It might handle different locations a bit differently, but whichever option I use (feed insects, easy care, soak up rain, etc) it basically suggests me to just leave the most common weeds grow to end up with a heath covered in brambles/wild roses/various thorny weeds.
Which is logical I guess, if the goal is just to have native plants that grow naturally here. But I suspect few people are looking for this without any consideration to the ornamental and recreational qualities of the plants.
In fact, that’s how I garden: I spray seeds and random plants and see what sticks. That’s literally all I do in terms of interfering; the rest I leave to nature. I don’t understand “weed” as terminology in that context; I am not a farmer, I don’t aim for specific “produce”.
I consider it highly recreational to have transformed the previously dead “garden” and “lawn“ with “colorful one-year flowers” and the need to artificially water and all that strange “active gardening“ maintenance but absolutely no insects and no smells into a beautiful wild garden with buzzing insect life and zero artificial watering in only a few years time. The smell of the different herbs and berries alone is doing much for my quality of life. And the flowers that made it are much more “beautiful” and year-round colorful than the weird catalog almost-plastic tulips that the previous tenant replanted each year to have a few weeks of color.
Your end goal focus is fine and I'm glad it improves your quality of life. Same here whether I focus on edibles or natives in my own garden. But I'd caution you about the "throw random seeds" approach, unless you've sourced those seeds and aren't using retail seed package "for pollinators". Those are notorious for frequently including invasive species.
"See what sticks" is inherently biased toward fast spreaders that will outcompete and outspread other plants regardless of their ecosystem services. Even if they're natives.
But like the parent comment, I can't tell you that your own enjoyment isn't warranted in your own garden. The whole point of Indigene is to try and contribute to tilting the balance a bit more into being more aware and focused on the ecosystem benefits of what you plant or don't plant.
You might have missed filters on the result page that specifically let you exclude plants that spread (width or height), thorns, and even stingers. Those can help you avoid what you describe, but they may not be obvious. Maybe I need to make those more obvious? Or default to them?
I think it's fine to have a human-centered mindset although the point of Indigene is to challenge that a bit and re-center the focus on the ecosystem services these plant provide. Ornamental and recreational qualities of a plant don't have zero value, but they definitely don't have 10 times the value of its ecosystem services.
But like everything, since the site is for humans, I do need to find a balance there. So point taken.
PS: swiftcoder and pavel_lishin had similar feedback about garden goals although with a different focus
Nice idea, but given how limited the coverage is, you could save people a lot of time with a map, rather than everyone having to parse 17 regions to work out if they are in one you cover.
FYI I’m not convinced this is easier than just letting folks find their region via GPS, postal code or city name but it’s now implemented at https://indigene.app/map for folks who prefer visually pickings a region or looking at worldwide coverage.
Although the entire point of the landing page spot finder flow is not to have to browse anything, regions are there for you to explore everything, not just where you are.
Was that not clear when you landed on the homepage? You can use your device location, zip code or town name to find natives specific to your location that way. It should find the region without you have to parse anything.
Curious what (broad) region you're in if you don't mind sharing. Just added Scottish Lowlands yesterday via 1.1 (https://indigene.app/release-notes/1.1/), but I need to tackle some Asian or African regions at some point as well. There's obviously a lot of North & South America to tackle as well...
This is very cool! I got into native plants a few years ago when I moved out to the country. I have been using LLMs to help me with what to plant and where but they aren’t perfect. A couple things that’d be handy / cool to add:
- more info on deer hardiness
- more info on growth speed (show only fast growers)
- would be cool to add a feature for sets of plants too like “good for wild turkeys” or “plant these for all season pollination eg a set of plants that bloom early, middle, late”
All excellent feedback thank you! I was just talking with a friend who's helped me test a few things about the wildlife support focus, which excites me too because I've been trying to see what particularly tickles Steller's jays. They show up at pretty specific points in the season but other than bird seeds I've never really noticed a specific plant tie.
For deer hardiness, do you mean a way to find more deer hardy plants directly?
Love the idea for fast growers, especially to counter boxstore-bought yews and other non-natives people pick up because they want fast-growing shrubs for privacy hedges.
Bloom-timing is a cool one too, I already love the info I get from the propagation section (https://indigene.app/planting) on timing specific techniques and I added plant-specific propagation timing fairly recently (https://indigene.app/plants/symphoricarpos-albus#propagation), but having something equivalent for what to expect in a given season so you might have overlapping ecosystem support instead of one big bloom and nothing for months.
