing, jointed appendage with skeletal support?

I think I'd argue that the starfish and their relatives have more like hundreds of feet than five -- check out this video of starfish locomotion. Keep in mind as well that not all starfish have five arms.

and it's not actually relevant to the discussion at hand, but echinoderm larvae have bilateral symmetry, and in the big picture the echinoderms are among our closest relatives.
posted by endless_forms at 2:25 PM on March 2, 2011

Here's an octopus walking.

("walking octopus" gets you a lot of neat-o youtube vids.)
posted by endless_forms at 2:43 PM on March 2, 2011

Well, if we're going to go in to 'what is a limb', there are plenty of microbes with flagella...

fsh - yes, I know what arthropods are, that's why I said OTHER than arthropods.

endless - when does the radial symmetry come in? And how on earth do they split in to 5? I just can't picture how they do that evenly...
posted by maryr at 2:43 PM on March 2, 2011

maryr -- I'm looking at pictures right now, and I still can't picture it.

If you have journal access, you might try: Smith 2008, Deuterostomes in a twist: the origins of a radical new body plan. EVOLUTION & DEVELOPMENT 10:4, 493–503

To recap slightly: Hox genes are the master regulators of body plan development in all* animals. In virtually all, their order in the genome is conserved, and their order in the genome is reflected in the physical bodyplan of the animals. That is, the first gene in sequence codes for the frontmost region of the organism, the second for the next region, and so on. This is kind of amazing.

Echinoderm larvae follow this pattern. These are larvae of a type where they float in the water column as larvae, and then settle down on a nice rock somewhere. The article cited above argues that the ancestor of the echinoderms developed a novel trait -- instead of optionally attaching to the substrate on their posterior end (backside), they developed obligatory anterior (near the head) attachment. This makes feeding difficult, so organisms which were partially "twisted" to bring the mouth into a better feeding position. Eventually, the actual axis of the body was 90 degrees out of true with the gradients laid down by the Hox genes. So the larvae have two body axes at odds with each other. Then most of the larval structures themselves are broken down, and a pentaradial adult emerges. Yes, I know that's unsatisfying. The above theory explains how the normal bilateral Hox gene patterning got out of sync with the body plan in echinoderms, and left the field open to "repurposing and co-option" of the gene product gradients for other purposes, but it doesn't explain where the pentaradiality comes from. I actually don't know if anyone knows (and if anyone does know I'd love to hear about it.)

Another paper abstract seems to imply that the Hox genes fell out of order in the ancestral echinoderm, leading to "genome disorder" and, again, allowing the gene product gradients to be repurposed.

It all boils down to, if you see the Hox gene products mapped on to the body plan of an adult echinoderm, they are inconsistent with its apparent pentaradial symmetry. (Unlike a true jellyfish, which has true tetraradial symmetry which conforms with its Hox gene map.)

*Except sponges, okay.
posted by endless_forms at 3:17 PM on March 2, 2011