What is a species?
 

The species concept has been a little awkward and troublesome since the start. In his earlier editions of Systema Naturae, Carl von Linnè considered swedes to be a species with different kinds of swedes as sub-taxa. The simple truth is that differences between organisms are often not clear cut and biologists today are still having trouble defining what a species is - what line you have to cross before you go from one species to another.

People in general are not as weighed down by academic considerations and may have a more clear cut idea of how to define species. As the world is not really managed by biologists, the general perception of species is immensly important to stuff like conservation and less academically hindered definitions may give valuable insight.

So, that was the intro :) Now, here's a question for you to answer :

How would you define a species?

Please, don't look up the answer before you write because you won't find one (at least not by consensus, only different ideas) and I'd like to see some unadulterated opinions. No definition is too silly and if yours contains one good idea and one bad one, we can possibly ditch the bad and build on the good!

species is a blurry category indeed.

first and foremost, the label "species" can only be applied retroactively. this is because it is impossible to draw a line across an evolutionary path, marking the moment when species A became species B. a parent can never be a different species than its offspring.

as such, it must be a very general category, and must be analogical rather than digital.

generally speaking, two organisms are of a different species if they will not interbreed. i imagine this is as much social as it is biological.

of course, this definition has problems. some animals are physically capable of interbreeding, but will not for social reasons. and vice versa...

I consider a species a group of animals that can interbreed and produce fertile offspring (My own words btw) as that's what I was taught, for instance, a horse and a donkey are not the same species in my opinion as they produce a mule which in turn is sterile.

Maybe you can help me answer this tore as I always wondered it.

Why is the mule sterile? Why is a "Liger" sterile? I was never taught that and I always wondered why two animals from the same family couldn't interbreed (well they can technically) but they cannot produce fertile offspring? It's always confused me.

Some interesting definitions so far :) You both suggest to define a species by reproductive isolation as it were, all organisms that can have fertile offspring together form a species. This makes sense because you would think that these will evolve independently from all other organisms and indeed that's pretty much what's traditionally now thought of as the biological species concept.

There are some deep underlying problems with this species concept, but I think I'll wait until a few more definitions come in before I comment any more on them. :)

That's a deceptively difficult question to answer. As you've probably guessed, it has to do with genetic stuff. You probably know that animal genes are organized into long strings of DNA and associated proteins called chromosomes. Diploid animals have one set of chromosomes from dad and one from mum. These chromosomes are copied every time a cell splits into two. It's quite simple, you get a duplicate of each chromosome and they align themselves at different sides of the cell core (nucleus) and then the whole thing splits down the middle and you get two new cells with the same two sets of chromosomes each. That's called mitosis.

However, when sperm and eggs are created, there's a slightly different process called meiosis. Instead of one cell dividing to two, one cell shall become four sperm or four eggs, each containing only half the amount of genetic material as the parent cell. They will be haploid, containing only one set of chromosomes with some from mom and some from dad. To ensure that each new egg/sperm cell gets a nice mixing of genes and chromosomes from both mom and dad in that 1 set, there's a step in meiosis where homologous chromosomes from mom and dad pair up and exchange genetic material.

When two organisms that are very dissimilar create a hybrid (f.ex horse and donkey), the diploid mule offspring's cells can have two very different sets of chromosomes. When meiosis is to take place in the hybrid, chromosomes can have a hard time pairing up with their homologues from the other dad/mum set or the mixing between them creates problems .. possibly because their homologues in the other set are so different or even because they're simply not present! Horses and Donkeys have different amount of chromosomes which would result in some chromosomes being unable to pair up with homologues. This can cause a multitude of problems, many which end in reduced ability to reproduce and even sterility.

I'm not sure if that made it understandable. Possibly, this meiosis illustration might help a little bit. :)

edit :

Gurgh, had to brush up a bit on that explanation. I think it's understandable now. :p:

edit 2 :

By the way, some "species" do hybridize quite a lot in nature, for example many birds. Hybrids are not always sterile, but may simply have lower reproductive ability which translates to lower fitness.

