Paleo Parts and Identification Question

[Jegelewicz]

I am working on the Alabama Vertebrate Paleo data and there are a bunch of trace fossils (trackways) that have reverse sides and multiple species identified on them. Before I rush to catalog, I'd like some advice on the best way to handle such things. Here is what I am thinking of doing.

Bulkload them initially with the "main" identification - what AL already had in their data as a scientific name - and a single part "trace fossil"

Bulkload additional identifications for the other species found in the trackway. (Can I have all the IDs be "accepted"?, cause they would be...)

Place any comments about the trackway such as "Undichna on the reverse side" in collection object remarks.

Anything else I should consider? @dperriguey @dustymc @mbprondzinski

[dustymc]

The usual advice to catalog the item of scientific interest aside, http://handbook.arctosdb.org/how_to/How-to-Use-Complex-Identifications.html is appropriate. Just link one identification to multiple taxa.

[Jegelewicz]

So I should identify these things as "Trackway" and then add all of the taxon IDs in associated taxa (can I bulkload that?)?

Would that allow these specimens to show up in a search of higher taxa?

So if one of the associated taxa is Ursus arctos, would the specimen show up in a search of mammalia?

[dustymc]

Trackway

I'd prefer something a little more descriptive, but that'll work.

bulkload

No, but I can help figure something out.

search

Yep.

http://arctos.database.museum/SpecimenResults.cfm?taxon_name=Cetartiodactyla&collection_id=76

finds

http://arctos.database.museum/guid/UAM:EH:0610-5898

because it's linked to http://arctos.database.museum/name/Rangifer (screenshot above) which contains

[dperriguey]

I'm really glad you brought this up. I've not understood how to deal with this in the past. I've just held off. I think I'll start getting some of our ichnofossils in Arctos. The problem I have is when they are really vague: burrow, vertebrate foot print. The reality is that the cast of the foot is never something that can "actually" be linked to the higher taxa. You'd have to have the foot fossilize with the cast (I'm unsure if this has ever been found). More likely is a mass death with some impressions preserved alongside the assemblage. Otherwise, the classifications are indirectly shape specific. For inverts it's a little bit easier when you find the fossil in the burrow, but there is still a whole mess of classified ichnofossils that are only classified based on modern analogues (and the ichnologists still debate...and did that animal actually make that burrow or just fall in it?).

I think a separate classification tree for ichnofossils is the most appropriate path for this problem. This tree could be linked to the ichnofossil classifications to make them come up in a search. Unless there is a better plan that's already in place that I'm unaware of??

[dperriguey]

Well, never mind. I guess I'm not allowed to use "trace fossil" for parts.

[dustymc]

I'm fairly sure that there are ichnotaxa in Arctos. I don't see any particular reason they'd need to be in their own source. The relationship between ichnotaxa and "real" taxa is a problem for someone else - Arctos just needs the name, and has a place for more if it's available.

You can use multiple taxa in identifications. If you find a fossilized chicken foot in a Eubrontes trackway you could just use "Eubrontes and Gallus gallus" or something.

"Really vague" isn't a problem either - worst case, something like "unidentifiable {burrow or root impression or maybe just a weird rock}" fits.

Why do you think you're not allowed to use trace fossil?

[Jegelewicz]

The problem I have is when they are really vague: burrow, vertebrate foot print.

Answer - ID them as Animalia, or Vertebrata. Vagueness is a signal that someone needs to do more research on this thing! But you can label it as something even if it is just "interesting rock".

The reality is that the cast of the foot is never something that can "actually" be linked to the higher taxa. You'd have to have the foot fossilize with the cast (I'm unsure if this has ever been found). More likely is a mass death with some impressions preserved alongside the assemblage. Otherwise, the classifications are indirectly shape specific.

I disagree. If the cast is of a mammoth foot, then identify it as such. In extant species collections we identify as "Peromyscus" feces that was in the animal, but isn't THE animal is it?

You can use multiple taxa in identifications. If you find a fossilized chicken foot in a Eubrontes trackway you could just use "Eubrontes and Gallus gallus" or something.

We need to talk about this though. If I proceed as per the parka example above, I am not able to designate who identified what and when (or changed the ID later). This is a problem as traces may be identified by different experts at different times. So for these weird things that have multiple species included in one "part" we need to have the functionality of the biological identification system and be able to have multiple "accepted" IDs or some solution that approximates that. Suggestions?

