Thank you so much Wesley! I appreciate your perspectives!
I agree with your interpretation of eco:isAbsenceReported and eco:isTaxonomicScopeFullyReported. That means the definition of the terms are clear (if the values are true or false), yay!
However, I am still looking for the answer of my first question - If nothing was caught in a trawl Event. Should eco:isTaxonomicScopeFullyReported (defined as "Every dwc:Organism that was included within the taxonomic scope, and was detected during the dwc:Event, was reported.”) be true, false or null? Is “nothing” considers as “within" eco:targetTaxonomicScope? If “nothing" is not considered as part of eco:targetTaxonomicScope, then eco:isTaxonomicScopeFullyReported will receive a false. If that’s the case, the user cannot make inference of non-detection of taxa within the eco:targetTaxonomicScope.
The reason I included eco:isAbsenceReported in the table I made was because this term is not restricted to the scopes. If there is only one Occurrence associated with the Event that says Animalia, occurrenceStatus=absent, hence eco:isAbsenceReported=true, it MAY be interpreted as none of my taxa in eco:targetTaxonomicScope (if they all belong to Animalia) and by-catch that is animal was caught.
Thanks again!!
Cheers Ming
On 3 Aug 2023, at 19:49, Wesley M. Hochachka wmh6@cornell.edu wrote:
Hi Ming,
If I am correctly understanding the use of eco:isAbsenceReported, then the value of eco:isAbsenceReported should be irrelevant to whether eco:isTaxonomicScopeFullyReported is set to be "true" or "false". I am assuming thateco:isAbsenceReported and eco:isTaxonomicScopeFullyReported are independent of each other, such that all four combinations of conditions can occur: eco:isAbsenceReported = false and eco:isTaxonomicScopeFullyReported = false: in this case the absence of the report of a taxon within the taxonomic scope cannot be interpreted. eco:isAbsenceReported = true and eco:isTaxonomicScopeFullyReported = false: in this case the absence of the report of a taxon within the taxonomic scope cannot be interpreted, because the combination of these two conditions says that "some counts of zero are being reported, but not necessarily for all of the species within or outside of the taxonomic scope. eco:isAbsenceReported = false and eco:isTaxonomicScopeFullyReported = true: in this case the absence of the report of a taxon within the taxonomic scope can be interpreted to mean that the species was not detected, but these non-detections will need to be inferred because they are not explicitly entered as records of counts of zero. eco:isAbsenceReported = true and eco:isTaxonomicScopeFullyReported = true: in this case, the absence of the report of a taxon within the taxonomic scope can be interpreted to mean that the species was not detected, and there is no need to infer these absences because counts of zero will be present for all taxa that were not detected. Again, what I have written above is based on the assumption that I fully understand the use of eco:isAbsenceReported. If this assumption is wrong, then I think that the definition should be tweaked to make its interpretation unambiguous, because my interpretation is consistent with the definition, as far as I can tell.
I do not think that there is any way to differentiate between a sampling event in which only by-catch was detected but it was not reported, and a sampling even detected literally no organisms of any taxon (within the constraints of the sampling methods). From an ecological perspective, the distinction could be important, because the two extreme alternatives are the difference between "our target taxa were not detected but there were still organisms (i.e. the by-catch) detected", and "the sampling area was devoid of detected life". So, these extreme alternatives could only be differentiated of at least one by-catch taxon was reported for the sampling event. However, this feels somewhat dis-satisfying to me, because it's still not possible to tell the difference between "none of the target taxa were detected and just a single organism of any taxon was detected", and "lots of different non-target taxa were detected, but we only bothered to report one of them". Reports of by-catch (which I'd generically call "presence-only data") is just massively limited in its interpretability.
So, anyway, those are my answers to your questions. You might want to forward my reply on to the Humboldt Extension listserv, because it's almost certain that my attempt to send my reply to this listserv is going to fail.
Wesley
Wesley Hochachka Senior Research Associate Cornell Lab of Ornithology ph. (607) 254-2484
From: Yi Ming Gan ymgan@naturalsciences.be Sent: Thursday, August 3, 2023 06:10 To: tdwg-humboldt@lists.tdwg.org tdwg-humboldt@lists.tdwg.org; Rob Stevenson rdstevenson10@gmail.com Cc: Wesley M. Hochachka wmh6@cornell.edu Subject: How to populate/interpret eco:isTaxonomicScopeFullyReported if dwc:Event caught nothing/something outside of eco:targetTaxonomicScope?
Hi all,
In my attempt to use Humboldt Extension to infer non-detection, I arrived at this question because it is not clear to me how to use null in the following situation.
Should eco:isTaxonomicScopeFullyReported (previously eco:isTaxonomicScopeComplete) be "true", "false" or null if an dwc:Event do not catch anything within the eco:targetTaxonomicScope and eco:isAbsenceReported == "false"?
I illustrate my understanding in the table below: assume that all the catch are reported in the Occurrence extension eventID catch eventRemarks targetTaxonomicScope (taxa are pipe-separated) isAbsenceReported isTaxonomicScopeFullyReported can infer non-detection of taxa within targetTaxonomicScope? event_001 A A | B false true true event_002 A A | B false false false event_003 A, C A | B false true true event_004 by-catch only A | B false true true event_005 by-catch only A | B false false false event_006 by-catch only A | B false false event_007 caught nothing A | B false true true event_008 caught nothing A | B false false false event_009 caught nothing A | B false false
How can I distinguish: an dwc:Event caught by-catch only, but did not report it (no dwc:Occurrence record associates with the dwc:Event) an dwc:Event caught nothing at all Is this distinction important? Can someone with ecology or relevant background please comment? @Rob Stevenson mailto:rdstevenson10@gmail.commaybe? Thanks a lot!
Cheers Ming