I thought people may be interested in a new paper, just published today, that I co-led with fellow iNatter Lizzy Wenk (and Will Cornwell, but David Keith not on iNat) for one of my PhD chapters.
Basically, we compiled a master all time plant species list for both Yosemite National Park in California, USA and Royal National Park in New South Wales, Australia, two of the five oldest national parks in the world, by combining herbarium vouchers + RG iNat records + an official expert-compiled park list to see what each data stream contributed and how complementary the three data streams were for each park. We spent several months manually curating tens of thousands of records across the two parks to correct for errors, trace taxonomic changes, etc.
One of the key take homes was that, to get the best/most comprehensive picture of the plant biodiversity in each park, you need to combine all three data streams, as each had unique species that the other two did not contain. From an iNat perspective in particular, these records were an important supplement, especially for Royal: 63 new species for the park were only recorded from the iNat data stream. Yosemite on the other hand had iNat only contribute 20 new species for the master list (although this is still impressive), in large part because quadruple the amount of herbarium vouchers have been collected for Yosemite vs. Royal over the last 160 years. So for areas with a less extensive collecting history, iNat plays a really important supplementary role.
Thanks! This looks like a useful paper for a project I am working on. I was hoping that you had solved the automated taxon-alignment problem, but apparently not. It would be wonderful if some organization (TNRS?) was able to develop a comprehensive historical synonomy API that could be used at least to resolve one-to-one matches and flag many-to-one and one-to-many matches for hand curation.
yeah that was certainly one of the most time-consuming (and frustrating) elements of our curation process. A flagging system like you describe would at least allow the huge pool of names to wade through to be whittled down to a much more manageable pool of actual changes. Hopefully someone is motivated by the paper to tackle this…
did you generate any figures on the impact of constant taxonomic change on the ability to do this sort of monitoring? I’ve seen it become a big burden making iNat an increasingly difficult tool to use to manage things like species lists as inat’s increasingly bizarre taxonomy diverges further and further from that which parks use, at least in the case of plants.
in the case of Yosemite, the official park list was surprisingly last updated more than 15 years ago (I was surprised by this given the high profile and popularity of Yosemite), and its taxonomy was significantly out of date relative to any taxonomy, iNaturalist or otherwise. The taxonomy used by the official park list for Royal aligned pretty closely to iNaturalist taxonomy of Australian plants, but that was unsurprising given the current taxonomy presented by the Australian Plant Census, the national consensus on plant names, is a roughly 93-94% match to current iNat taxonomy for all Australian plants overall
Thank you for sharing this, I cannot wait to read it! I have been working on a checklist of Coleoptera in Massachusetts since 2018, it includes journal articles, books, and online databases, and the synonomy for some of the 100+ year-old records is maddening. If I spend more than 30 minutes trying to determine if something is a synonym then I will add it as a new species and try to make peace with the database containing a small percentage of errors that I can correct down the line.
I still have not quite figured out how to deal with pulling iNat records that can change over time - i.e. RG observations that are disputed so that something might be Species A in 2020 but then reassigned to Species B in 2022.
Yes in my experience this is the case in most US federal and state entities. Given my own opinions i wouldn’t say it’s necessarily a bad thing, but i’m sure others would feel differently. It’s the main reason i don’t use iNat for work, save as a field notebook type thing.
It takes a really long time for those agencies to adopt new things, and i personally am glad they haven’t invested a whole lot of time in chasing taxonomy but obviously others disagree and this forum isn’t the best place to debate that.
Anyway… i’m glad it was successful in Australia and if i ever go there i won’t know any of the plant names anyway.