What things are misidentified as large milkweed bug?

One of the common name of Caesalpinia pulcherrima is also “red bird of paradise”, which contributes also to the confusion with Erythrostemon gilliesii, competing with Strelitzia. But this may not be the exclusive explation, (1) Caesalpinia pulcherrima and Erythrostemon gilliesii are in the same tribe and (2) there exists a yellow subspecies of Caesalpinia pulcherrima making it visually closer to Erythrostemon gilliesii.

You could check this one also against Cassia fistula:

Goldenrain tree
Koelreuteria paniculata

Golden shower tree ; Golden rain tree
Cassia fistula

1 Like

you could run the script using the code and instructions above, or just look at the Similar Species tab of the Taxon page. it presents the same information, just in different format as alexis18’s figures.

Yes I can! And thank you for the tip!

Other species commonly misidentified as Koelreuteria paniculata:

Besides other Koelreuteria species, Cassia fistula (same common name) is the most often misidentification.

Based on your explanations, this label in the “Similar Species” panel:

(1) Other species commonly misidentified as this species

should be instead:

(2) Commonly misidentified as these other species

For instance, if we are in the Koelreuteria paniculata taxon page and we see Cassia fistula in the “Similar Species” panel, it means that some actual Koelreuteria paniculata have been misidentified as Cassia fistula, so statement (2) applies, not statement (1).

2 Likes

I think - based on the description - that the similar species are:

Things originally identified as Koelreuteria paniculata that have been corrected to the listed identifications. That’s the way I integrated our misidentifications

Though it would be very interesting to see the alternative

it looks like it actually goes both ways, though maybe the similar species function / count is somewhat ephemeral since seems to rely on both taxa still being active identifications on any observation that is counted. in the Similar Species tab, you can click on the number in the corner to open up a new window that shows you what gets included in the count.

here’s an example for Pisum:

3 Likes

a tangent, but your use of a Sankey diagram here inspired me to add one to something i created back in the day: https://jumear.github.io/stirfry/iNat_obs_counts_by_iconic_taxa.html. (i think it’s interesting to see how the diagram changes when run by different users.)

2 Likes

Very very cool

1 Like

Interesting that such a large proportion of unknowns go to Casual and not to Needs ID.

1 Like

yes. i think that’s just the result of some identifiers focusing specifically on getting unknown to at least a high level taxon like plants (see https://forum.inaturalist.org/t/amount-of-unknown-records-is-decreasing/8594).

what didn’t make sense to me at first was the small but significant number of bird observations going to casual. my first thought was that maybe there are zoos or aviaries that people visit a lot, or maybe some folks are purposely looking for captive or poached birds, but i later realized it’s just a case where veteran birders are noting birds that they observed on site, without loading photos or sounds as evidence. so for example, you can tell that kueda is one of these birders: https://jumear.github.io/stirfry/iNat_obs_counts_by_iconic_taxa.html?user_id=kueda.

Does RTools also have to be installed? I got some messages when I tried to install the packages that it was required, and was successful installing them after I downloaded it, but I’m not sure if that was because it was needed, or just because I relaunched RStudio.

from https://cran.rstudio.com/bin/windows/Rtools/:

Note that rtools40 is only needed build R packages with C/C++/Fortran code from source. By default, R for Windows installs the precompiled “binary packages” from CRAN, for which you do not need rtools!

(i just ignored the rtools messages when i did my installation.)

1 Like

Somewhat surprisingly, the most common mis-ID is a two way street:

given that L. turcicus and L. reclivatus look much more like L. kalmii. Perhaps due to CV vs. trained human ID?

I think it’s more about the volume of Lygaeus kalmii and O. fasciatus observations, even though turcicus and reclivatus look more similar, there are just so many more O. fasciatus observations than turcicus and reclivatus.

But I know you and I and @fabienpiednoir have been going through to try and find some of the hidden ones

1 Like

Love these graphs…thanks so much for posting code @alexis18!

Quick question: is there any way to show observations that were initially made to the species level but RGed to genus (and not to any species)?

Example: I’m interested in ids of Anolis lizards, and do lots of iding. A non-negligible chunk of observations are initially posted with a species ID, but are not definitely verifiable as a specific species. They are, however, often able to be ided at RG to the genus level (Anolis). It would be cool to have the capability to include these in the viz if there’s an easy way to specify in the code. I did try digging into the objects (nodes, links, genera) a little bit and can see that the genus alone is a node, so it’s maybe possible to have this but maybe not an option in the sankeyNetwork function (and rest of code is a little over my head!). Thanks for any thoughts anyone has!

it looks like the similar species endpoint only does comparisons at the same rank. so species would only compare to other species and genera to other genera. so you wouldn’t be able to use that particular endpoint to do genus to species.

however, this will identify cases where there is a species-level id on a genus-level research grade observation of Anolis: http://api.inaturalist.org/v1/identifications?current_taxon=false&rank=species&observation_rank=genus&current=true&quality_grade=research&taxon_id=36362&observation_taxon_id=36362&order=desc&order_by=created_at. so you can visualize or otherwise use those results as you like.

UPDATE: i made something (see https://forum.inaturalist.org/t/search-and-filter-identifications/1304/23) that might presents the results of the API identifications endpoint in a more human-friendly format. so if you just wanted to see a list of identifications – not charts and other visualizations – you could use it. for example: https://jumear.github.io/stirfry/iNatAPIv1_identifications.html?current_taxon=false&rank=species&observation_rank=genus&current=true&quality_grade=research&taxon_id=36362&observation_taxon_id=36362.

1 Like

@pisum, wow that is really neat and a good visual display of that too…really cool, thanks! And thanks for the info on how to do the search, I was never actually sure how to specify that. The different search flags have always confounded me a bit, so good to have some one more advanced show me how it is done! Thanks again!

2 Likes

Yes, this window listing all misidentifications shows that both ways are taken into account. Thanks for clarification.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.