it may be worth noting that visualizing the ecoregions as a map background or basemap (this request) is conceptually different than adding the ecoregions as a set of places in iNaturalist (which would allow people to search for observations by ecoregion).
if you just want to visualize the ecoregions as a background, then i think the best request would be to have iNat add a way for users to specify their own basemap tiles (as described here: https://forum.inaturalist.org/t/use-openstreetmap-maps/2588/74).
if youâre wanting ecoregions to be created as a set of places in iNat, then thatâs something different than whatâs contemplated in this request, and you should probably make a separate request. but as bobmcd notes above, marine places have been been requested to be added, amd thereâs been no action on that as far as i can tell:
Iâm curious⌠why would raster tiles be a better solution than using the vector data? In other GIS contexts, I think creating a raster of the ecoregion boundaries for the purpose of visualising a shapefile would be a pretty unusual and inefficient way of doing it. However, I donât know how well-developed iNaturalistâs vector display abilities are. âWe canât display the shapefileâ is the only reason I can think of to go with a raster in this context.
iâm not sure exactly where youâre going with this, but in most cases, the things that you see in iNat maps are delivered via tiles. so having a mechanism that allows a user to specify their own set of tiles is relatively close to what the system is already designed to do, meaning that should be relatively small effort to implement such a change. and once implemented, that would allow someone to use all sorts of existing tile sets available out there for all sorts of different kinds of data â different flavors of OpenStreetMap maps, different kinds of ecological data, etc.
a web map and a desktop GIS application are different beasts. tiles allow for relatively efficient delivery of data to a client on just about any device, and someone has already decided how to best represent the data visually â colors, labels, lines, etc. â at different scales.
if youâre thinking of trying to load a shapefile into a web mapping interface, itâs possible of course with the right code implemented, and with a client setup capable of handling the given shapefile, but without doing a lot of extra configuration and setup to tell the map how to render the features and labels, etc., you wouldnât necessarily end up with a very useful visualization by default.
Integrating with iNaturalistâs interface might be easier for a tile service, but I think everything else would be much harder. For instance, I know I can get a vector of this scale into web mapping apps. Iâve tried the same with rasters a couple of times and given up. Same with symbology⌠I know how to get it non-awful as a vector, as a tile service sounds like a bit of a nightmare. OTOH, this could just be my lack of familiarity.
And I know iNaturalist does vector display, since thatâs what this is:
because the tiles are formatted specifically to work well for visualization, you have a much better idea of whether the observation pins fall in or out of the polygon here. also when you compare the boundaries, you can see how much the boundaries in the first screenshot were simplified (to make sure that all devices could handle the vector).
this is all dealing with just one feature, too. imagine you had multiple features to visualize. in that situation, you could preformat your tiles to visualize one feature one color and the other feature another color, or you could have labels that help you figure what was on different sides of the boundaries. in a GIS application, you might be able to just click on a feature to interact with it and get its details, but in this kind of simple web map, youâre going to be relying on the map symbology and labels, etc. to tell you what different things on the map represent.
iâm not saying that tiles are going to solve every visualization need, but my point is that given a good set of tiles and the flexibility to incorporate those tiles into a given web map, youâre generally going to end up with better visualizations for web map purposes. if you want to do something more complicated, thatâs fine, too, but thatâs when youâre going to be working in a GIS application anyway.
Interesting, thanks! That seems to answer my questionâthe iNaturalist vector handling is pretty limited. There seem to be web maps that are pretty capable with vectors, and can handle clicking on features. But I just know that it appears to be possible, I donât know anything about how difficult it is. I suppose it is more difficult than I would have guessed.
Do you have a solution for getting a map tile service online? Last time I tried getting rasters online, I made a local map tile set that I could look at with desktop GIS, but I couldnât find a viable option to do anything else with it.
if you can think of it, there is likely some way to implement it. itâs always just a matter of how costly it would be to build and maintain. i think it would be a stretch to ask iNat staff to specifically add a visualization just for ecoregions in its maps, especially with interactive functionality, because that seems like it would be a lot of work to do something that â to me â seems relatively far from the core mission of iNat. but who knows? maybe they will decide that itâs worth the effort to do it. (personally, i probably would never use such functionality, or if i did, i would use it only very rarely.)
a web search is going to provide you better advice than i could provide here.
This is exactly what Iâm talking about, thank you for taking this discussion deeper.