[dustymc]

cast is of a

I don't completely understand it, but there's some formal-ish distinction between ichnotaxa and critter-taxa. I think in my (horrible) example they'd still call the impression Eubrontes on the not-so-off chance that the chicken was somehow not in fact responsible for the track.

https://doi.org/10.1016/j.jafrearsci.2011.01.003 may be a better example.

who identified what and when

If this is what I think it is, the wrong thing is being cataloged; tradition is getting in the way of good science. As always, the advice is to catalog the item of scientific interest. If someone identified a piece of some sort of conglomerate, then that piece should be recataloged so it can be cited properly (which also sets up the formal relationships to identifiers and methods and such). It doesn't matter if the specific bit/area/fossil/track/whatever is already cataloged as part of the assemblage, there's no reason to try and pry it off or anything, just give it a useful identifier so it can be properly treated as the item of scientific interest instead of a component of something else.

[campmlc]

That was my thought - why not give each item in the assemblage a different catalog number? And then relate them all together with relationships and a common shared ID. Then each one could have it's own identification history and who/when/why. I know the rock probably already has a number. But this is how we deal with lots (of fish, bugs etc) that need to be split because one individual is pulled for measurements or gene sequencing etc, and the same could work here.

On Mon, Feb 4, 2019 at 1:28 PM dustymc notifications@github.com wrote:

cast is of a

I don't completely understand it, but there's some formal-ish distinction between ichnotaxa and critter-taxa. I think in my (horrible) example they'd still call the impression Eubrontes on the not-so-off chance that the chicken was somehow not in fact responsible for the track.

https://doi.org/10.1016/j.jafrearsci.2011.01.003 may be a better example.

who identified what and when

If this is what I think it is, the wrong thing is being cataloged; tradition is getting in the way of good science. As always, the advice is to catalog the item of scientific interest. If someone identified a piece of some sort of conglomerate, then that piece should be recataloged so it can be cited properly (which also sets up the formal relationships to identifiers and methods and such). It doesn't matter if the specific bit/area/fossil/track/whatever is already cataloged as part of the assemblage, there's no reason to try and pry it off or anything, just give it a useful identifier so it can be properly treated as the item of scientific interest instead of a component of something else.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ArctosDB/arctos/issues/1900#issuecomment-460400453, or mute the thread https://github.com/notifications/unsubscribe-auth/AOH0hM_0mo7oPjrx6D9dX4Yxj9zX0pckks5vKJfegaJpZM4afsPP .

[Jegelewicz]

Let's see if I can convince the fossil curators....

[dperriguey]

Why do you think you're not allowed to use trace fossil?

I tried and it didn't work. I have a specimen in our browse/edit that I've given the ID Animalia but when I try to use trace fossil for part name (I got from http://arctos.database.museum/info/ctDocumentation.cfm?table=ctspecimen_part_name&coln=ES) I get this response:

I disagree. If the cast is of a mammoth foot, then identify it as such. In extant species collections we identify as "Peromyscus" feces that was in the animal, but isn't THE animal is it?

These are not the same thing at all. Paleontologists give tracks a different name than their most likely source. The name effectively means "Dimetrodon like foot print," its name is Dimetropus. If it is a total cast of the foot that's different. Check this article out for my point: https://geoinfo.nmt.edu/tour/federal/monuments/prehistoric_trackways/home.html

I definitely name things like a bivalve cast with the actual taxonomic name if all the morphological characteristics are there, but I note that it is a cast and not a body fossil; that is I have the full(ish) body cast to make this determination.

Regardless the points Dusty made about leaving that to the experts is the way to go. If it is not identified to some lower classification, I'd just leave a note that it is a mammal like reptile foot print or something like that. We don't even name fossil poop with the name of the dinosaur or deer mouse it likely came from; it has to have a different classification.

[dustymc]

Is that an ES specimen? Is it still in the bulkloader?

[dperriguey]

I'm pretty sure we are Earth Sciences. Our name is UNM:ES. Yes it is still in the brows/edit table under acotter.

