dwc
dwc copied to clipboard
tdwg
New term - organismStatus
This proposal is under active development in the 'OSR - How Did It Die?' Task Group.
- Submitter: Sophia Ratcliffe - NBN Trust (https://nbnatlas.org/)
- Proponents (at least two independent parties who need this term): I have had requests from several of our data providers asking for the ability to supply and filter records by alive/dead status
- Justification (why is this term necessary?): We receive many records where the occurrence was of a dead animal and at the moment there is no way store the status of the organism in a DwC term with a controlled vocabulary. It is really important in marine records, which often contain stranding information, and for road kill records.
Proposed definition of the new term:
- Term name (in lowerCamelCase): organismStatus
- Class (e.g. Location, Taxon): organism
- Definition of the term: A description of the status of the organism (alive or dead)
- Comment (examples, recommendations regarding content, etc.): Recommended best practice is to use a controlled vocabulary, with the terms alive, dead, unknown
- Refines (identifier of the broader term this term refines, if applicable): Currently we've been using occurrenceRemarks, but it is not easily searchable.
- Replaces (identifier of the existing term that would be deprecated and replaced by this term, if applicable): N/A
- ABCD 2.06 (XPATH of the equivalent term in ABCD, if applicable): N/A
I support the concept, but not so much the suggested name. Status can mean so many things. Can't think of a better one yet.
Also, there is the issue of dead museum specimens and living culture collections. I'm not certain these matter if they are described as dead or alive, as long as it is appropriate, but it is worth considering. I think basisOfRecord is currently used for some of these distinctions.
Yes, I didn't like using status, but it felt consistent with other DwC terms.
The new term would be used for field observations where the basisOfRecord is HumanObservation (or Occurrence, as some data providers use).
A possible name could use vitality.
Or viability, which is already used for seed collections (values might be viable, dormant, inviable, dead)
Would reproducing or fertile be appropriate values for this term in that case? It's important for determining if a non-native species is established. (Excuse and ignore if another term already takes this into account.)
Annie Simpson
From: Quentin Groom [email protected] Sent: Wednesday, May 29, 2019 10:43 AM To: tdwg/dwc Cc: Subscribed Subject: Re: [tdwg/dwc] organismStatus - proposal for new term (#228)
I like vitality.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/tdwg/dwc/issues/228?email_source=notifications&email_token=AELXSJLSBECTAETW7V32U6TPX2I7NA5CNFSM4HQLPV62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWPR3PQ#issuecomment-496967102, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AELXSJKNXNV3GUF5XZIMXNTPX2I7NANCNFSM4HQLPV6Q.
Would reproducing or fertile be appropriate values for this term in that case?
dwc:reproductiveCondition and dwc:lifeStage cover this
Could the definition be changed to
An indication of whether the organism was alive or dead at the time of collection or observation.
Comment: Generally Intended to be used with a dwc:basisOfRecord of PreservedSpecimen, MaterialSample, or HumanObservation.
A further thought
If this field is added to Darwin Core then next obvious need would be causeOfDeath.
Rather than just having a flag to say if something is dead, why not jump straight to a causeOfDeath field with a vocabulary?
causeOfDeath would be really useful, then we can differentiate strandings, road kills etc.
A single term for vitality, with @qgroom's suggestion for the definition and comment, would still be useful, so that we can easily select all 'dead' records, but if needs be, we can derive that from causeOfDeath.
I don't like boolean flag fields in the standard when they can be derived from other fields. This doesn't mean they can't be implemented in a database to make searching faster, but as long as they are always derived from one source field there is no danger of a mismatch, where for example one vitality says living, but the cause of death is roadkill.
It is a good question for GBIF though. @MattBlissett was saying dwc:year is derived from dwc:eventdate if it is empty. How much is GBIF willing to add non-standard fields for indexing if they are derived from standard fields?
I would like to urge some caution here. In other words, please be very explicit about your needs or use-case(s) for a controlled vocabulary describing whether an organism is/was dead or alive. In many 'collecting events", whole organisms are encountered alive, but the process transforms them to dead. Organisms can be observed dead or alive in the field. An organism could have been collected alive (a long time ago), lived many years in captivity, and be dead now. "causeOfDeath" could venture easily into necropsy results. In the case of cetacean and pinniped "strandings" (i.e., organisms encountered dead on the shoreline), the cause of death is of great interest and represents a summary conclusion based evidence found in necropsy, pathology, and/or toxicology.
On Thu, May 30, 2019 at 9:50 AM Quentin Groom [email protected] wrote:
I don't like boolean flag fields in the standard when they can be derived from other fields. This doesn't mean they can't be implemented in a database to make searching faster, but as long as they are always derived from one source field there is no danger of a mismatch, where for example one vitality says living, but the cause of death is roadkill.
