I have gone on somewhat of a crusade against cultivated plants in Hawai’i. At one point I was going through the identify modal spamming the X and right arrow key on species which only occur in gardens in the area. I think that this aggressive approach is needed across much of the world for garden species.
I am rather liberal at marking other people’s observations as cultivated, if it’s a plant that’s commonly in cultivatation, and the GPS point is within 50 ft of a structure, I’m just going to mark it cultivated. I think that the damage done by research grade records which are actually cultivated far outweighs a few false positives of wild plants getting marked as cultivated.
The enormous quantity of cultivated plants which are research grade is a turn-off for some researchers I have talked to.
I’m sorry but in my opinion this is a bad practice and results in the loss of many legitimate data points. I photograph/record weeds + naturalised plants all the time that are within 50 ft of a structure, literally hundreds of my legitimate observations would be erroneously marked as cultivated if you applied this logic. You should only mark an observation as cultivated if the available evidence supports that judgement. Does this require more time investment? Of course yes, but I think it’s worth it. I spend a huge amount of my time doing exactly this for Australian observations, so I understand the time sink, but I think it’s very unfair to observers to ‘blindly’ mark their observations as cultivated without actually checking first.
This practice also hinders a very useful aspect of iNat, which is early detection of new weeds/naturalising species. If a species starts to escape from cultivation, where are the first observations most likely going to be? In many cases, pretty close to the parent plant(s).
First, thank you for the effort you’re doing.
But…
I think this is true in theory… but not in practice, because of the issues raised by the OP. The dataset is already so poisoned that it’s hardly effective for this use case.
Again, I’m with you in spirit, but in practice this model will have trouble because it requires the observers to actually respond. In my experience (asking people to verify the location/dates for preserved specimens or screenshots of flies), I’m lucky if half of them ever follow up. In plants it’s gotta be way worse because “cultivated” is such a murky category.
Adding a pre-emptive comment like you do should keep yours safe, or make it easy to correct if someone like @kevinfaccenda mows through them. But I don’t fault them for trying to clean up their corner of the dataset.
I agree, but I usually at least give people a grace period. Unambiguously cultivated? I immediately mark as such, not need to comment. But edge cases where it’s murky? I’ll leave a comment and bookmark the observation, if no response in a few weeks (and they’ve been active in that time), then I will mark as cultivated
I’m also more punitive (for want of a better word) if a user has already posted a number of other clearly cultivated plants without marking them, and am far more likely to assume that their edge cases are also cultivated and mark without asking
I don’t deal much with plants on iNat, but with animals I always use the criterion that “wild/captive/cultivated” is a description of the organism’s “state” rather than its “origin”. It is not the opposite of native.
If we go in flagging every non-native as captive/cultivated, we lose data about real invasives.
It is a tricky call in many cases.
I have commented on this problem several times in the past and have sent a suggestion to the administrators of Inaturalist. The problem is that the system DEFAULTS to the observation being wild, and many observers do not notice or ignore that check box entirely. If Inaturalist would simply change the default to NOT WILD that would eliminate most of these erroneous postings. Serious observers who care about the research value of Inaturalist would know to mark the observation as being wild when it really is wild.
Since iNaturalist can automagically flag many species as “cultivated” (because “in this area they are frequently cultivated”), why not make it the default behaviour across the various apps/website for all new uploads?
That would be a good start, without making everything cultivated by default.
There are so many species that are not even capable of being cultivated, across the entire spectrum of life, that doing this would create an extremely annoying extra step to have to deal with.
Maybe it would work for a small subset of plants/animals that are more often casual than not, but having to mark ‘wild’ for everything just seems extremely tedious
The idea would be to lower the bar (at the moment plants are automagically flagged as cultivated only above some elusive numerical/geographical threshold) - not flag everything as cultivated by default, of course. Cf. the last sentence of my previous message :)
edit: an example, maybe not the best: Eucalyptus globulus is a native of Australia… but (rightfully) marked by iNat as ‘Introduced’ everywhere else, where it’s planted/cultivated in many cases. Why not automatically enforce the ‘cultivated’ flag by default for this taxa in these areas? (rather than assuming by default that ‘introduced plants’ are exotic but wild nevertheless.) Or simply lower a lot the threshold, so that any taxon that has been marked ‘cultivated’ just once or twice in an area, will immediately default to ‘cultivated’ for all future occurrences in this same area.
It’s a matter of adding some more weight (or making new factors such as ‘Introduced’ add extra weight) towards ‘cultivated’, in the current algorithm.
