Pretty much what you would expect here, the places with the most species happen to also be the highest in biodiversity and human population (more people = more iNat observers = more species found).
Obviously, southern Appalachia holds the greatest number of species of Salamanders, but I was previously unaware of the diversity in Alabama and in California.
Nice maps! You might be interested in rarefaction analyses that can help in partially taking into account varying observer effort across states. See the iNEXT package and this paper https://doi.org/10.1111/2041-210X.12613.
Thanks so much for the links, I’ll definitely have to fiddle around with that package tomorrow! I actually have been trying to find a package like iNEXT for a long time haha.
Having been a little kid spending most of my time outdoors and muddy in coastal northern California, on the southern margin of the Pacific Northwest, the salamander diversity of that region has always been one of my favorite things.
It was exciting to work in Shenandoah NP a while back and be in an area that also has high diversity, but I didn’t get to see as many as I’d have liked (too much time on rocky outcrops instead of damp areas).
If you ever make it out to Northern California during Orange-bellied Newt (Taricha torosa) breeding season it can be pretty impressive. I’ve seen tangled mating balls of dozens of them slowly tumbling down shaded streams and once, high in the Marble Mountains, came across a small alpine lake with thousands of them congregating.
Very nice charts. I’d like to generate some for my place of interest so how do I go about easily obtaining a list of standard place_id values within a country? Doesn’t seem possible via the API.
Not going to lie, the US states are simple because their place_ids are just 2-52, but I’m working on a script to pull place_ids using the API right now, it’s just a little cumbersome.
if you want to automate that process a bit, you could save off a(n abridged) version of that CSV file or put it on a server somewhere, and then you could parse through it programmatically for any country and admin level. the standard administrative places shouldn’t change often. so you typically wouldn’t need an up-to-the-minute version of the CSV file.
Cool stuff whimbrelbirder! When I looked at iNat data across Alaska boroughs (the equivalent of counties), here’s a twist that I found related to the relationship of iNat observations and people within a spatial unit at that scale:
More observers & more observations per borough → more taxa recorded on iNat. In general, there’s a positive relationship between the number of taxa recorded in boroughs on iNaturalist and the number of observers (R2=0.47) and observations (R2=0.87). For every additional observer in an area, roughly about 13 additional observations and 1.5 additional taxa are expected to be recorded on iNaturalist. The boroughs with the most observers? Anchorage (1,023 iNat observers; Alaska’s largest population center with almost three times the residents than the next biggest borough), followed by Kenai (910 observers). The boroughs with the most observations? Sitka (38,505 observations), followed by Kenai (12,526 observations). The boroughs with the most taxa recorded? Same pattern as observations: Sitka (2,740 unique taxa), followed by Kenai (1,680 unique taxa).
More people living or visiting a borough for nature → more iNat observers. What explains variation in iNat activity across Alaska boroughs? I spent a bit of time compiling datasets on factors that I suspected might be important: population size, area, visitor volume, visitor volume engaged in nature activities (including wildlife viewing, birdwatching, hiking), broadband service1. Of those, the two factors that seem the most predictive2 in terms of explaining # of iNat observers are (a) borough population, and (b) number of visitors that engaged in a wildlife viewing activity in the borough3. Multiple regression analysis indicated that those two predictors explained a decent 84% of the variance in the number of observers between boroughs. So, there are more iNaturalist observers in boroughs with more residents and more visitors that want to see wildlife – makes sense, right? This appears true even when controlling for other factors.
More people living or visiting a borough ≠ more iNat observations…. except…. . In contrast to iNat observers, the same pattern does not hold true for iNat observations when looking across all boroughs. In fact, none of the factors I looked at significantly explained differences in observations between boroughs EXCEPT when I dropped a single borough from the analysis. When I looked at all boroughs except for Sitka, the importance of population size and wildlife-viewing visitors re-emerged as significant variables, with a simple model explaining 75% of observation variance4. So, what makes Sitka special?
true, true