[dperriguey]

Key is 96506588

[dperriguey]

The part code table said ES and I assumed that meant Earth Sciences. Is that right?

[dustymc]

UAM@ARCTOS> select ':'|| part_name_1 || ':' from bulkloader where collection_object_id=96506588;

':'||PART_NAME_1||':'
------------------------------------------------------------------------------------------------------------------------
:trace fossil :

1 row selected.

get rid of the trailing whitespace and it should work.

[dperriguey]

I'll be sure to check that next time. Stupid spaces. Thanks! No error now.

[Jegelewicz]

Paleontologists give tracks a different name than their most likely source. The name effectively means "Dimetrodon like foot print," its name is Dimetropus. If it is a total cast of the foot that's different. Check this article out for my point: https://geoinfo.nmt.edu/tour/federal/monuments/prehistoric_trackways/home.html

So identify it as "Dimetropus"!

[dperriguey]

I always identify things as they are identified first. Using A {string} method. If I am a trackway expert, I might change the identification. I'm not though. So I leave that to an expert. But to keep it visible I would make it Dimetropus {Dimetrodon trackway} if that's how it was identified. Someone might have identified it as a rock. Then I am asked to enter it into a database. I enter it in as Bivalvia {rock}. Maybe I have time to identify the bivalve later. I go back and give it a second identification of Euspira cragini.

I have a problem. Although I don't have a better way to do this, I'm calling a burrow Animalia. It's not an animal. It's the trace of an animal. For now I'll keep doing this because I need to get records in, but this is not a long term solution.

[Jegelewicz]

I always identify things as they are identified first. Using A {string} method. If I am a trackway expert, I might change the identification. I'm not though. So I leave that to an expert. But to keep it visible I would make it Dimetropus {Dimetrodon trackway} if that's how it was identified. Someone might have identified it as a rock. Then I am asked to enter it into a database. I enter it in as Bivalvia {rock}. Maybe I have time to identify the bivalve later. I go back and give it a second identification of Euspira cragini.

This all sounds reasonable and I don't see a problem. It is exactly what I would do.

I have a problem. Although I don't have a better way to do this, I'm calling a burrow Animalia. It's not an animal. It's the trace of an animal. For now I'll keep doing this because I need to get records in, but this is not a long term solution.

In my opinion, when you use the A {string} identification, you are saying this is {string} and A is associated with it, not necessarily IT. Does that make sense? I mean, this parka is not a mink, it just has parts of one. I do think it would be a good idea to add a remark in your burrow case, but I am also wondering if we need a new Nature of ID type - maybe "trace evidence".

[dperriguey]

Thanks. I'll stick to A {string}.

The idea of nature of ID for paleo ichnotaxa is weird because then you need a fossil nature of ID, biologic nature of ID, etc. Nature of ID is still not well defined on Arctos documentation.

I prefer to use parts for this. Using parts you can say trace fossil then leave a remark stating what type of trace fossil. Maybe editing trace fossil to trace fossil/ichnofossil would be useful for searching or something.

[Jegelewicz]

So, I think we have the ichnotaxa worked out? Now we need to document.

Back to my original post though - please read this: https://github.com/tdwg/dwc-qa/issues/119#issuecomment-385222655

[Jegelewicz]

It doesn't seem good to have duplicate catalog numbers (which I believe is what John is suggesting?)

[dperriguey]

I just use associated taxa. This is arbitrary, but searchable (I think).

[dperriguey]

Op, just checked. I can't find where it's searchable.

[dustymc]

There are two general possibilities: catalog the item of scientific interest (in which case tuco's comments will work themselves out and never become an issue), or catalog something else and struggle with everything forever. John's comments regarding duplicate catalog numbers is just a side effect/inevitability if you're applying one catalog number of a bunch of "items of scientific interest."

Searching by taxonomy is documented here: http://handbook.arctosdb.org/documentation/taxonomy.html

[campmlc]

Is using relationships an option? They are searchable.

On Tue, Feb 19, 2019 at 11:33 AM dperriguey notifications@github.com wrote:

Op, just checked. I can't find where it's searchable.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ArctosDB/arctos/issues/1900#issuecomment-465254703, or mute the thread https://github.com/notifications/unsubscribe-auth/AOH0hJbmYQ5nH9Ar68DMOemoS80Hu2LNks5vPENwgaJpZM4afsPP .

