Yes, I am aware that a system exists. What I’m saying is the current system is insufficient as is, as demonstrated by people having to manually mark 1,000s of individual identifications beforehand (like the original poster does with gymnosperms).
Instead, we need a system where species can be nominated. For example, someone says “Norway Spruce is basically exclusively cultivated in Ohio. Let’s set it so that all observations are automatically marked cultivated” instead of relying on people manually annotating what could be thousands of cultivated photos trying to hit the 80% mark.
(Keep in mind too, that that 80% going to become a higher and higher bar to climb, in terms of total observations needed to annotate, as inaturalist becomes more widely used).
I am afraid I find this faintly ridiculous. Most of the countryside in the UK is managed. Miles of roadside and fieldside hedges are planted. Many of the rivers in Essex have willow planted along them. Sometimes these occured decades and perhaps centuries ago. Where do you draw the line, one year, two? ten? The fact is humans have completely dominated and transformed the natural environment in the much of this country.
The original plants are never wild, because they were planted. However, there’s a work-around.
My approach to this managed landscape (including our planted Douglas-fir forests) is to remind myself that some of those plants came up on their own (sometimes from seed from ones that were planted). I can’t know which were planted and which weren’t. Therefore, I mark them wild.
Also, some plants are only cultivated in their early years. When an area was scheduled to be leveled for highway construction, my father would go out before the bulldozers rolled in and dig up saplings of species that struck his fancy. He transplanted them into the small woodlot at the back of the property. They got watered for a few months after which they were on their own. If those trees are still there, the current property owner would have no way of knowing that they were originally “cultivated.” They are species native to the region and they’re just out there living in the woods.
I’ll just say, “theoretically” this is a good practice. However, this takes away precious identifier time away from wild observations. Wild observations has been the focus of iNaturalist from the very beginning, is my primary focus, and is hopefully the focus of most naturalists on this site (it is iNaturalist after all and not iGardener). These days, it’s nearly impossible to find interesting observations past all the cultivated and super common species. The cultivated ones are the most frustrating because they shouldn’t be there in the first place. The observer’s decision to intentionally not mark these as cultivated equates to a violation of the community guidelines in my mind (i.e., deliberately adding false information). The only reason I ID cultivated observations these days is to keep wild plants from being flagged as cultivated by iNaturalist. It’s just not worth my time unless someone specifically asks me.
I also want to emphasize the cost from a research perspective when “Research Grade” observations of cultivated species get sent to GBIF. When I talk to researchers that use iNaturalist data, our inability to get rid of cultivated species is the number 1 complaint. It causes all kinds of problems with niche modeling. My advisor spent months curating a dataset for one of the grad students in my lab partially to get rid of cultivated species.
I don’t have a problem with iNaturalist accepting cultivated species, observers who post them, or identifiers who are interested in them. But please understand that not sending these observations to casual grade places a significant burden on identifiers and researchers and causes tension in our community between research-focused identifiers and casual observers.
As for the “gray zone” of cultivated plants, if it’s ambiguous, it’s ambiguous. Deal with it however you see fit. That’s not what I’m talking about here. In my opinion, it’s not a good excuse for leaving plants that you know are cultivated as not cultivated. I mean no disrespect when I say this, but this is my position on the issue.
Considering iNaturalist is intended for the observation of wild organisms, I think it would make more sense to automatically mark all the observations of observers with high rates of captive observations captive / cultivated by default.
This is actually already a feature. I don’t remember what the cut off is, but if a high percentage of the species in a place is cultivated, iNaturalist will consider it cultivated. Unfortunately, it’s not enough to keep the flood of cultivated observations at bay.
Yes, but I’m talking about the observers, not the taxon. Theoretically, observers should have a very low percentage of captive observations since iNaturalist is meant for wild observations. What I am suggesting is if an observer has a certain percentage of captive observations, all of their observations by default would be captive, and they can check it as wild if that is so the case.
I don’t mean to go off topic, but I would rather not have to manually check every observation I make to ensure it hasn’t been incorrectly marked captive upon upload.
Experienced observers who know how to use iNat will often post Casual observations. The key point is whether the user is diligent about marking them as captive or whether they neglect this or deliberately misrepresent in order to get an ID. So if something along the lines of what you’re suggesting is done, let’s make sure it only counts observations that are corrected to captive/cultivated, not ones that were originally entered as such.
There are so many gray areas here, as has been discussed many times, but still there’s a huge category of observations of landscaping plants that are 100% obvious. I will try to be a bit more aggressive and mark them captive where at times I gently suggested “Please mark this as captive/cultivated” out of a misguided sense of politeness.
The one area that is still gray for me is forestry and restoration work, where the plants are actually native plants. I do this kind of work myself on my own land, and always mark them if they are planted. However it is difficult to assess this when looking at observations of others when there’s a deliberate intention to restore “natural” conditions and therefore “deceive” the naturalist. Also many observers resist having their rare and interesting forestry trees discounted just because they are planted in a managed forest.
I keep typing and deleting broad comments along the lines of,
“Don’t you think that interest in What plant is that on the lawn? qualifies as engagement with nature?”
“What about What breed of cow is that in the field? Doesn’t that show curiosity about the animal world, and isn’t it a form of engagement with nature?”
Will these “low-hanging fruits” get picked up by Google Lens and other AIs? Will those AIs scrape iNat data to become the be-all and end-all of organism ID?
Just some things that have been bouncing around my head.
Hi Erika - I used this message for a couple of observations users had made of cultivated plants in my local botanical gardens. As I know many use INaturalist to assist with plant ID learning I think it is very useful to alert people to this - especially as data in Tasmania, Aus is transferred to Atlas of Living Australia, which is often used for species distribution mapping.
Thank you!
I agree it’s a bad practice to intentionally mark cultivated plants as wild. However, I think most of the observers posting unmarked cultivated plants just don’t know any better. I’ve gone through entire collections of observations lately that have lots of unmarked cultivated plant observations and found that not one identifier left the observer with a note explaining how cultivated plants work on iNaturalist. So if we want to reduce the number of unmarked cultivated specimens, we need to inform the observers with notes.
On the other hand, it’s important to keep in mind there are some legitimate reasons for posting cultivated plants, such as documenting the species corresponding to animal interactions (pollination, predation, parasitism) or documenting where cultivated plants are capable of growing. So long as each cultivated observation gets marked as such.
That’s kind of just the sad story of nature on the English Isles. Most wildlands on those islands were exterminated centuries to millennia ago, and if there simply aren’t wild plants there, there aren’t wild plants there. You can’t just ignore the fact that the natural landscape has been nearly completely destroyed or altered and just upload cultivated plants as wild because “that’s the best we’ve got.” They still are essentially just a large scale garden, and should be marked as such. Only truly wild, non-human planted plants are actually considered wild.
Fun thing about fungi is that they are almost never captive or cultivated (at least the majority of observations here on iNat) so there is rarely times when fungal observations need to be downgraded to casual. But that’s not really on topic (sorry)
I bet you could geofence urban and suburban areas to automatically filter many of these. In the US, a good starting point would be the parcel-level lot descriptions from ReGrid. Very likely that an observation on a private urban parcel is cultivated, and very likely that urban wild parks could be identified either by a city or state owner name, or am unusual density of observations, or some combination. Public lands are pretty clearly identified owners as US, state, USFS, DOI, BLM, National Park, etc.
I post a LOT of weeds in urban areas, from private lots and from roadsides, etc. And the mapping usually isn’t precise enough (accuracy circle small enough) to distinguishing between private lots and adjacent roadsides, vacant lots, etc. I don’t think this will work.