Ecoregions arenât a niche thing, they are critical to understanding our immediate environments. Our immediate environmentsâ natural boundaries donât match county boundaries, in many cases. Being able to see results that fall WITHIN the boundaries, or OUTSIDE of them is a big deal.
Here is an example: in my area, it is popular to encourage the planting of Zizia aurea (Golden Alexanders) as a host plant for a certain butterfly. In looking at iNat, using the shape that was (wonderfully) created by someone else in 2019, and viewing that plant, it becomes clear that plants that may (or may not) be Z. aurea are only found OUTSIDE of the Southern Limestone and Dolomite Valleys (Ecoregion 67f) and typically on top of ridges. In asking for IDs of the few sightings from the ridges, it has become clear that the plant being identified there is actually in the genus Thaspium. So there are NO sightings in either of those Ecoregions thus far, for this plant. That would indicate that Z. aurea is not a plant that should be suggested in 67f at all, and that there must be another host plant that does occur in 67f naturally. This tool should make it easier to identify the local, native that performs the needed function for the butterfly (and is likely the plant that would import the most health and vigor or the offspring of the butterfly for our conditions here.)
I am lucky to have the shape already in place, but others likely do not have it, and thatâs a shame. Thereâs so much more to know. We can have a much more finely-tuned understanding of local ecosystems, RIGHT NOW, with the item suggested above by @pisum.
As a side note, it takes a bit of doing to see the results using the shape, as it currently stands. One must bring up the shape, then bring up the species, to see where the species falls on the map with the shape outlined. If the Ecoregions were offered in a more robust fashion I assume one could review a species and automatically see the Ecoregion boundaries on the map, if the Ecoregion Map view were chosen. I would think it could also serve up the list of Ecoregions in which the species has been identified, which would make it useful for creating plant tags and seed packaging that clearly lets the purchaser know where that plant is naturally found. There are all sorts of benefits to putting the effort into creating an Ecoregion Map view.
i assume youâre talking about having some mechanism for users to specify their own map layers to be visualized in iNat maps, not specifically adding an ecoregion layer to iNat maps.
personally, i think ecoregions are niche, and hereâs why:
although iâm not denying that itâs nice to be able to visualize organism distributions in the context of habitat in a few situations, most users in iNaturalist arenât going to ever do this, even if they could do so easily within the system
there are lots of ways to visualize habitat, and ecoregions are only one of them. some folks might like to use Ecoregions, but some folks might prefer some other type of visualization.
if other folks really want to see observation distributions in greater context, there are already ways to do this. as iâve mentioned before, just about anyone who has a device with internet access can visualize iNat observations over whatever other map layers they like relatively easily and quickly.
iâve also mentioned above that things like EPA ecoregions seem like sort of a dated concept to me. you almost need a manual to decipher exactly what each ecoregion means. i would much rather see the kind of visualization shown below, complete with waterways. i can view this without any keys and still get a good idea of whatâs going on:
compare that to something like this, in which the orange polygon tells me nothing about what could be included in its boundaries unless i know in advance what that polygon is supposed to represent:
if you want to visualize against ecoregions specifically, itâs pretty easy to do this in ArcGIS Online. you donât even need an account, although you can get one for free.
So what Iâm looking at in these visualizations is that the few sightings (that turn out to be Thaspium) are seen in the Ecoregion that has âfingersâ within the orange polygon. Zooming in you can definitely see that. Ecoregions have distinct soils and mother rock, and are a pretty exceptional system of showing relatively discrete (and sometimes very discrete) ecosystems. When you look at this in the other views you have provided, you lose the distinct differences of the ridgetops.
ecoregions have a place when it comes to habitat visualization, but i donât see ecoregions as being so superior to all other visualizations that we should put a lot of effort to promote those while ignoring all the other possible visualizations out there.
I have to admit I donât have the first idea how you are overlaying the data over all these other visualizations. I guess because I am just an average person and not tech savvy.
We discussed this request and are not moving forward with it. We donât plan to host more maps on iNaturalist that iNaturalist isnât using. This is not a judgement on the value of ecoregions or any other kind of map, itâs about whatâs feasible to implement and maintain on iNaturalist.
If anyone familiar with ArcGIS or qGIS wants to make a post in #tutorials for how to map iNat data there, that would be awesome. And perhaps there could be a wiki for map sources people could add to?