[dustymc]

relationships

Relationships can be (should be, if there's any chance anything has been cited) used to link https://github.com/ArctosDB/arctos/issues/1900#issuecomment-460400453 back into a source lot. Is that what you're referring to?

[campmlc]

Yes. I haven't been following the entirety of the discussion, but it seems to me that the suggestion is to give each taxon in a rock cast a separate catalog number makes sense, even if the rock itself has a separate catalog number. Then use "collected with" or some other more specific relationship analogous to "same lot as" to link them. I agree with Dusty's comment below:

If this is what I think it is, the wrong thing is being cataloged; tradition is getting in the way of good science. As always, the advice is to catalog the item of scientific interest. If someone identified a piece of some sort of conglomerate, then that piece should be recataloged so it can be cited properly (which also sets up the formal relationships to identifiers and methods and such). It doesn't matter if the specific bit/area/fossil/track/whatever is already cataloged as part of the assemblage, there's no reason to try and pry it off or anything, just give it a useful identifier so it can be properly treated as the item of scientific interest instead of a component of something else.

On Tue, Feb 19, 2019 at 11:41 AM dustymc notifications@github.com wrote:

relationships

Relationships can be (should be, if there's any chance anything has been cited) used to link #1900 (comment) https://github.com/ArctosDB/arctos/issues/1900#issuecomment-460400453 back into a source lot. Is that what you're referring to?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ArctosDB/arctos/issues/1900#issuecomment-465257844, or mute the thread https://github.com/notifications/unsubscribe-auth/AOH0hPaTjC8SRrNmg2RqQdCKdiWsTEd1ks5vPEVGgaJpZM4afsPP .

[mbprondzinski]

Yes. I haven't been following the entirety of the discussion, but it seems to me that the suggestion is to give each taxon in a rock cast a separate catalog number makes sense, even if the rock itself has a separate catalog number. Then use "collected with" or some other more specific relationship analogous to "same lot as" to link them. I agree with Dusty's comment below: If this is what I think it is, the wrong thing is being cataloged; tradition is getting in the way of good science. As always, the advice is to catalog the item of scientific interest. If someone identified a piece of some sort of conglomerate, then that piece should be recataloged so it can be cited properly (which also sets up the formal relationships to identifiers and methods and such). It doesn't matter if the specific bit/area/fossil/track/whatever is already cataloged as part of the assemblage, there's no reason to try and pry it off or anything, just give it a useful identifier so it can be properly treated as the item of scientific interest instead of a component of something else. … On Tue, Feb 19, 2019 at 11:41 AM dustymc @.***> wrote: relationships Relationships can be (should be, if there's any chance anything has been cited) used to link #1900 (comment) <#1900 (comment)> back into a source lot. Is that what you're referring to? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#1900 (comment)>, or mute the thread https://github.com/notifications/unsubscribe-auth/AOH0hPaTjC8SRrNmg2RqQdCKdiWsTEd1ks5vPEVGgaJpZM4afsPP .

This seems logical to me, especially since we are the culprits!

[dustymc]

http://arctos.database.museum/info/ctDocumentation.cfm?table=CTID_REFERENCES

"same lot as" should be used for splitting lots. If you do pry an individual (or whatever) off prior to cataloging, "collected with" might make more sense.

[Jegelewicz]

Yes. I haven't been following the entirety of the discussion, but it seems to me that the suggestion is to give each taxon in a rock cast a separate catalog number makes sense, even if the rock itself has a separate catalog number. Then use "collected with" or some other more specific relationship analogous to "same lot as" to link them. I agree with Dusty's comment below:

I agree with this treatment - BUT I think the first item cataloged should be "rock" or some version thereof to indicate the physical object collected, with all the other cataloged items being identified as whatever they are identified as. Then, we need a new relationship, "part of" perhaps, so that all identified organisms on the object can be related to the object (and thereby to each other.

Make sense?

The problem is we can't catalog anything without some association to a biological name (the A {string} formula) and I would prefer the first item NOT be given any taxa associations so that it becomes the "point of contact" for all of the individual identifications that I know about now, or may be generated in the future. How in the world do art and ethnology collections do this? Maybe I am just missing something...