I didn't even think about the bolded. I feel stupid for not realizing something as important as that. I'm assuming either horse or donkey has more information than the other, leaving spare chromatids over? I never learned any of this during school, I'm just guessing btw. So if two animals in a similar situation to horse and donkey reproduced but had an equal number of chromosomes the product would be fertile? Or would the only time this happened be within the same species? Meaning that in every circumstance different species never have an equal number of chromosomes? I think I may be complicating things.

what an interesting question.

i recently watched a documentary called "earthlings." despite the fact that there were video clips depicting what man does to species lower on the totem pole and it brought me to tears on multiple occasions (ps, i love eating meat... a lot), it brought a new point to my perspective going by the name "Speciesism," wherein humans believe that because we are more ingenuous, we are able to without guilt do things to animals that if happened to us via the hand of our leaders, would no doubt result in violent uprising.

it made me think a little about the fact that we're the only species capable of looking at all the different species and making a decision to try and understand them, whereas they just go about their lives being a part of the system.

there is also some science floating around concerning the universe and the possibility that it might be conscious, like a species of it's own. if you examine the way the universe works, you'll find that yes, indeed, it WORKS, making it possible to call it a SYSTEM. an autonomous one at that, and thusly alive...? it doesn't reproduce the entire thing (obviously another convo, involving multidimensional thinking, so to speak), but planets die and are born within it... galaxies... quasars... all just parts of a seemingly ceaseless show.

is anything classifiably ALIVE a member of some species? and if so, to what species does the universe belong, for it too exhibits the traits of life?

that might sound like a really dumb question, and it's on the fly, so jumbly it is. but, if we can look at the universe as the mother of all things, the complex within which our planet and many other planets reside, on which we and our dogs reside, inside which the minutiae of reality resides, the word "species" becomes obsolete or irrelevant, particularly when considering evolution as it is there that we see nothing is quite as static as we'd like it to be.

i guess i'm saying that classifying things as species is not a tangible concept... still not sure if that's what i believe though, it's just what came out.

maybe there is just life.

Very interesting thread, Tore. I love your thread ideas. I'm going to have to sit on this question for a bit, though. Interesting discussion so far, though. I had no idea that some hybrids were sterile.

You may be complicating things, yes, but don't feel stupid :) I find most people have little real knowledge about this and that often includes myself! Considering horses and donkeys are quite similar, if they had the same amount of chromosomes, I'm thinking their chances at producing fertile offspring should at least be much better, so I'm guessing that yes, they could create fertile offspring! That's just a guess though. There are still more ways for genetic material to be different and even incompatible other than the count of chromosomes. For example, on what chromosomes certain genes are located can differ and so I'm sure there could be a lot of possible negative effects from unhealthy genetic interactions between chromosomes from differing species. Simply in order to illustrate, imagine if species A has a vital gene on chromosome nr. 1 and it hybridizes with another species B that has the same gene, but on chromosome nr. 2. If a hybrid's sperm cell contains chromosome number 1 from species B and chromosome nr. 2 from species A, the sperm would not contain the gene at all. Maybe that's an overcomeable problem, but imagine that variations of the same problem also happens with other genes. I'm not saying that's what happens, genes in different locations, but it illustrates how differing genetics could cause trouble. Plus, I'm sure that before you get to the point where you actually have sperm/egg, there could be problems with pairing and recombination as well with some unhealthy results, but let's not focus too much on that otherwise this post will get horribly lengthy.

Also, there are a couple more here with knowledge in biology, Vegangelica being one of them, and they might know more about this than I do.

As I wrote above, many hybrids are fertile. Very many in fact - and not all of them are animals either. Hybrids are very common with plants for example and the variation in the amount of chromosomes going on in the plant kingdom quickly gets high and quite chaotic. If we stick to animals, birds are good examples of animals that hybridize a lot. Not all birds of course, but many species. Negative effects from what I'm taking an educated guess at what must be negative genetic interactions and so on usually cause these hybrids to have a lower fitness than most average members of their parent species. For this reason, it should be generally adaptive to try and avoid producing hybrids, but of course birds can get confused as well when different species look so alike!

Anyways, there's a lot of stuff going on so it's hard to say "this explains all of that" if you know what I mean, but at least I can add something to help solve pieces of the mystery. :p:


To zevokes, about you wondering what species the planet is, back in the days of Linnè when the binomial taxonomy as we know it was brand new, people actually tried to organize minerals the same way, with families and genera and species and so on. They stopped at some point because there's no real evolutionary relationship between minerals like there is between organisms (even if they had no concept of evolution at the time). Your point was a little more profound, but I thought it might be an interesting nugget of trivia.

Thank you, I think you answered my question as much as it can really be answered and defined. :D

And you can only do so much in a forum posts .. There are several text books that cover these problems in more detail. :p:

Back to the species definition, sexual isolation has been mentioned - that members of a species can have reproductive offspring or - turned on it's head, that they are reproductively isolated from other species.