For deer hardiness, I meant for plant info pages something like “deer resistant but cage young plants, or deer avoid, or deer favorite —needs protection” I live in an area with high deer pressure and they eat aggressively, so I tend to avoid planting something that will need to be in a forever cage if its exposed to deer.
I would strongly advice to just pay a few bucks to somebody with experience for a quick consulting job in place.
Garden design is painting with life beings, substrate, and time. (With a paint that has its own ideas about remaining where you put it or doing what is expected).
Is basically an art; and AI is a really bad choice for doing art (unless you want to be chained to the safest, blandest, more average trophy garden possible. The same boring garden that everybody has).
A wrong garden design may make your next thirty years pretty miserable and even cost a lot of money trying to fix the mistakes. A wrong design can help to burn down your house.
As a bonus befriending a real plant lover and collector, if you can find a real one, can be also a treasure of rare green gifts. The problem is finding one of those elusive garden gnomes because there are plenty of pretenders and sellers, and just a few real wizards.
I don't think we disagree but none of what you describe is what Indigene does: it's not doing art with AI. You're mixing the development tools with the actual mechanism of the system, which in no way uses LLM output for suggestions presented to users. Indigene doesn't design gardens and I'm not sure I ever want it to. It's a spot and condition-specific tool that lets unfamiliar people discover the most ecosystem-positive specimens they could plant. Sure, it helps focus your efforts, but it's not a whole garden design tool. Far from it.
Landscape designers who are familiar with natives don't cost "a few bucks". I've hired one and not even in a particularly expensive area (Central Florida) just a year ago. So I think we have a different target audience in mind. If you're ready to spend hundreds or thousands of dollars on native landscape design assistance, Indigene is unlikely to be the thing that pushes you to action. You've already been convinced.
On the other hand, encouraging people to iteratively rewild home gardens with natives with the the Doug Tallamy "Homegrownt National Park" (see https://indigene.app/homegrown) frame of mind isn't something that's hinged on whether a garden would earn landscape design trophies. A few native plants can already have an outsized impact on the native wildlife ecosystem in a given area.
I'd second your pitch for befriending folks that are also into native plants (or any plants) however. That can start with planting a handful yourself and offering seeds or cuttings to neighbors too.
Wow this worked really well for me. It suggested some plants I already have from the botanical garden recommendations, and a few new ones I've considered but haven't tried yet. Now I know how they would interact with the bugs and critters too!
If I may offer a suggestion, it would be great if I could exclude trees from the list of suggested plants - it's much easier for me to print wildflowers and things like that, than an actual tree, where I live.
Who doesn't love a tightly scoped request like this. Would an anti-filter mode do the trick for you?
Click twice on a filter to flip it to exclude instead? I can noodle on it for sure.
Did you try the print stylesheet? I did some work on it a few months ago but I really like the idea of being able to print out instructions, especially for the propagation technique pages.
Other super fun fact, do you know there are standards for needle punctures? hexarmor is expensive but its about the only thing I've found that will stop autumn olive woody thorns. They will punch through canvas and leather. Probably also useful for locusts.
Welp you're in luck because my friend Rikki Ocampos who's been helping with Indigene told me yesterday about CISMAs:
> CISMAs are organizations for different management areas, they’re part of the Florida Invasive Species Council.
I told her guidance on "unplanting" invasives is definitely something I want to tackle but it can be tricky to navigate so the first update is this PR (https://github.com/olivierlacan/indigene/pull/247) providing better guidance on whether, when, how to "unplant" invasives and whose permission you might need.
Definitely curious about any and all invasive management resources so thanks fo the https://woodyinvasives.org link, I don't think I was aware of that one yet!
Invasive worldwide are going to be such a wrench. The simple fact is there's no going backwards, and accelerating forward (with importing biocontrols) takes a lot of time and research to make sure you're not making the problem worse.
In the middle we're simply delaying. And each plant x biome has slightly different advice. We had a hard winter last year, and the damage done to autumn olive here in West Michigan was excellent. If we had consistently cold winters it wouldn't be a problem this far north, alas climate is going the other direction. If we had a native fungus like we have for our swamp alders, or a disease to prevent sprouting, the alders and the ash would out compete it.
As it stands the "best" way to remove large trees is via tractor due to its shallow roots, with dry summer tricloyr sprays on the root edge resprouts. (and never, ever backhoe it out, the disjointed root system splits and no longer spreads the triclopyr). And we have to worry about the permaculturists introducing things like seabuckthorn, another Elaeagnaceae that will probably be worse.