Thanks for bringing this up, because I have been trying to find an answer to this question. When does a reconstructed prairie, presumably planted with native seed, become a non-cultivated site? What if it has become totally self-seeded? Will it always be cultivated, even if it is self-sustained? Do we have any prairies that are not re-constructed or in some way cultivated anymore? For instance, what if someone goes out and cuts down a tree that has sprouted in a prairie- does that make it cultivated? If a prairie is part of a nature center, but self-sustaining, does that automatically make it cultivated? What if they didn’t add any seed, but did prescribed burning?
What if some good samaritan takes all the invasive plants from a newly established, never cleared, wooded conservation area, resulting in abundant growth of rare natives? Has that become cultivated? Sorry, but I get into long arguments with myself on this. Often I find that answering whether something is cultivated or uncultivated becomes more complicated than it seems at first glance. I, for one would appreciate some clarification on this.
Hi @elaine3 and welcome to the iNaturalist forums!
The questions you are asking are ones that occur to many iNat users, and have therefore been discussed many times in this forum. A search like this will bring up a lot of those threads: https://forum.inaturalist.org/search?q=cultivated%20native
I don’t at all mean to suggest that you shouldn’t ask about specific scenarios again on the forum, but you may get quicker answers to these questions if you look at the discussion in those previous threads. Please do post topics on the forum if the scenario you encounter doesn’t seem to be covered!
I think you misunderstood. If it’s a common cultivated plant AND it’s within 50 ft of a structure. If it’s a common cultivated species far from a structure I don’t mark it as cultivated, nor if it’s a weed close to a structure.
I also will ask people to clarify if things were planted or not if it’s a weird case, but rarely do as most tourists visiting Hawai’i and posting to iNat have no idea what’s cultivated and what isn’t.
No misunderstanding. My point was that there are ‘many’ (proportionally not really, but absolute number yes) cases of common cultivated plants within 50 ft of a structure that are growing as weeds, having self-seeded/escaped from cultivation and started to naturalise (I guess the term ‘doubtfully naturalised’ would be most appropriate here), and it’s these records that are being falsely marked as cultivated when the blanket approach is applied
A classic example I see in Australia is the Norfolk Pine. Extensively planted along coastlines throughout the country, especially at beaches in major cities. 99% of observations uploaded to iNat in Australia are very clearly planted and I mark them all as such. But, there are a handful (and seemingly increasing over the past few years) of genuine ‘wild’ records where seedlings have popped up in coastal dune situations, having self-seeded from planted parent plants nearby. Some of these legitimate records are within 50 ft of a structure. These are important records to help document these new invasions
I absolutely agree that such records exist, but as you mentioned, they are a minority, and unless the observer indicates that they are wild, they’re often impossible to seperate from garden plants.
The truly naturalizing plants are also impossible to find (even if they are labeled / commented as wild) if the map is 99% cultivated records in the area, so I still stand by liberally marking garden plants as cultivated as it will make the truly naturalizing plants apparent.
I have recently been contemplating my own bias when I approach these edge cases.I find I am happy to let the observer decide usually. However, I recently ran into some observations that were new species for my area. They were ornamentals (ground covers) on a residential property. The observer was unsure about whether the individual plants were planted on purpose, but reported that they seemed untended. Based on them doing this in multiple locations, I don’t believe the observer is the property owner.
My impression is that this that when we don’t know and we are adding to the list of locally present wild plants, having some evidence that it actually is wild is appropriate. I have no problem leaving weeds as wild. But weedy is subjective and I think applying some knowledge about the local context is fair.
Of course I don’t know everything about the plants in my area and conditions will change over time. So I remain open to evidence that an observer provides. I always leave a comment when marking something as not wild, so the observer can provide feedback.
There is the danger that a “garden” plant which gets invasive is not recognised as invasive because there are people marking wild offspring as [cultivated/planted] even if they are not cultivated and clearly “runaways”
I agree that information about native/non-native range is important. I would like to know the native range of the species identified, and it could be helpful for deciding how to label the species.
Welcome to the iNat forum.
For many species, this information is present on iNat - there will be a little icon to the right of the species name with an exclamation point if the species is introduced. You can also go to a species page and look at the maps to see its distribution. The “Status” tab will also show the establishments means checklist (though it can be hard to find location on here if a species is widely invasive). If you find any errors, you can flag the taxon in question, describe the issue and ideally provide sources so that curators can update.