But how about similarities? Shouldn't members of a species also look similar and have similar ecology?

I thought the dividing line between one type of animal with another regardless how similar they are to each ofther is the number of chromosomes they possess. And generally I always thought that a species can only reproduce with it's own kind. There is always an exception to every rule e.g. the mule, being an offspring of a donkey and a horse, both being of different speciesfrom each other.

And since I'm following your orders not to look up any information beforehand, I have no way of verifing this but there is another example of a corss-species breeding I want to bring up and that is the "Cabbit," an offspring of a rabbit and a cat. I have no idea if the Cabbit is classified under the rabbit's species or the cat's species or does it share a dual citizenship between both species???

Another mysterious animal that is a result cross-breeding species is the "Camelopardus," which I became fimiliar with oddly enough not through Zooology but through Astronomy. I presume by it's Latinish name, that maybe in the ancient times the ancient Romans cross-bread the leopard with the camel for who knows what reason, but maybe they had a utilitarian purpose for it. By combining the endurance of a camel and the speed of a leopard the ancient Romans found a solution to quickly build their roads in record time using their hybrid animal that had specific qualities build into it that made it more advantageous to use then using a slow stubborn mule that would just hold up progress of said road building projects. All thanks to their scientific knowledge of cross-species breeding. The reason why I brought up the Camelopardus is that while modern scientist are endlessly experimenting with genes of animals (e.g. cloning sheep, cross breeding mules and glow-in-the cats) they've yet to reproduce an hybrid animal from antiquity.

The Cabbit and the Camelopardus throw a monkey wrench in the cog wheels of the machinery of science, when it comes to classification of animals by species. What is a "species" and how should we classify them, and their hybrid-off-spring? What is the critea for classification, and is it worth editting the hundreds of textbooks for the millions of students in the high schools across the United States when the definition is changed. It reminds me of the kind of dilema we had with Pluto's declassification from Planet to Planetoid.

Scientifically, a species is in the order below genus.
here's an easy way to remember it, tough it's missing domain:
King Phillip Came Over From Great Spain.

My bio teacher got us to remember that as King Philip Came Over For *** Sex.

It was much easier to remember that way.

But...your response doesn't really answer tore's question.

I'm a supporter of the reproductive isolation bit. Unless someone finds an organism (say, an amphibian) that can breed with a totally different organism (say, a mammal) and produce viable offspring, I'll stick with that.

not necessarily...

http://pics.livejournal.com/ecctv/pic/00afe831

http://www.stanford.edu/~siegelr/ani...dimorphism.jpg

http://www.cathouse-fcc.org/imgs/nyala.jpg

The horse and donkey are still seen as viable species from the species definition that species are reproductively isolated. The reason is the importance is not on whether or not two species can create a hybrid, but whether or not they can create a hybrid which itself is capable of sexual reproduction. A mule isn't, so horses and donkeys are generally seen as viable species according to the sex-definition. :)

About chromosomes and using those, a lot of species are species simply because they were described back in the days when they had no way of checking this. Today we can check it, but if we rearranged species by this criteria, humans would be the same species as tobacco. Silkworm and elephants would also be the same!

They don't really throw a cogwheel in the machinery of science because cabbits and cameloparduses only ever existed in people's imagination. Cats and rabbits and camels and leopards are not capable of hybridizing together. They're intriguing fantasy animals, but that's it I'm afraid. Where did you learn about them?

Many fantasy animals thought up in the olden days were basically just mixes of other parent species. Some other examples are the griffin (lion, dragon, eagle), centaur (human, horse), harpy (human, bird) or hippogriff (deer, griffin), manticore and sfinx (lion, human) .. Creatures everyone who's ever played Dungeons & Dragons will be familiar with.

The Cockatrice for example is a mix between basilisk, lizard or dragon with a rooster.

http://xs.to/image-23C8_4B8E2B64.gif

At some point, medieval scholars thought they were created when a cock laid an egg and that egg then got incubated by a toad or possibly a snake. Needless to say, this has little to do with nature, but everything to do with imagination. :)

In theory it works well for many organisms, it would seem, but it sure is hard to check since all organisms in nature don't occupy the same geographical areas and don't all have the same courtship behaviour and so on.

You get further troubles when you get to plants. They hybridize a lot and such evolutionary events are thought to have given rise to a multitude of the species out there. Another problem is this; imagine that species 1 can hybridize with species 2 and 3, but 2 and 3 can't hybridize with eachother. The sexual isolation definition would have to refuse and accept 2 and 3 as the same species at the same time!