[mbprondzinski]

To stir the pot further….https://en.wikipedia.org/wiki/Form_classification

From: Teresa Mayfield-Meyer [mailto:notifications@github.com] Sent: Tuesday, February 19, 2019 1:26 PM To: ArctosDB/arctos Cc: Prondzinski, Mary Beth; Mention Subject: Re: [ArctosDB/arctos] Paleo Parts and Identification Question (#1900)

Yes. I haven't been following the entirety of the discussion, but it seems to me that the suggestion is to give each taxon in a rock cast a separate catalog number makes sense, even if the rock itself has a separate catalog number. Then use "collected with" or some other more specific relationship analogous to "same lot as" to link them. I agree with Dusty's comment below:

I agree with this treatment - BUT I think the first item cataloged should be "rock" or some version thereof to indicate the physical object collected, with all the other cataloged items being identified as whatever they are identified as. Then, we need a new relationship, "part of" perhaps, so that all identified organisms on the object can be related to the object (and thereby to each other.

Make sense?

The problem is we can't catalog anything without some association to a biological name (the A {string} formula) and I would prefer the first item NOT be given any taxa associations so that it becomes the "point of contact" for all of the individual identifications that I know about now, or may be generated in the future. How in the world do art and ethnology collections do this? Maybe I am just missing something...

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/ArctosDB/arctos/issues/1900#issuecomment-465275110, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ApNlkNFHPpfXCXNAG_nF84cSvWXRLXemks5vPE_BgaJpZM4afsPP.

[dustymc]

"rock" or some version thereof

Not a problem, although I'm not sure I see the point for new material. Something that's been cataloged (and potentially cited) as the lot should definitely be cataloged as the lot to preserve the link between publications and specimens.

new relationship, "part of"

How is that not "same lot as"? I don't care much about the terminology, but I think you're trying to introduce a second way of saying one thing.

can't catalog anything without some association to a biological name

http://arctos.database.museum/name/unidentifiable exists for just this reason. (Even if it's mis-used for things that can be identified - I'd wager that http://arctos.database.museum/guid/UAM:Herb:235303 is a plant!)

[dustymc]

@mbprondzinski I don't see a problem with that. http://handbook.arctosdb.org/documentation/taxonomy.html. If someone at some point thought it was taxonomy, then it's taxonomy in Arctos. If they didn't (eg, it's a working name) then the A {string} formula is available for that informal usage.

[mbprondzinski]

How about this? What do we call the substrate? Rock? Shale? Sandstone? The Ichno fossikers don’t seem to document the substrate in their records, at least not in our collection. https://link.springer.com/chapter/10.1007%2F978-3-642-65923-2_4

From: dustymc [mailto:notifications@github.com] Sent: Tuesday, February 19, 2019 1:35 PM To: ArctosDB/arctos Cc: Prondzinski, Mary Beth; Mention Subject: Re: [ArctosDB/arctos] Paleo Parts and Identification Question (#1900)

@mbprondzinskihttps://github.com/mbprondzinski I don't see a problem with that. http://handbook.arctosdb.org/documentation/taxonomy.html. If someone at some point thought it was taxonomy, then it's taxonomy in Arctos. If they didn't (eg, it's a working name) then the A {string} formula is available for that informal usage.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/ArctosDB/arctos/issues/1900#issuecomment-465278444, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ApNlkPHZFLlJz_NYAV3aKJ2pM6vlGIJDks5vPFHxgaJpZM4afsPP.

[mbprondzinski]

How about referring to the substrate as “sedimentary rock” as a general term and catalog that as such?

From: dustymc [mailto:notifications@github.com] Sent: Tuesday, February 19, 2019 1:35 PM To: ArctosDB/arctos Cc: Prondzinski, Mary Beth; Mention Subject: Re: [ArctosDB/arctos] Paleo Parts and Identification Question (#1900)

@mbprondzinskihttps://github.com/mbprondzinski I don't see a problem with that. http://handbook.arctosdb.org/documentation/taxonomy.html. If someone at some point thought it was taxonomy, then it's taxonomy in Arctos. If they didn't (eg, it's a working name) then the A {string} formula is available for that informal usage.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/ArctosDB/arctos/issues/1900#issuecomment-465278444, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ApNlkPHZFLlJz_NYAV3aKJ2pM6vlGIJDks5vPFHxgaJpZM4afsPP.