No zip code input, no zoom out. It's either scroll repeatedly or allow you to get my location so that I can check some location in California. No thanks.
It's broken. I entered a zip code (sacramento) and it said it needs more detail but can't offer a map any closer than pittsburg to narrow. Pick by region is a dead end if that's not your region.
That's what I get for trying to limit myself to client-side APIs but I'll take your repro and try to address it. Here's the tracking issue if you're interested: https://github.com/olivierlacan/indigene/issues/251
Yeah, that's not great. You missed "Don't want to use your location? Search by town or pick a region instead." because it's lower on the page (see edit below for fix).
I specifically limited map levels to reduce the app's size and that's a big downside.
I don't get your location, your browser does since everything is client-side by the reverse geocoding API definitely does for the map (and you can't take my word for it) so let me see if I can make this less annoying.
No point of trying to be privacy-conscious if the alternative is a pain in the ass, no argument there.
That's pretty nice. I had to use mediterranean France as the closest analogue to North Atlantic Spain/Portugal, but the recommendations look about in line with our native flora.
However, I don't think any of the stated goals lined up with my use case here - in this specific region, we're trying to deal with rapidly warming/drying summers, and I'd be looking for recommendations to help with soil water retention, shade, etc.
Thanks! There's actually a "Soak up rain" and "Stop erosion" goal for the spot-centered default flow to find natives, but it seems like your need is a bit more heatwave resiliency focused?
The alternatives (get it? https://indigene.app/alternatives) section focuses a bit more on the swapping non-natives or turf grasses that do poorly or require intense irrigation during heatwaves, but I really like the idea of adaptation planting.
Generally most natives are inherently better suited for heat waves and warming/drying summers since they tend to need little to no watering once established, but I agree it'd be nice to get a subset of plants that can create shade and help create a bit of an oasis.
I think in areas experiencing rapid climate change we're having to go a little further - augmenting with plants from more desert-like climates to help our natives out as the summers go from traditionally rainy -> drought
as a native plant nerd and GIS dev i love the idea
however
i think for an experienced developer using claude the app itself would be easy to build
it is the underlying domain knowledge of "what plant actually grows where" is the actual hard part and im not getting the impression that your system is reliable in that area yet
Sure. But this same argument goes for every open source project ever built. At some point some convergence is in our best interest one some things, or we all burn tokens recreating the same thing until the planet is mostly data centers.
You'll note I'm not reinventing things, most of this project is connecting dots between open science and already amazing projects like iNaturalist. But if you're familiar with it, you might also be frustrated with its extremely simplistic approach to identifying natives.
The reason I suspect is why your last point exists, except I've spent an inordinate amount of time and effort ensuring that Indigene doesn't claim a plant is native to a region (vs. a whole continent) when it's not. That makes it less comprehensive than other resources, but unless you can point to specific issues it doesn't make it less reliable.
I can't make you care, but if you might this is where I'd look if I were skeptical:
First, really cool project and love what you're doing.
I'm working on something VERY similar, but focused on a particular country, so I'm in the weeds (hah) on this and here's my feedback:
* I absolutely do not trust the data provenance here, and reading through all your links doubly so. You cannot trust claude to do analysis. Not only does it misattribute often, but fake sources and poor interpretations seep in. Without url sources tied to each claim it's inevitable that it rots. You'll have no way to keep track of claims across plants, and over time things slip.
I did see your data structs in github, but a lot of it is just LLM summaries of what it read. This isn't bad (and at some point necessary), but Claude in particularly is notorious on tasks like this, that don't pass human evaluations.
HOWEVER, that said, it's probably okay because it's not a life or death subject. Claims like "full sun" aren't very usable as is, and these can just serve as pointers for people to go research plants on their own.
Thank you for the feedback, I'd love to see your thing and what angle you're focusing on. I see what you mean, but I've never trusted Claude to do analysis. And I'm still only trusting it to help me build acceptance criteria that I can validate against a multitude of region-specific sources for different data points:
What is your skepticism based on? If you've identified a gap I'd be very obviously interested to close it. But if you think I'm allowing Claude to editorialize without verification, then I can assure you I'm not. I've definitely caught it trying to skip steps and "leave out" verification steps, especially for new region coverage where we have insufficient data for a clear cut boolean answer.
And like everything, I keep expanding the rings of what I cross-reference for each region. Hopefully the more keen eyes on this project will mean folks can spot fallacies and holes. But to be frank, I'm in this for the long run so I very mulch welcome it :-)
It works well in the sense that it suggested to plant all the wild plants that I see daily in the woods and nature around me.. on the other hand I guess I could have come to that conclusion myself :D .
