Using Darwin Core for Cataloguing Organism Photos

Post Reply
gcorbin
Posts: 86
Joined: 21 Aug 06 12:31
Location: Brisbane

Using Darwin Core for Cataloguing Organism Photos

Post by gcorbin » 02 May 21 14:27

Darwin Core defines a standardised set of fields to facilitate the sharing of information about biological diversity. It includes fields for Occurrence, Organism, MaterialSample, Event, Location, GeologicalContext, Identification, Taxon, MeasurementOrFact and ResourceRelationship so is extremely comprehensive. Photo Supreme natively supports Darwin Core, so ideally, people should be using Darwin Core fields when recording organism photos to allow simple interchange of the cataloguing data with others.

For my cataloging, I am currently only using the fields

Code: Select all

dwc:Taxon.dwc:scientificName
dwc:Taxon.dwc:vernacularName
by completing one or both fields for the organism in the photo.

There are many more fields which I should be completing to define the Taxon. For example, for an Australian native orchid I recently photographed, I should be completing all the Darwin core taxon fields as:-

Code: Select all

Kingdom: Plantae
Phylum: Tracheophytes
Class: Equisetopsida
Order: Asparagales
Family: Orchidaceae
Scientific Name: Acianthus exsertus
Vernacular Name:
Genus: Acianthus
Specific Epithet: exsertus
Scientific Name Authorship: R.Br.
Original Name Usage: Acianthus exsertus R.Br. 1810
Identified by: Graham Corbin
Date Identified: 02/05/2021
I do not currently complete all these fields due to the amount of work involved to do this by hand. I could however make an ‘Acianthus exsertus’ label which applies all the required Darwin core field data and I would just need to add the label to the photo to have all the Darwin core data applied in a single click. This would be ideal, except I now need to create a label for each organism I photograph. I have created a few labels to prove the concept works, but I haven’t followed this approach due to the amount of work involved to create all the labels.

Theoretically, I could download a list of organism species from an internet checklist site such as the Global Biodiversity Information Facility (https://www.gbif.org/) and programmatically create a label for each species to correctly fill out the Darwin core xmp. This would make my cataloguing of organism photos much easier as I just need to apply the species label, an identifier label and a location label and I have completed the basic Darwin core cataloguing of the organism.

I see two challenges with this approach.
  1. According to the Global Biodiversity Information Facility, there are currently 2,555,332 identified species, so we would need to create that many labels. We could reduce this by only including some kingdoms such as Animalia or Plantae depending on your interest, but the bottom line is there will be tens of thousands of labels required irrespective of whether we limit the organisms imported. Can the Photo Supreme database support this? Presumably, but this would need testing.
  2. Taxonomists regularly split, combine and restructure species, so we would need a method of periodically updating the labels to the latest taxonomy.
Is there any interest in looking at the possibilities of programmatically creating labels to complete the Darwin core XMP or any alternate method of completing the Darwin core XMP with Photo Supreme?

fbungarz
Posts: 1706
Joined: 08 Dec 06 5:03
Location: Arizona, USA

Re: Using Darwin Core for Cataloguing Organism Photos

Post by fbungarz » 03 May 21 5:21

Hi gcorbin,
I'd be interested in helping with this. GBIF is not necessarily the best source when it comes to taxonomy. It is more focused on specimens.
There are different resources for different groups of organisms. For Fungi for example, both Mycobank and Index Fungorum are the name repositories most up-to-date.
I would like to collaborate on this, ideally establishing some sort of Controlled Vocabularies that could be shared through the IDI resource repository.
But we should perhaps discuss the details not necessarily here on the forum. It may get fairly technical...
You could contact me via email.
Cheers,
Frank

Bongo
Posts: 60
Joined: 22 Mar 10 19:11
Location: Ingolstadt, Germany
Contact:

Re: Using Darwin Core for Cataloguing Organism Photos

Post by Bongo » 03 May 21 6:52

Hi Gcorbin and Frank,

Thanks for the commitment.
I think this large amount of data records for importing PSU is not realistic. You would have to be able to select the records for import. One would have to be able to consider the region of the species.
I'm curious.
Regards, Bongo
Photo Supreme V6 Single User Edition, Windows 10.

fbungarz
Posts: 1706
Joined: 08 Dec 06 5:03
Location: Arizona, USA

Re: Using Darwin Core for Cataloguing Organism Photos

Post by fbungarz » 04 May 21 18:32

I agree that it would make sense to focus on supplying controlled vocabularies for specific taxonomic groups and ideally for specific regions as well. But compiling those lists and keeping them reasonably up-to-date can be challenging. There are some groups (orchids, birds) that are very popular and thus reasonably well known, their taxonomy is a bit more settled and not changing all that frequently. But the amount of names is still overwhelming - even in these groups.

