Oh me too!!
2 Likes
just out of curiosity, i made a quick plot of research grade vs verifiable observations (red is high, blue is low):
what jumps out at me here is that Russia seems to have a distinctly high identification rate.
looking at the iconic taxon breakdown, it looks like the big difference may be that plants in Russia are much more likely to get identified. are Russian identifiers just that much better at identifying plants? or is there something else to it?
10 Likes
Much of this is due to the efforts of @apseregin who has recruited plant people to make observations and do identifications, and who has done some studies using the iNaturalist data.
7 Likes
Fascinating. I guess my first question would be how does Russia compare to the world overall in terms of percent of Verifiable observations that are Plants, number of Russian observers as a percent of overall population, and so on. I’m wondering if there are relatively fewer plant observers than in the world as a whole and, if so, do the plant observers in Russia tend to make “better” observations, that are easier to bring to Research Grade? Or do identifiers working on Russian observations have a greater tendency to mark genus-level observations “as could as can be” than IDers world-wide? Or other such possibilities?
There are several factors here. Firstly, the flora of most of Russia is relatively sparse. If you look separately at observations of the floristically richer Russian Caucasus, Altai Mountains or Russian Far East, the percentage of RG will be lower (find Доля бэклога в структуре данных in this ).
Most observations of plants in Russia were made not by random people with smartphones, but by active project participants who know how to photograph a plant so that it can be identified, they are familiar with regional binary keys and can find a description of the plant online. For example, before I got to iNat, I participated in a similar Russian website for several years. The core of the identifiers are professional botanists, herbarium workers, teachers and graduate students of universities, or enthusiastic regional amateurs. And also identifiers from Europe and Asia, of course.Identifier statistics
10 Likes
there are names on your list that I call on for plant groups I don’t know - so that knowledge is generously spread across other countries too.
2 Likes
Looking nice and orange in southeast Texas. Maybe I should go work on North Carolina again… sometime later. I’m taking a break from IDing to work on my obs uploading backlog.
4 Likes
yup, but it might look a little different than you would expect when you zoom in. i can make this publicly available later, if you all are interested in this kind of visualization.
i’ve added 3 months worth of observations in the last week. 4 months more to go…
3 Likes
That’s really cool, worthy of its own forum thread imo. Coasts seem to be relatively well-identified in general. Certainly it’s a combination of things to explain why some places have a higher RG:verifiable ratio, but I also wonder if relatively less speciose areas (like Russia and Alaska) are more likely to be red, somehow controlling for effort.
4 Likes
It certainly seems reasonable when you how many isolated red dots there are in gray (mostly unsampled) areas like the high Arctic. I’m more curious about why there are so many red dots scattered across oceans where there aren’t any islands.
3 Likes
I reckon it’s because on the open sea, these are going to be mostly birds and cetaceans, so relatively ok to identify. There aren’t so many plants and insects available to be observed on the high seas, even if there is a lot of life below the surface.
5 Likes
Yes, I’d like to see a zoomed in view of east Texas and Louisiana
1 Like
i updated my existing iNat map page to include this RG / Verifiable visualization. it can be invoked by adding &view=rgratio to the URL or by selecting the layer from the layer selector.
https://jumear.github.io/stirfry/iNat_map?view=rgratio&verifiable=true&place_id=18,27
that’s exactly right. compare mammals + birds vs plants vs insects.
not sure. i’ve been wanting to make a visualization like this that shows species density, but there’s not an efficient way to get this kind of data in real time. there definitely is more blue near the equator and more red near the poles though.
9 Likes
i changed this to make it more generalizable so that it’ll be possible for the user to specify more types of comparisons.
it is still possible to compare RG to verifiable, but it will require different parameters:
- all RG vs all verifiable: https://jumear.github.io/stirfry/iNat_map?view=subsetratio&verifiable=true&quality_grade=research&compare_exclude_param=quality_grade
- taxon variants: plants, birds, insects
the nice thing about doing it this way is that it’ll be possible to do other kinds of comparisons:
- casual vs all observations
- observations identified by an identifier: semiaquatic bugs identified by mpintar
- app obs vs all obs: Android, iOS, both
- comparison vs a previous year: 2023 vs (2022+2023)
- observations during a given month vs all obs (useful for migratory animals): Monarch butterfly in September
- species vs other species in the same genus: loblollies vs all pine species
- observations in a given genus stuck at genus: pines
10 Likes
So here’s my identfying approach this week: under the “Community” tag, select “People.” This generates a grid of 12 avatars (I assume the 12 most recent activities?). I pick one and see what I can do for that user’s observations. I’m more likely to pick one if it says that they just added an identification, because I know that those are the ones giving back.
7 Likes
i scanned around my local area for places that had a relatively low RG to verifiable ratio, and i found a park next to a high school east of Downtown that seemed like it needed some help:
i started identifying observations in the park and almost immediately regretted my decision to go down that path. it looks like the local high schoolers have filled the place with repeated observations of questionable quality. still, did 100+ identifications (which is a lot for me in one sitting), including 30+ Ligustrum lucidum identifications, and i’ll probably come back next time to continue the slog.
12 Likes
Robinia of Canada
Isn’t that phenology graph beautiful! I annotated every observation I reviewed, which amounts to way more than my total IDs. As you can see Robinia flowers are very showy, hence the much greater number of observations in May. Relative to flowers and flower buds, I encountered few seed pods, even in late summer, autumn and winter. Can’t tell why exactly
Total verifiable IDs
Total RG IDs
Total Casual IDs
Gleditsia of Canada (mostly G. triacanthos)
Once again an amazing phenology graph, albeit featuring much fewer Flowering and Flower Budding annotations. Gleditsia flowers are incredibly discrete compared to Robinia’, which it’s often confused with. However, I encountered more seed pods, presumably because they tend to be produced in greater numbers during mast years and persist longer both on the ground and in the tree.
Total verifiable IDs
Total RG IDs
Total Casual IDs
I must have marked 200+ RG observations as cultivated as I annotated them. C’mon observers!
12 Likes
I jumped in to assist, and caught a copyright violation. I got suspicious when I clicked the compare button, and the observation photo was identical to the taxon photo. 
EDIT: Got a bunch more out of the unknowns. Haven’t found any more cheating.
4 Likes
That is … they think no one will notice?!
1 Like
Possibly, they think that it just has to go un-noticed long enough to get the class credit. Oh, well. 
1 Like