Identification Of New Species

[Jdeutsch]

I am curious- how can one be sure when a fossil is identified, if it is just part of normal biological variability versus a new species? On the forum there is often a variety of opinions on ID based on phtographic data. Ultimately, the species is identified by "eyeball"

What kind of rigor is applied in the paleontology literature? What ability is there to search a database to make sure the same "pattern" identified in rock hasn't been called something else in the literature?

To complicate things- number of species x time suggests the fossil record must have thousands fold or more variability compared to what is alive now. Huge numbers to deal with.

My working definition of species is that it is a restricted breeding group in nature. For example- pintail ducks and mallards- different species- don't breed in the wild. Quite happy to interbreed in a cage. Wolf, dog and coyote do the same, and perhaps may mix in nature if given the opportunity but general behavior keeps them apart, etc. A german shepard looks more like a wolf than a german shepard looks like a dachshund- but wolf is a different species from a shepard, and the dachshund is the same as a shepard. Humans with Achondroplasia have similar appearence and skeletal characterisitics but are simply homo sapien with mutation in structural development.

Sometimes seems hard enough to identifiy species on living specimens, with DNA sequences to boot. Morphological information could be misleading in my mind. Humans with Achondroplasia all have similar appearence but are very much homo sapien.

How do you experts do it?

[pleecan]

From my limited experience ....Identification via photos are just a "best guess". If the photos are interesting enough.... Paleo guys want to handle and examine fossil first hand... and use a number of methodology to ascertain the best guess identity... sometimes the fossil remains an unknown and undescribed,.... I will let the experts comment further...

[Stingray]

The PALEONTOLOGY section is basically a replica of your thoughts and points,, however it has a bit more explenation of how Morphology is determined.....I guess its all best summed up by saying "the only thing certain is uncertantity" ...............

http://www.bluelion.org/basicinfo.htm

P.S. I follow this theroy myself but others may not.......

" All organisms have some variation among their populations, different sizes, body shapes, etc. To propose a fossil is a new organism, there should be one of two factors present. One is amount found. Usually a single fossil, or specimen, is not enough to prove this was a separate species. In most cases a sample must be found. A sample is a large collection of fossils sharing similar characteristics. A sample can be evidence that a fossil specimen's appearance is not only variation."

Edited April 1, 2012 by Stingray

[Jesse]

There is a system to identify various types of fossil species. This involves a process called morphometric analysis or landmark analysis. For this to work someone would need to gather a large volume of a given species and take a series of linear measurements (length, width, height, etc) to establish the variation within a species. The landmark analysis is similar but slightly more complicated to explain, and would be a very long answer. I will attach a publication that does a beautiful job of explaining morphometric analysis in several forms.

The kicker to this is that the amount of acceptable variation is very subjective and varies person to person. For example Joe can analyse a specimen and say it belongs to X species, while Jim can measure the same specimen and say that it does not lie within the predetermined range of variation and assign it to new species. It boils down to lumpers and spliters and the repeatability of specimen analysis.

Shark morphometric analysis paper

http://www.tandfonline.com/doi/full/10.1080/02724630903409055

And just for a point of clarification, mallards and pintails crossbreed quite often in the wild, with beautiful results!

[Jdeutsch]

To Jesse

The paper is interesting-the statistical overlap is also interesting. By definition (in my mind) a venn diagram of species shouldn't overlap, but the morphogenic measures do. The pintail-mallard picture is great- and shows species overlap-assuming the crossbreeding occurred in the wild-but it also raises the question-if crossbreeding is common in the wild-why are the species seperate? What species is in the picture? How did the species seperate? Will the species stay seperate over time? What duck will that Drake choose to breed with? Which gets me back to fossils-it is easy to tell a crinoid from an ichthyosaur- but what about distinguishing various brachiopods? Are the boundries between species fuzzy? How much of fossil morphology differences occur based on the local environment and not through "DNA differences"?

This comes to mind when I look at hashplates of shells and feel absolutely lost as how to get a grip on ID. I'm trying to get a feel for what trained people think.

[Jesse]

You have some great questions that are not so easy to answer!

As far as the duck, it is easy to id it as a cross when it is seen as above, but what if all you had was the skeleton? That is when a solid comparative collection comes in handy. Most "professional" paleontologists have access to vast comparative collections where they can compare specimens to known species to help narrow down an id. In the case of the pintail-mallard, a real id may be impossible from the skeleton, absent DNA analysis.