And Bongo is correct that a user might need only a small subset of those names, say all names from vascular plants in Germany, rather that the vascular plants of the world.

There are lots of resources like The Plant List, or Index Fungorum, some of them can even be accessed via API, but they not necessarily have sufficiently granular data, where those species actually occur. Some regional resources like the Lichen Consortium do have distribution data linked to specimens and tools built in to create species checklists for a particular region. But it is a bit of a challenge. Many regions world-wide are very poorly know, most of their species remain undiscovered ...

The irony: there are species going extinct as we speak and the planet is facing an enormous biodiversity crisis. But the majority of biodiversity remains essentially unknown, especially in species groups that are much less iconic than a tiger, birds or butterflies.

Compiling any such reference lists is therefore not just a technical challenge. For many regions of the world and most species groups the information available is not particularly accessible.

Still, compiling your own controlled taxonomic vocabulary within PSu is cumbersome - especially because you have to manually map the different taxonomic ranks to their corresponding DarwinCore fields.

I think it would already be a huge time-saver, if it was possible to import a checklist of species names with the taxonomic hierarchy mapped to their corresponding fields. This would be fantastic. But it is something that Hert would have to support. A format to import such a taxonomic thesaurus for a particular species group (perhaps filtered by region). Once such an import structure exists, we could then try to compile at least some useful lists for particular species groups.

It is a feature request that probably doesn't attract a huge audience of potential PSu customers. Most wildlife photographers that I am aware off are much more interested in capturing a great looking shot, hardly any are really interested to accurately name the frog or orchid that they photograph. So, the dedicated taxonomist using images for his research is a bit of an obscure niche audience. I am already more than grateful to Hert that he decided to fully support the DarwinCore, but I also do understand that him developing an import format to upload a taxonomic thesaurus as a controlled vocabulary might not be ranking very high on his priority list...

gcorbin
Posts: 86
Joined: 21 Aug 06 12:31
Location: Brisbane

Re: Using Darwin Core for Cataloguing Organism Photos

Post by gcorbin » 06 May 21 13:25

fbungarz, we have independently come to the same conclusion. Thinking about loading all species into Photo Supreme, I also conclude it would be a massive, never ending task which would be effectively impossible in practice, ignoring the practicalities of the massive database this would create.

A practical method of achieving the required result would be when an identified species photo requires the Darwin core fields to be populated, a lookup from an Internet resource could be made to retrieve the required information and then to populate the Darwin core fields. This method has the advantage that it requires no local species database but instead gets the Darwin core information from the internet on demand. As taxonomy changes, the Darwin core catalogued photos would not update but the photographer would need to review their taxonomy occasionally and update it as required using the same method initially used to populate the Darwin core information. This is a simple and practical method.

To implement this would require two things.
  1. Firstly and most importantly an internet source (or sources) of information to populate the Darwin core fields. This will be a challenge, but hopefully achievable. Without a source of implementation, we have no solution.
  2. Secondly, we need some code to retrieve the information from the internet and populate the Darwin core fields. This could be built into Photo Supreme and would require Herts support to implement, but as was pointed out, the number of users wanting this feature is probably low so it would need to be very simple to implement for Hert to be interested.

    This potentially could be also implemented as an external program which updates the database with the Darwin core data. This would also require Herts assistance and support on how to correctly update the database tables for this to be a successful method.

    Lastly, there are probably other options Hert would know which I have not considered for implementation.
I suspect the implementation code is the simplest component to achieve. Our first task is to identify the information source or sources for the Darwin core data. If/when we have a viable information source and understand what is required to retrieve that information, we can consider the best method for code implementation.

Sound reasonable or has someone got a better though?

fbungarz, have you looked at the resource API functionality to see if this gives a viable source of Darwin core information?

fbungarz
Posts: 1706
Joined: 08 Dec 06 5:03
Location: Arizona, USA

Re: Using Darwin Core for Cataloguing Organism Photos

Post by fbungarz » 06 May 21 17:14

Hi Gcorbin,
this forum used to include a private message board, where users could get in touch directly with one another. Unfortunately, this is no longer the case.
I don't think continuing this discussion, posting text messages here on the forum, will get us anywhere. Could you please contact me directly? We could meet via Zoom and discuss possibilities more directly.
I am managing the Consortium of North American Lichen Herbaria, a biodiversity data platform build on the software Symbiota. That software was developed here at Arizona State University. The software is now widely being used to share biodiversity data from portals dedicated to specific regions and organisms. We should discuss how it might be possible to export taxonomic data sets from these databases and load those as controlled vocabularies into PSu.
Hert has been very supportive of the DarwinCore and I am sure if we could find a viable solution to generate controlled vocabularies from these platforms, he might find a way how PSu would support such an import option.
Cheers,
Frank

Post Reply