It is a good question for GBIF though. @MattBlissett https://github.com/MattBlissett was saying dwc:year is derived from dwc:eventdate if it is empty. How much is GBIF willing to add non-standard fields for indexing if they are derived from standard fields?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/tdwg/dwc/issues/228?email_source=notifications&email_token=ACKZUDLS645A42HIJ2EWT43PYAAU7A5CNFSM4HQLPV62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWS3RVY#issuecomment-497400023, or mute the thread https://github.com/notifications/unsubscribe-auth/ACKZUDNM76ZYRUNX5NLL5E3PYAAU7ANCNFSM4HQLPV6Q .
Potential user here: we work on wild ungulates, and would be interested in vitality and causeOfDeath terms.
Some of our data is road kill data (individual observed dead, with an obvious cause), hunting data (dead animals, through various types of hunting methods), monitoring data (live animals / animals found dead).
The executive of TDWG thinks this is an important issue to consider. The first step would be to engage in discussion with the Darwin Core Maintenance Group to figure out the next steps.
@tucotuco @peterdesmet @mdoering @timrobertson100 @baskaufs @pzermoglio @morr @chicoreus
Consideration by the Darwin Core Maintenance Group should include whether development of this term and a controlled vocabulary of values for it (including the development of user stories and competency questions) would fall within the scope of the maintenance group, or whether this work is of sufficient scope to merit a new task group.
Consideration also needs to be paid to alignment with ABCD, including the PaleontologicalUnit/Preservation/Taphonomy and PaleontologicalUnit/Preservation/Completeness terms.
It seems to me that this term should be organized under the Occurrence class rather than Organism. In the thread, we've been talking about the status of the organism at the time when it was observed or collected. That makes it specific to a particular Occurrence of the organism, since an organism can be observed repeatedly and on some of those Occurrences it might have been alive, then dead on its last occurrence (perhaps being made to be dead by the collection).
I very much like the term vitality and we should probably think about a short accompanying vocabulary that is more expressive than just the binary dead or alive.
Would you also use it for exuviae or shells?
I found a slide from 2014 about establishmentMeans and its actual content in GBIF which contained a few indications to dead organisms. It seems causeOfDeath is a good company for vitality:
Notes from the Biodiversity Next Unconference 25 October 2019
Vitality & CauseOfDeath
Introduction Linked to the...
- decay process
- health
- taphonomy
- species interactions (disease, pests, predators)
Relates to dwc:basisOfRecord
- PreservedSpecimen
- FossilSpecimen
- HumanObservation
- LivingSpecimen
Use Cases
- Disease Ecology
- Causes of death in a living collection (animals and plants)
- Historical studies
- The last specimen of an extinct taxon
- Understanding the genomics of dead organisms (e.g. non-target organisms)
- Communication with the public about the origins of collections
- identifying roadkill hotspots
- Documenting whale strandings
- Studying the phenology of death
- Quantifying by-catch in invasive species control programs
- Reducing unnecessary mortality of endangered species.
Scope
- When did it die? (also see dwc:GeologicalContext)
- Collection methods (dwc:samplingProtocol)
- Should close to death be included?
- Uncertainty
- "Natural" versus "unnatural" death
- Multiple causes of death e.g. killed on road as a cause of blindness caused by disease.
Consider the public sensitivity towards death and killing animals.
Team 1:
- Collected Dead
- "Natural Causes"
- Predation
- Storms/weather
- Disease
- Age/Senescence
(Fossils: Location) +Confidence level of above
- Killed when collecting
- Pitfalls
- Biobanks: alive?
- Trapped
- Shot
- Picked
- Harvested
- Cause of death ambiguous Further data needed Unsure at time of collecting
Team 2: #causeOfDeath
- trapped
- poisoned
- starved
- drowned
- shot
- of old age
- road kill
- disease
- weeding
- collecting
- burned
Direct cause of death linked to event in time, such as a volcano eruption Are there pathology ontologies?
Team 3: SALAFSKY, N. , SALZER, D. , STATTERSFIELD, A. J., HILTON‐TAYLOR, C. , NEUGARTEN, R. , BUTCHART, S. H., COLLEN, B. , COX, N. , MASTER, L. L., O'CONNOR, S. and WILKIE, D. (2008), A Standard Lexicon for Biodiversity Conservation: Unified Classifications of Threats and Actions. Conservation Biology, 22: 897-911. doi:10.1111/j.1523-1739.2008.00937.x
-
Experimental [death]
-
Anthropogenic [death]
-
Biotic/Abiotic [death]
-
Environmental [death]
-
Serves as marker/flag for follow-up in downstream uses
-
medical concepts/terms might already exist
-
10-20 Use cases to test against
Age at death? Would it have died anyway? Whose fault is it? Collection event
Team 4: Cause and qualifier When?