[dperriguey]

Mary Beth's point about the substrate is very important to the few ichnologists out there. Calling it sedimentary rock makes the point of even mentioning the substrate useless though. Almost all fossils are preserved in sedimentary rocks.

When I have a gastropod steinkern (internal cast) with bivalves that have encrusted on it, I choose what to identify it as. If someone collected it and named it gastropod and made notes that there are associated bivalves, I just follow by putting bivalves in the associated taxa. I do the same thing if the collector identified it as bivalve and associated gastropod. It all depends on the focus of the collection. If you have an assemblage with multiple taxa within one large rock specimen, I think Dusty's "unidentifiable" works well. As long as the associated taxa are searchable. So far I've not been able to get this result. Do I have to go in and manually make the associations to each record?

Both the following cataloging methods are something I encounter all the time. Sometimes I have a rock with multiple taxa in it and the whole rock was given a number. Other times I find a large rock with multiple catalog numbers associated with it.

Where are the relationships in Arctos? Collected with or same lot as makes better sense to me than associated taxa unless I'm making a general note.

[Jegelewicz]

new relationship, "part of"

How is that not "same lot as"? I don't care much about the terminology, but I think you're trying to introduce a second way of saying one thing.

lots can be separated - these things cannot be removed from the "rock", but if everyone else is OK with that, I won't argue. @dperriguey @mbprondzinski

[Jegelewicz]

When I have a gastropod steinkern (internal cast) with bivalves that have encrusted on it, I choose what to identify it as. If someone collected it and named it gastropod and made notes that there are associated bivalves, I just follow by putting bivalves in the associated taxa. I do the same thing if the collector identified it as bivalve and associated gastropod. It all depends on the focus of the collection. If you have an assemblage with multiple taxa within one large rock specimen, I think Dusty's "unidentifiable" works well. As long as the associated taxa are searchable. So far I've not been able to get this result. Do I have to go in and manually make the associations to each record?

Both the following cataloging methods are something I encounter all the time. Sometimes I have a rock with multiple taxa in it and the whole rock was given a number. Other times I find a large rock with multiple catalog numbers associated with it.

Where are the relationships in Arctos? Collected with or same lot as makes better sense to me than associated taxa unless I'm making a general note.

@dperriguey after cataloging the large rock as "unidentifiable {fossiliferous rock}", I would then catalog each identification with a new catalog number using identical collecting event, locality, collectors, etc., then add to the new cataloged item an other_id that is the catalog number of the "unidentifiable {fossiliferous rock}" with the relationship "same lot as". Everything searchable and all identifications include a "by who".

[Jegelewicz]

@dperriguey don't be afraid of assigning new numbers - we can talk to Cori about it if necessary....

[dustymc]

substrate

Sounds like geology to me, but the real answer probably depends on who cares and why.

Arctos will handle mineral taxonomies - if you have such a thing we can load it, if not the A {string} formula will suffice as much as any string can.

putting bivalves in the associated taxa

That's fine until someone cites the bivalve, at which point we're failing to adequately link publications to material.

result

Example?

[Jegelewicz]

Arctos will handle mineral taxonomies

Excellent point and we should work on this one - we are going to have some mineral collections coming in so it will happen eventually.

BUT - will we need a new set of ranks? I have no idea how mineral taxonomy works!

[dperriguey]

@Jegelewicz ok, I just didn't know that was available. Makes total sense to do the A {string} and add these relationships to other identifications.

@dustymc I just didn't know how to assign relationships. Substrate does not equal geology. Substrate has the same definition in paleontology as it does in biology with regard to taxa that live on or in some sedimentary mass (or other taxa when encrusting). Ichnologists care for the level of anatomical resolution that is preserved (different depending on grain size). Paleoecologists care for environmental context.

@Jegelewicz @dustymc Minerals might not be as hard as petrology (or otherwise known as rocks!) Minerals are going to be grouped into sulfides, silicates, phyllosilicates etc. Petrology is not going to be as hard to classify as geology (=lithostratigraphy !!!), but similarly as hard to classify as species but an almost inverted hierarchy (Sandia granite).

