I am setting up our workflow with Collecting Event (CE) in shared mode, and I have a question about how to define/create “unique localities” without a CollectorNumber. In our case, the difference between samples is the depth, which is recorded in the LocalityDetail table (StartDepth/EndDepth).
For example (same place and coordinates, but different depth and/or date):
(A) Date: 01/01/2026 — Port de sa Dragonera Lat: 39.5868; Long: 2.3279; CoordUncert (m): 151.26; StartDepth: 1; EndDepth: 2
(B) Date: 01/01/2026 — Port de sa Dragonera Lat: 39.5868; Long: 2.3279; CoordUncert (m): 151.26; StartDepth: 8; EndDepth: 20
(C) Date: 21/12/2020 — Port de sa Dragonera Lat: 39.5868; Long: 2.3279; CoordUncert (m): 151.26; StartDepth: 1; EndDepth: 2
In practice, we think A, B, and C should be different (“unique”) records:
(A) and (C) have the same depth but different dates.
(A) and (B) have the same date but different depths.
My questions:
If we work with shared CEs, what is the best practice to handle different depths and dates?
Should StartDepth/EndDepth be part of the Locality table or is it better to put the depth field in the Collecting Event so we can reuse one single Locality for the coordinates?
Can we have two Localities with the same LocalityName but different lat/long?
There are many ways to solve this issue, and it will depend on how much you want to do manually / how much info you want to share between Collection Objects from the same within-event habitats.
We use shared CEs in our database, and handle this type of issue by recording the differing habitat/depth info in Collection Object - Remarks (though I plan to implement a Collection Object - Habitat field soon to take some of this info). We record the whole range of depths / habitats in the shared CE.
If it is important to preserve the specific depth an individual specimen / lot came from, and you don’t want to deal with multiple lots from the same depth/habitat needing info added to each individual CO record (i.e. not a shared record in a linked table) then you should record the different depths as different CEs with the same Locality.
There should generally be one date or date range associated with a Collecting Event record. Repeated samples of the same location should be recorded as new CEs of the same Locality. For example, your A and C would never be the same CE in our database, and I think most databases around the world would not record a sampling effort that happened 6 years apart as the same event. They are different Collecting Events that sampled the same Locality.
I think it makes the most sense to have Depth fields be part of the CE table, because it prevents having to georeference multiple instances of the same Locality at different depths. In your case, I would think about how much effort it would take vs how much it would save to record sampling of the same place/time at different depths as different CE records vs as a shared CE record with the specifics dwc:habitat or dwc:depth fields stored in the CO records. If the number of lots from an event is generally small in your collection, it might make more sense to go the CO route for these fields. If it’s large, it probably makes sense to make each different depth its own CE.
Nothing prevents this, but you may not want to allow it because it seems confusing. Someone might come along later and think this is an error. Why would you want to have two locations with the same information associated with two different lat/longs? If the issue is that the original collectors didn’t record specific info and didn’t use a site/collector number for a given collecting effort/location, and only provided a lat/long, what we do is actually record more specific information based on the GPS point [in brackets] as part of the Locality. In our DB, the Locality is not necessarily verbatim what the collectors wrote, we alter it if it is found to be unclear or incorrect, as the original information is recorded in the record history / our scanned ledger books if people care about the verbatim information.
Thank you very much for your helpful replies, recommendations, and clear explanations, they have been greatly appreciated.
At the moment, we are using embedded Collecting Events in other collections, but for the Botany discipline we are planning to work with shared Collecting Events. For this reason, I want to be sure that I fully understand how shared CEs function and which fields make each CE unique, in order to document our data correctly.
In our case, the same toponym (LocalityName) may correspond to samples collected at different collecting points with different latitude and longitude values. My understanding is that, in such situations, the LocalityName should include both the toponym and a reference to the specific point. Therefore, to avoid confusion, it is recommended that the LocalityName be unique.
We also document different georeferencing methods, for example based on original information provided by the collectors, retrospective georeferencing, or reverse geocoding when only a GPS point is available and we assign a toponym as the LocalityName.
Finally, I understand from your explanation that it is preferable for depth information to be recorded at the Collecting Event level rather than in the Locality.
I think you understand the issues at hand based on this comment . The only other thing I would stress based on this comment is that if you add locality info yourself based on the point (i.e. by reverse geocoding) that you use database conventions or remarks field to document the inferences you make, like the example I mentioned above where I say information is in [brackets] when it’s not original/verbatim.
We do this in the localityName in our database but we do not do it in the geography table.
@Anna One more thing I thought of that might be useful for you – if you want to link CollectingEvents that sample different depths at the same location together while still making separate CE records, one way to do that is to use the CollectingTrip table.