This problem does happen in nature, sometimes involving a lot more species than 3. Brassicas are an example. Although I don't know for sure, I assume this must happen in some bird species as well.

I agree and if you flip the problem over on it's head, many species that should obviously be different look almost exactly the same. A good example is the phylum Nematoda (nematodes), a huge group of wormlike creatures that have differing ecology. Some are parasites while others for example live in the soil. However, they all follow the same general body plan and look similar to eachother. Two seemingly identical nematode worms can be further apart genetically than humans are to any other mammal.

So I would argue that both sexual isolation and similarity/dissimilarity are not good species criterias.

I'm speechless.

That is why I mentioned both similarity of the aminal and the number of chromosomes. I didn't say just the number of chromosomes alone. It's blatantly obivious that you examples (humans & tobacco; silkworms and elephants) aren't even remotely close. I meant similarity of the aminal and the number of chromosomes for donkeys and horses. Yes from a standpoint from of a pedestrian observer they might seem like the same species, but through scientific means we know they have different amount chromosomes. A horse has 64 chromosomes and the donkey has 62 chromosomes. Mules and hinnies have 63 chromosomes. Mules and hinnies together don't not have their own species classification but it's combined [Equus caballus + Equus asinus] right?

I don't know if it is the best idea to divided a living creature into seven ranks:
Kingdom, Phylum, Class, Order, Family, Genus, Species. The number Seven is generally prefered because of the history of it being known as a mystical or magical number. Having such a system seems like science is relying on Metaphysics rather then a rational scientific schema. If Evolution is to be true such a neat system of diving life into 7 ranks seems to be a bit of an archaic notion. Shouldn't some species have 5 ranks and other species 500 ranks depending on how many stages it took of evolving into different animals to suddenly morph into it's present form?

While looking up information which was something you told us not to do in the opening post, I came across a fact I found out interestinly enough that a Camelopardus (which is a modern constellation) is named after the ancient latin word for Giraffe, a Camelopard. Infact a Giraffe's binomial name is [Giraffa camelopardalis.] You said it only exists in people's imagination. I will not celebrate the fact that you are completely wrong that a Camelopard isn't real, I'll compromise you're only 50% wrong.

The side of the road you drive on could be a good start.

Where are you getting this information from?

There are 8 major taxa in taxonomy starting with domain. By only the 7 you mention, there's no way to squeeze in bacterias the way they are currently classified ;) Also, something like the flexibility you criticize taxonomy for lacking is actually present. Each taxon like family can, if needed, be split up into several ranks, for example you could take the family step and substitute it for all these levels :

Superfamily
Family
Tribe
Subtribe

I think this pretty much invalidates your criticism.

Then I will say Giraffa camelopardalis IS NOT camelopardus which must be regarded as a True statement and add that the name camelopardalis only exists in the minds of people - the imagination as it were - and therefore my previous statement is also true. ;)

edit :

By the way, where did you learn about the cabbit and camelopardus? Was it something you learned during your education?

I still hold my ground, but, you are right about plants. Even so, all plants diverge genetically until they can no longer form hybrids (without the help of humans, anyway). The same could be said about the Prokaryote kingdoms. Maybe there is a need to refine the system, but since even plant species head out on different evolutionary paths to the point where they can't form hybrids where they once might have tells me my "reproductive isolation" standard still holds. To me, anyway.

I think a problem with cladograms and that representation of phylogeny at the moment is that when branches are resolved, they only allow for dichotomous divergence. There's only one evolutionary event in a cladogram and it's something splitting up to form two somethings.

I guess some might be confused about what I'm writing about so here's a cladogram :

http://nimravid.files.wordpress.com/...am-options.jpg

It's essentially a hypothesis about the evolutionary history and relationships between taxa such as species. A problem with it is branches can only split, there's only divergence - and no convergence. In nature, plant species sometimes converge to form new species.

duga, you've got a good point about species eventually diverging anyways, but just like they will keep diverging, they will also keep converging in the future so the problems I pointed out with the sexual isolation definition still won't go away. Furthermore, it doesn't seem practical that the species definition will be valid eventually, possibly some million years in the future. It seems to me like a species definition should be applicable here and now.

You make a good point about evolutionary paths, though. I personally like the idea that a species is something on an evolutionary "path" which is different from that of other species, an idea that - as you also mention - is applicable even if there's a little flow of genetics with other groups.

That would only apply to Cladogenesis, but what about Anagenesis? If you take in consideration the working hypothesis of Evolution, a species on a "path" may be entirely difference from the beginning of it's evolutionary "path" to the end it's evolutionary "path."