I work exclusively with vertebrate material, so have a very limited knowledge of invertebrate material. I do know that to positively id brachiopods, for example, you really need a complete and exposed specimen as the folds, hinges, ribs, and growth lines can be very diagnostic. Fragments and hash plates can be very problematic for positive id's, especially for me. On the plus side, most inverts only occupy a certain area, for a certain period of time, with little variation. for this reason many invert species' are used for index fossils to identify time periods and even locations in some instances.

When I need help identifying a brachiopod I go to my boss/adviser and he helps me in the following method. First, is it an articulated, inarticulate, productid,etc? Next what is the stratigraphic reference or formation? Third, what is the scale of the brachiopod? And so on, each step narrows it down a bit more until a relatively good ID is made.

That long winded and most likely less than helpful answer brings me to this. It depends on the researcher who discovers the "type" specimen of a given species as to whether the specimen is a variation of a known species or belongs to a new genus or species.

[Auspex]

We are cursed/blessed with the need to understand, and compartmentalizing things in ever narrower pigeonholes helps, up to a point. Nature is not so restricted, and blurred boundaries are reality.

[Boneman007]

The first steps to describing a new species or genus:

Step 1: The easy step. Find it.

Step 2: Get on JSTOR.COM and download all papers from the formation, and the type of fossil you are hoping to describe.

Step 3: Read all documentation. Make sure it isn't already described. Become an expert on every part of the animal you are describing. Become a expert in scientific nomencature as it pertains to your specimen.

Step 4: Go to a museum in the area. Show your fossil to the curator of collections. Ask his opinion.

Step 5: Go to the local fossil society. Most large cities have one now. Show your fossil to the advanced members. Get their opinion.

Step 6: After all this, if you still believe that it is a new species, the actual work begins...

Be patient. It took 4 years of work, a PhD thesis by a Phd candidate, and a international publication by the top permian guy in the world for my massive permian reptile Acheloma dunni to be described...

It was worth it though! I can now say that I discovered the T-Rex of the Permian time period!

Edited April 2, 2012 by Boneman007

[jpc]

I wish I had time to discuss this one... this is an essay subject that is often discussed by paleo types at conventions over a glass or two of delicious adult beverage. As the duck/dog discussion shows, the idea of species is even hard to describe with things that we can actually see in the living form. Paleospecies.... now that's a duck of a different color.

Edited April 2, 2012 by jpc

[Boneman007]

] And just for a point of clarification, mallards and pintails crossbreed quite often in the wild, with beautiful results!

In paleontology, mallards, pintails, maltails (mostly mallard), and pinards (mostly pintail), would all be one species. unless there are structural differences that would differenciate them. I expect if you looked at the skeletons of all these types of birds they would be virtually the same. Edited April 3, 2012 by Boneman007

[Auspex]

In paleontology, mallards, pintails, maltails (mostly mallard), and pinards (mostly pintail), would all be one species. unless there are structural differences that would differenciate them. I expect if you looked at the skeletons of all these types of birds they would be virtually the same.

The osteological differences between species in the genus Anas can be vanishingly subtle, and it can take scrutiny of several associated bones to inspire any confidence in a diagnosis (very rare for fossil bird bones).

[Jesse]

In paleontology, mallards, pintails, maltails (mostly mallard), and pinards (mostly pintail), would all be one species. unless there are structural differences that would differenciate them. I expect if you looked at the skeletons of all these types of birds they would be virtually the same.

I have seen several duck fossils and you are correct, they are simply Anas sp. I was just pointing out that ducks do crossbred with some regularity in the wild.I suspect that for any real id of fossil material you would need a very complete skull.

[Jdeutsch]

I have found the replies here most interesting.

Which brings up the topic of mutations in structural genes that don't define speciation. What if there is an interbreeding population that carries a mutation. Would a paleontologist assume a new species for those with the mutation? I mentioned achodroplasia in the original post as it affects skeletal structure in a characteristic pattern http://www.newsrx.co...ndroplasia.html Would a similar gene in structural development lead paleontologists to believe they had a new species? Are there stuctural gene mutations analogous to this in all classes of living organisms- including brachiopods? Mendel had smooth and wrinkeled peas to describe his law of independent assortment. Different characterisitc morphology, same species. Are there examples of modern brachipods with populations containing a characteristic structural mutation, yet by DNA and behavior still be contained in the same species? Can paleontologists define presumptive evidence for structural gene mutations within a species in the fossil record?

Edited April 6, 2012 by Jdeutsch

[Auspex]

For genetic mutations to persist long enough to leave a fossil record, they must be, at least, benign in terms of reproductive success, and carry no metabolic cost. Any mutation that yields an ill-adapted lineage will not naturally last for long, unless they can access and exploit a new or vacant niche (in which case, they are on their way to speciation).