That seemed to be my story as well, which is good. I'm low on nannyberry and ninebarks, and it missed tamarack and poison sumac which grow together, and sweet gum which is endangered, but overall good.
For people with degraded ecology though this is a great resource.
That actually may be more due to the fact that I've tried to ensure we have sufficient cross-referencing before we claim things are native to a specific region.
This one deserves a little build-focused post on the blog I referenced in the OP. But in summary, it was a frankly shocking roughly 60-minute session with a surprisingly narrow prompt in Claude Code.
I didn't use a cloud environment which is what I typically do on this project. I set it up to use OpenRouter to delegate to appropriate voice model for the narration (which still blows my mind, especially the French version as a native speaker). That cost barely $2 of credits.
The art direction was simply to remain coherent with both the tone and voice of the project but the visual identity and existing design and illustration elements.
All told I think the final cost was around $30 (at API rates) for the both the US English version and the French version. It was much cheaper to French, because it produced reusable HTML, CSS, and JS assets to animate the scenes in the video.
As someone who's done a lot of tedious video editing work and (lousy) motion graphics in the past for some educational content, it both terrified me (for all the hard-working folks who do this for a living) and made me realize how accessible producing this kind of video content is. It's a bit more palatable given the focus of this project, but I know many content producers, video editors & audio engineers (even voice actors!) who likely wouldn't find this exciting at all.
Given this project has no profit motive, I haven't paid artists (or programmers to make this hit home) to contribute to the project but if people find Indigene helpful (outside of me and a few friends) it's something I'd much rather do. I've done that in the past for Keep a Changelog for instance.
Once you got to results you mean? I just made the numbered steps at the top of the spot result pages clickable so you can use them to navigate back to soil but that may not be obvious. So thanks for the feedback!
Biggest obstacle for me is the giant invasive Ailanthus tree in my back yard. Apparently its roots produce allelopathic chemicals, and I’m finding that some native plants are unaffected (spicebush, wood fern) but others can’t cope at all (asters, cup plant).
I can’t find much if any info on this, so it’s been an ongoing experiment.
Have you considered killing the alianthus tree? I know loosing a big tree is a tough sell, but tree of heaven is one of our worst invasives in the eastern US. Every year that tree is producing wind-born seeds that mean it is going to be causing huge problems for anyone attempting to plant natives anywhere in your neighborhood and actively degrading any natural spaces in the area. You’d do the environment a lot more good just by getting rid of that tree than via anything you plant under it.
Also as an aside, it’s possible that the problem for your cup plant and asters is shade, not chemicals. There are a lot of species of aster so I can’t say for sure there, but cup plant is a prairie species that absolutely wants full sun.
Either way though, please, please, kill your tree of heaven.
It might handle different locations a bit differently, but whichever option I use (feed insects, easy care, soak up rain, etc) it basically suggests me to just leave the most common weeds grow to end up with a heath covered in brambles/wild roses/various thorny weeds.
Which is logical I guess, if the goal is just to have native plants that grow naturally here. But I suspect few people are looking for this without any consideration to the ornamental and recreational qualities of the plants.
In fact, that’s how I garden: I spray seeds and random plants and see what sticks. That’s literally all I do in terms of interfering; the rest I leave to nature. I don’t understand “weed” as terminology in that context; I am not a farmer, I don’t aim for specific “produce”.
I consider it highly recreational to have transformed the previously dead “garden” and “lawn“ with “colorful one-year flowers” and the need to artificially water and all that strange “active gardening“ maintenance but absolutely no insects and no smells into a beautiful wild garden with buzzing insect life and zero artificial watering in only a few years time. The smell of the different herbs and berries alone is doing much for my quality of life. And the flowers that made it are much more “beautiful” and year-round colorful than the weird catalog almost-plastic tulips that the previous tenant replanted each year to have a few weeks of color.
Your end goal focus is fine and I'm glad it improves your quality of life. Same here whether I focus on edibles or natives in my own garden. But I'd caution you about the "throw random seeds" approach, unless you've sourced those seeds and aren't using retail seed package "for pollinators". Those are notorious for frequently including invasive species.
A few examples among many similar studies:
https://neobiota.pensoft.net/article_preview.php?id=118480
https://www.washington.edu/news/2002/04/18/wildflower-seed-m...
https://www.dtnpf.com/agriculture/web/ag/crops/article/2016/...
https://wssa.net/2017/08/make-sure-your-seed-mix-is-tested-f...
https://pmc.ncbi.nlm.nih.gov/articles/PMC8176916/
"See what sticks" is inherently biased toward fast spreaders that will outcompete and outspread other plants regardless of their ecosystem services. Even if they're natives.