Codes used to express plant losses at Meise Botanic Garden
DIE_WHY_CODES DIE_WHY_CODES DESCRIPTION D D discarded ES ES environmental stress (cold, heat, drought) FR FR failed to root H H horticultural failure N N neglect / error ND ND natural disaster (hurricane, flood, etc.) P P pest R R re-accessioned under a new name RF RF rootstock failure SO SO sold V V made into a voucher specimen NV NV no viable seed X X not coded ANS ANS annual - seeds saved but original plants discarded ANU ANU annual - died after flowering, no viable seed produced A A animal damage DI DI disease FG FG failed to germinate G G given away NC NC natural cause (old age, etc.) O O other S S stolen U U unable to locate W W weeds / competition VA VA vandalism SH SH shaded out / overgrown WT WT changes to water table
Is it still possible to contribute to this conversation related to OccurrenceStatus. Currently there are 2 terms: present or absent. if an animal/plant is dead it is still present. I would like to propose the addition of a new term called 'trace'. This term could be assigned to observations of shells - these animals should not be flagged as present/dead if the shell is empty. a crab shell can be considered present/dead but an empty snail shell should be flagged as 'trace'. Other cases where 'trace' could be assigned are when one observes scat, tracks, nests, etc. In these cases there are 'traces' that an animal was present at this location at some time in the past. the animal is not present at the time that the observation was made. It does not imply that the animal is dead.
Many different things are added to the Darwin Core field establishmentMeans and published on GBIF. This file contains a list of the terms relating to death that have been put in establishmentMeans. DeadEntriesPutInEstablishmentMeans.xlsx
Causes of Death from the UK Predatory Bird Monitoring Scheme (PBMS) From Lee A. Walker of the UK's Centre of Ecology and Hydrology.
| BOP_USER_BOP_COD_DESC | ||
|---|---|---|
| COD_ID | COD_DESC | COD_ABRIDGED |
| 2 | Window collision | Accident |
| 3 | Shot | Shot and Special Collection |
| 4 | Collision | Accident |
| 5 | Starvation through injury | Accident |
| 6 | Unknown Trauma | Unknown |
| 7 | Disease | Starvation and Disease |
| 8 | Other trauma | Accident |
| 9 | Starvation | Starvation and Disease |
| 10 | Poison | Poison |
| 11 | Road accident | Accident |
| 12 | Human predation | Accident |
| 13 | Animal predation | Accident |
| 14 | Drowning | Accident |
| 15 | Experimental collection/Euthenasia | Shot and Special Collection |
| 16 | Electrocution | Accident |
| 1 | Unknown | Unknown |
Email from Lee states
Although the CoD list is in our standard operating procedures these are not citable. However the list is published in the supporting documentation for a dataset that we have published in the UK EIDC data catalogue.
The record is here: https://catalogue.ceh.ac.uk/documents/e39e3afb-eb90-4a7d-921d-e03b1a0f414d
The specific file can be accessed through the link below which downloads a zip file. The CoD list is in the column_headings_predatory_bird_livers.csv file.
https://data-package.ceh.ac.uk/sd/e39e3afb-eb90-4a7d-921d-e03b1a0f414d.zip
Curious if this has advanced since 2019? I have a dataset of dead marine mammals and seabirds that wash up on shore and I haven't heard a new term has been added.
It hasn't progressed further than agreeing to set up a task group to explain the rationale of adding new terms for vitality and cause of death and their vocabularies. I've started the charter and meant to send it round to very one for comment. Let me know if you're interested in being involved. Thanks.
@sophiathirza I'm still keen to make progress on this. Let me know if there is anything I can do. @albenson-usgs It would be useful to know what terms you want to use for your standings dataset.
Yes please include me! Thanks Mary Kennedy
Sent from Mail for Windows 10
From: Sophia Sent: April 15, 2020 6:42 PM To: tdwg/dwc Cc: OBISCanada; Comment Subject: Re: [tdwg/dwc] organismStatus - proposal for new term (#228)
It hasn't progressed further than agreeing to set up a task group to explain the rationale of adding new terms for vitality and cause of death and their vocabularies. I've started the charter and meant to send it round to very one for comment. Let me know if you're interested in being involved. Thanks. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
@qgroom I believe if I'm reading this thread correctly that I would want to use vitality. I'm not sure about causeOfDeath- I haven't seen the dataset yet to know if that's in there.
When will a resolution be made on this matter? It appears that a great deal of valuable energy and thought has gone into it, so are we at a stage where a decision can be made? I have a colleague who is eager to submit data into OBIS but needs to have this clarified to ensure his work is interpreted correctly.
Also curious if there are plans to address the question about 'evidence' when one doesn't know if dead or alive? in the case of shells maybe one can assume dead but maybe the organism did not die at that location. evidence might be of more interest to terrestrial critters - tracks, faeces, nests, molts, etc.