[campmlc]

Relationships are other IDs. You just add an additional column of OTHER_ID_REFERENCES_1" to a bulkload file, or you choose the dropdown in data entry. Or use the Other ID bulkloader for existing records: https://arctos.database.museum/tools/BulkloadOtherId.cfm. Creating the one way linkage this way will autocreate the reciprocal, even outside your collection, with an email to the corresponding collection manager who can choose to approve. You can also use the "pull" function next to Other IDs in data entry to pull data from an existing record into a new record - makes creating these kinds of relationships really easy.

On Tue, Feb 19, 2019 at 1:42 PM dperriguey notifications@github.com wrote:

@Jegelewicz https://github.com/Jegelewicz ok, I just didn't know that was available. Makes total sense to do the A {string} and add these relationships to other identifications.

@dustymc https://github.com/dustymc I just didn't know how to assign relationships. Substrate does not equal geology. Substrate has the same definition in paleontology as it does in biology with regard to taxa that live on or in some sedimentary mass (or other taxa when encrusting). Ichnologists care for the level of anatomical resolution that is preserved (different depending on grain size). Paleoecologists care for environmental context.

@Jegelewicz https://github.com/Jegelewicz @dustymc https://github.com/dustymc Minerals might not be as hard as petrology (or otherwise known as rocks!) Minerals are going to be grouped into sulfides, silicates, phyllosilicates etc. Petrology is not going to be as hard to classify as geology (=lithostratigraphy !!!), but similarly as hard to classify as species but an almost inverted hierarchy (Sandia granite).

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ArctosDB/arctos/issues/1900#issuecomment-465301722, or mute the thread https://github.com/notifications/unsubscribe-auth/AOH0hJTsCtjo8K_DLn1HY21ETT4-oMSWks5vPGHLgaJpZM4afsPP .

[Jegelewicz]

Minerals might not be as hard as petrology (or otherwise known as rocks!) Minerals are going to be grouped into sulfides, silicates, phyllosilicates etc. Petrology is not going to be as hard to classify as geology (=lithostratigraphy !!!), but similarly as hard to classify as species but an almost inverted hierarchy (Sandia granite).

Would this include something like "Jagger Coal"? I have been struggling with how to handle this term, which isn't really lithostratigraphy (I don't think...), but is important for describing what the specimen is.

Maybe these would work just like the 4.0 for man-made items classification terms:

Coal, Jagger Granite, Sandia

So I could ID the 'chunk of rock' as Coal, Jagger (assuming we relax the name constraints) and then the tracks in the chunk as described above with appropriate relationship...?

[dustymc]

@Jegelewicz yes I think mineral taxonomy is something like the man-made thing. I don't think there are ranks involved, but there seem to a bunch of things that people don't quite use so it's possible we'll find ranked data.

Hooking a specimen to the string "Jagger Coal" + unidentifiable lets you find things identified to "Jagger Coal."

Hooking a specimen to the data object

rocks -- black rocks -----shiny black rocks ------ burney shiny black rocks --------coal -----------Jagger Coal

lets you find the specimen by any of that junk, and introduces a place for relationships between names, common names, publication, etc. It also provides one way of spelling Jagger Coal, where the string-approach does not have that limitation.

[dperriguey]

The Jagger Coal is a coal bed within the Mary Lee Coal Group, a local lithostratigraphic unit (Upper Pottsville age). These are the odd balls; regional and local lithostratigraphy. They are related to mineral/petroleum or paleontological resources within some smaller area. You have to do some investigation to figure out how it relates to the international chronostratigraphic chart. See: https://babel.hathitrust.org/cgi/pt?id=mdp.39015078539924;view=1up;seq=7 and https://pubs.usgs.gov/bul/1182b/report.pdf . Regional chronostratigraphy example: http://www.stratigraphy.org/upload/OrdChartHigh.jpg

I'd say if they were in a mineralogical collection (although it's not a mineral it's a sedimentary rock with many minerals and organic matter!) then keep them in there. It should be in a petrology collection though.

[Jegelewicz]

Not to rehash but please read https://github.com/ArctosDB/data-migration/issues/53#issuecomment-560941969

[Jegelewicz]

I think this is applicable to this issue - https://github.com/tdwg/dwc/issues/314#issuecomment-827707288