It would apply to both. Though one results in a split while the other results in a replacement, they both have the same result by our current "species" definition: reproductive isolation. Though it may be harder to see on a cladogram, it still holds...so I fail to see your point.

I agree with duga and would argue further that it actually holds up well for every monophyly, potentially every taxonomic level and some new ones we don't have yet. If you did a genetic analysis of different populations of a species and created a cladogram, you could find that each population forms a monophyletic group, even if there is gene flow. It's a sign that the populations are diverging, they are on different evolutionary paths - although that wouldn't mean they won't converge again some time in the future.

Similarly, you could do it on a larger scale and say that wasps have taken a different evolutionary path from beetles, which is basically what taxonomy does already.

Then you can have a new way of defining species not yet discussed in this thread based on genetic analysis. At the moment, results vary with markers used, but imagine a future where it's easy to do whole-genome sequencing and you had a sequencer where you could just put in some rat, cat and dog DNA. They should come out as separate monophylies for you to base species on which is actually a proper result and not just something theorized, so it has something going for it. It would require some equipment we don't have yet and a helluva lot of processing power if you are to compare whole genome sequences of many species, but it might still be just a matter of time.

It would still run into the problem of changes between many species being transitional though and where to draw the line is a problem. Rats, cats and dogs are easy, but what about the brassicas? Still sounds like it could be an effective way of defining species to me.

So are you agreeing with me by saying that the critea of defining a species should have two qualifiers:

1. hybridization barriers
2. genetic information

Whether there are other qualifiers, I can't think of any at the time, but I think at least both should be taken in consideration.

On what grounds are you making these arguments? Do you know what these things mean? Hybridization barriers result in reproductive isolation...and well...every living thing has genetic information. Which is exactly what tore and I have been talking about. And you have been disagreeing with. Maybe you can expand a bit because I am not getting your point.

No to hybridization barriers. I've already written a few points now about why I don't think it's a very good criteria in previous posts.

Genetic information? I'm slightly confused as to what you mean.


In my latest post, I mentioned it might be possible to determine species based on sequencing data. It's not a new idea, it's already being done. However, I mentioned that in the future, it might be possible to do whole-genome sequencing quickly and if you store whole genomes of organisms in a database with some tremendous calculating powers, that could be used to infer phylogeny on a grand scale.

A bit like this (for ghost insects) :

http://www.nature.com/nature/journal...01313-f3.2.jpg

I argued that species should be based on monophylies. That would be one criteria, but you would need more of course. A monophyly is a group in a cladogram like the one above which includes all descendents of an ancestor. For example the bottom two are a monophyly. So are the bottom 3 or the bottom 7, but if you then leave out the last (Haaniella dehaanii), you would have a paraphyletic group - one which does not include all descendents of an ancestor.

Monophylies can be said to have been on an evolutionary path which differs from other monophylies - which is why they form monophylies in the first place. It does not mean sexual isolation which promotes divergence, but rather divergence itself - coupled with common ancestry. Evolution takes them in a different direction from other monophylies. A different direction/divergence basically means their genes are becoming different.

This is perfect. This would probably be the end all for the species debate, but as you said, we simply lack the tools to sequence whole genomes as rapidly as would be needed for this level of bioinformatics. We are getting close, though! We have several organisms sequences (including humans, obviously) and can now use those existing genome sequences to "map" another individual's genome to determine if they have certain genetic disorders or predispositions to one. This level of computation was unheard of and unthinkable even 15 years ago...10 more years and I'm sure we will be at the level needed for mass genome sequencing.

I agree. I keep feeling there's a huge snag somewhere, but all I can see at the moment are problems that look/sound like they can be overcome, but that's no different from and perhaps even slightly better than for the other species definitions. :)

Comparing whole genomes would need some smart algorithms and so on so you'd need some clever programmers and mathematicians. However, even if there's a problem with an algorithm used to define species, at least you can just improve the algorithm and run it again. Although one should definetly try and find one best way to do it, you could easily define species in a number of different ways with something like this. For example, one could use different algorithms or perhaps restrict analysis to only certain bits of DNA (f.ex mitochondrial DNA) and so on.

Yes, and depending on the organism being studied, I don't think it is nearly as difficult as it would first seem. Most coding regions of DNA (exons) are flanked by repetitive introns and mRNA primers making the task of finding protein coding genes that much easier. No doubt, there are still exceptions and problems that will arise but my point is that it is entirely possible and has some very exciting applications.