Big-Bee-Network / trait-intertegular-distance

trait repository
Creative Commons Zero v1.0 Universal
1 stars 0 forks source link

translate one (or more) intertegular distance metrics into wikidata as a quantitative trait #1

Open jhpoelen opened 1 year ago

jhpoelen commented 1 year ago

Hi @Daniel-Mietchen ,

@seltmann et al. have prepared some example of intertegular distance measurements. This distance (as described in https://github.com/Big-Bee-Network/trait-intertegular-distance) is a proxy for body size/mass for bees.

I was wondering whether you have any ideas on how to represent this metric in the wiki-verse. . .

Looking forward to our session tomorrow!

jhpoelen commented 1 year ago

For examples see:

https://github.com/Big-Bee-Network/trait-intertegular-distance/blob/main/UCSB-Microscope.csv

and

https://github.com/Big-Bee-Network/trait-intertegular-distance/blob/main/UCSB-NfN-2022-07-22.csv

jhpoelen commented 1 year ago

Oops, mistyped your handle @Daniel-Mietchen earlier. Apologies.

jhpoelen commented 1 year ago

@Daniel-Mietchen pointed out that it may be more practical to use a more generic trait like body size or body mass. Then, as a reference, point to the data publication. This data publication would be expected to include a description on how to the body size/mass estimate came about.

Daniel-Mietchen commented 1 year ago

By "practical", I mean doable with existing Wikidata properties or properties likely to get community approval.

jhpoelen commented 1 year ago

Here are some example of traits as listed in the biological traits project -

https://www.wikidata.org/wiki/Wikidata:WikiProject_Biodiversity/Traits#Morphological_traits

jhpoelen commented 1 year ago

@Daniel-Mietchen suggesting to use some of the individuals of https://www.wikidata.org/wiki/Q179959 as a model for type specimen traits , a nearly extinct bird, with individuals documented on wiki data.

jhpoelen commented 1 year ago

We need to select an appropriate trait related to body size/mass available in wikidata .

See related discussions on various wikidata discussions -

https://www.wikidata.org/wiki/Wikidata:Project_chat/Archive/2023/02#best_practice_for_the_representation_of_quantitative_data_and_their_metadata

fyi @katjaschulz @jhammock - any thoughts on defining a body size/mass trait property in wikidata?

jhpoelen commented 1 year ago

@Daniel-Mietchen should we create a proposal to introduce body size as a wikidata property?

jhpoelen commented 1 year ago

@Daniel-Mietchen can you please share some examples of type specimen recorded in Wikidata?

KatjaSchulz commented 1 year ago

I agree with @Daniel-Mietchen that intertegular distance has little prospect of getting approved as a WikiData property. However, if you can develop a formula that converts intertegular distance measurements into estimates of body length (P2043) or mass (P2067), you could add statements for those properties to Wikidata with references to the publication where you present the intertegular distance value and to the paper where you explain the data conversion. You could also add a nature of statement (P5102) qualifier using something like estimate (Q37113960) or extrapolation (Q744069) as the item. See here for available items for this qualifier: https://www.wikidata.org/wiki/Property:P5102#P2302

Daniel-Mietchen commented 1 year ago

@Daniel-Mietchen can you please share some examples of type specimen recorded in Wikidata?

@jhpoelen @KatjaSchulz @seltmann @jhammock

Here is a list of several hundred type specimens. The links between them and the respective species are not modeled in a consistent fashion, so the query does not show the taxon names.

There are also several thousand individuals from various taxathat have their own Wikidata entries.

Daniel-Mietchen commented 1 year ago

@Daniel-Mietchen should we create a proposal to introduce body size as a wikidata property?

Yes, given the popularity of body size-related statements in the literature, this is worth exploring. As always, the devil here is in the details, e.g. in properly representing the link between the actual measurements and the inferred body size.

seltmann commented 1 year ago

Hi @jhpoelen @KatjaSchulz @Daniel-Mietchen the primary paper for conversion of intertegular span (HAO:0002584) is Cane, 1987 https://www.jstor.org/stable/25084877.

Span is related to dry weight (body mass) by the equation y = 0.77(x)^0.405, where y is intertegular span and x is dry weight.

jhpoelen commented 1 year ago

@seltmann Thanks for the reference and I am excited by the prospect of having Big Bee contribute to wikidata and other trait datasets.

@KatjaSchulz @jhammock would you like Big Bee to deliver some bee body mass measurements to EOL directly, or would you like to calculate them from intertegular span yourself. Are you still interested in trait datasets for EOL?

jhammock commented 1 year ago

This is what we've done in EOL in similar cases in other groups. Some are sourced from a single paper (which provided linear measurement and conversion formula), eg: https://eol.org/records/R556-PK219451427

and some use separate sources for measurement and conversion method, eg: https://eol.org/records/R556-PK219453343

Our own data model is pretty crude, but for export to wikidata we might use "inferred from" the source of the measurement. I'm not sure what property is appropriate to associate the source of the conversion method, though...

jhpoelen commented 1 year ago

@jhammock thanks for sharing these example of converted datasets . I am hoping to motivate myself to come up with some way to follow your example for the intertegular distance-based dry mass estimates in collaboration with @seltmann with the goal to distribute as widely as practically possible. This would include the provenance/origin of the data and their associated authorities (e.g., people and their publications).