[IPT] how to do when you have more than one sampling method for the same plant plot

Tue Jul 21 13:28:40 UTC 2020

Hi Rui,

If using the ExtendedMeasurementsOrFacts, as Andrea recommends, I should include the occurrence ID from table occurrences in the column `occurrenceID`.

Is this correct?

Ø  In this case, I would say yes. In general, that would depend on whether the measurement or fact relates to the event, or to the occurrence. The specification says about occurrenceID: "The identifier of the occurrence the measurement or fact refers to. If not applicable, it should be left empty.". This refers back to the introductory paragraph (see https://tools.gbif.org/dwca-validator/extension.do?id=http://rs.iobis.org/obis/terms/ExtendedMeasurementOrFact), stating that " When used with Event Core it allows to create an additional link between the eMoF and the occurrence extension." In the case of measures of coverage, it would indeed be occurrence related.

However, please be aware that this is a contruct that no longer falls withing the "start schema" setup that an IPT supports (and GBIF indexes from): links can only go one level deep from the center, and links between satelite tables (in this case occurrence and eMoF) are not supported. That said, it is a good format for a dataset to exist in though, regardless of GBIF, at indexing, losing some of that information.

Alternatively, you could consider using the combination of https://dwc.tdwg.org/terms/#dwc:organismQuantity and https://dwc.tdwg.org/terms/#dwc:organismQuantityType on occurrence records directly for the measure that this fits best, and only use the extension for the other one.

There are also the "dynamicProperties" (https://dwc.tdwg.org/terms/#dwc:dynamicProperties), though these are now so generic that I am not sure how useful they would be for your case.

From: IPT <ipt-bounces at lists.gbif.org> On Behalf Of Rui Figueira
Sent: Tuesday, 21 July, 2020 12:12
To: André Heughebaert <a.heughebaert at biodiversity.be>
Cc: ipt at lists.gbif.org
Subject: Re: [IPT] how to do when you have more than one sampling method for the same plant plot

Hi all,

Thanks, Andrea and André :-) for your replies.

Just to be clear, this is how I should do:

In the event core, I leave `samplingProtocol` blank, because each event will have two sampling methods. The sampling method should be included in `measurementType`of the (Extended)MeasurementsOrFacts DwC Extension.

I am curious what will happen when GBIF ingests the data. Will any of the sampling methods be associated with the occurrences?

Cheers,

Rui

------------------

Rui Figueira

Coordenador do Nó Português do GBIF

ruifigueira at isa.ulisboa.pt<mailto:ruifigueira at isa.ulisboa.pt>

Instituto Superior de Agronomia

Herbário

Tapada da Ajuda, 1349-017 Lisboa, Portugal

Tel. +351 213653165 | Fax. +351 213653195

http://www.gbif.pt

http://www.isa.ulisboa.pt
On 7/21/20 10:47 AM, André Heughebaert wrote:
Hi Rui,
What about using SampleEvent core with MeasurementOrFacts<https://tools.gbif.org/dwca-validator/extension.do?id=dwc:MeasurementOrFact> DwC extension to store both event measurements(BraunBlanquet Scale value and % of coverage.) in addition to the Occurrence DwC extension for the occurrences.
No duplicates and you keep all original data.
My2cents

Ir Andre Heughebaert
GBIF Node Manager at Belgian Biodiversity Platform<http://www.biodiversity.be>
+32(0)2238 3796
WTC III, 30 Bld Simon Bolivarlaan; B-1000 Brussels, Belgium
ORCID 0000-0002-7839-5300<http://orcid.org/0000-0002-7839-5300>

On Mon, 20 Jul 2020 at 18:34, Rui Figueira <ruifigueira at isa.ulisboa.pt<mailto:ruifigueira at isa.ulisboa.pt>> wrote:
Dear all,

One of our publishers is preparing an event based dataset with plant
surveys, where two measurements were taken at each location: a Braun
Blanquet Scale value and % of coverage.

The normal option to create this dataset would be to make two event
records for each plot, one for each survey method. This allows to
preserve all original data, but will result in two occurrences for the
same species at the same plot.

The other option would be to report only one of the survey methods,
avoiding occurrence duplication, but not preserving data for the other
method.

Would anyone have a suggestion how to prepare the dataset, so that it
keeps values for both methods and does not create duplication of
occurrences?

Or the duplication is not an issue, and we should publish data for both
methods?

Thanks,

Rui

--
------------------

Rui Figueira
Coordenador do Nó Português do GBIF
ruifigueira at isa.ulisboa.pt<mailto:ruifigueira at isa.ulisboa.pt>
Instituto Superior de Agronomia
Herbário
Tapada da Ajuda, 1349-017 Lisboa, Portugal
Tel. +351 213653165 | Fax. +351 213653195
http://www.gbif.pt
http://www.isa.ulisboa.pt

_______________________________________________
IPT mailing list
IPT at lists.gbif.org<mailto:IPT at lists.gbif.org>
https://lists.gbif.org/mailman/listinfo/ipt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.gbif.org/pipermail/ipt/attachments/20200721/7321f9f3/attachment-0001.html>