But like the parent comment, I can't tell you that your own enjoyment isn't warranted in your own garden. The whole point of Indigene is to try and contribute to tilting the balance a bit more into being more aware and focused on the ecosystem benefits of what you plant or don't plant.
You might have missed filters on the result page that specifically let you exclude plants that spread (width or height), thorns, and even stingers. Those can help you avoid what you describe, but they may not be obvious. Maybe I need to make those more obvious? Or default to them?
I think it's fine to have a human-centered mindset although the point of Indigene is to challenge that a bit and re-center the focus on the ecosystem services these plant provide. Ornamental and recreational qualities of a plant don't have zero value, but they definitely don't have 10 times the value of its ecosystem services.
But like everything, since the site is for humans, I do need to find a balance there. So point taken.
PS: swiftcoder and pavel_lishin had similar feedback about garden goals although with a different focus
Nice idea, but given how limited the coverage is, you could save people a lot of time with a map, rather than everyone having to parse 17 regions to work out if they are in one you cover.
FYI I’m not convinced this is easier than just letting folks find their region via GPS, postal code or city name but it’s now implemented at https://indigene.app/map for folks who prefer visually pickings a region or looking at worldwide coverage.
Although the entire point of the landing page spot finder flow is not to have to browse anything, regions are there for you to explore everything, not just where you are.
Was that not clear when you landed on the homepage? You can use your device location, zip code or town name to find natives specific to your location that way. It should find the region without you have to parse anything.
It didn't where I live so this site doesn't do much for me. I'd love it though, I'm a huge fan of the idea.
Curious what (broad) region you're in if you don't mind sharing. Just added Scottish Lowlands yesterday via 1.1 (https://indigene.app/release-notes/1.1/), but I need to tackle some Asian or African regions at some point as well. There's obviously a lot of North & South America to tackle as well...
That's true, good reminder to prioritize that. Thanks.
This is very cool! I got into native plants a few years ago when I moved out to the country. I have been using LLMs to help me with what to plant and where but they aren’t perfect. A couple things that’d be handy / cool to add: - more info on deer hardiness - more info on growth speed (show only fast growers) - would be cool to add a feature for sets of plants too like “good for wild turkeys” or “plant these for all season pollination eg a set of plants that bloom early, middle, late”
Congrats on the launch!
All excellent feedback thank you! I was just talking with a friend who's helped me test a few things about the wildlife support focus, which excites me too because I've been trying to see what particularly tickles Steller's jays. They show up at pretty specific points in the season but other than bird seeds I've never really noticed a specific plant tie.
And for now Indigene only has a group category for Jays, turkeys & woodpeckers: https://indigene.app/wildlife/acorn-birds
Ditto on that "good for wild turkeys" idea.
For deer hardiness, do you mean a way to find more deer hardy plants directly?
Love the idea for fast growers, especially to counter boxstore-bought yews and other non-natives people pick up because they want fast-growing shrubs for privacy hedges.
Bloom-timing is a cool one too, I already love the info I get from the propagation section (https://indigene.app/planting) on timing specific techniques and I added plant-specific propagation timing fairly recently (https://indigene.app/plants/symphoricarpos-albus#propagation), but having something equivalent for what to expect in a given season so you might have overlapping ecosystem support instead of one big bloom and nothing for months.
Thanks so much for all these great ideas!
For deer hardiness, I meant for plant info pages something like “deer resistant but cage young plants, or deer avoid, or deer favorite —needs protection” I live in an area with high deer pressure and they eat aggressively, so I tend to avoid planting something that will need to be in a forever cage if its exposed to deer.
I would strongly advice to just pay a few bucks to somebody with experience for a quick consulting job in place.
Garden design is painting with life beings, substrate, and time. (With a paint that has its own ideas about remaining where you put it or doing what is expected).
Is basically an art; and AI is a really bad choice for doing art (unless you want to be chained to the safest, blandest, more average trophy garden possible. The same boring garden that everybody has).
A wrong garden design may make your next thirty years pretty miserable and even cost a lot of money trying to fix the mistakes. A wrong design can help to burn down your house.
As a bonus befriending a real plant lover and collector, if you can find a real one, can be also a treasure of rare green gifts. The problem is finding one of those elusive garden gnomes because there are plenty of pretenders and sellers, and just a few real wizards.
I don't think we disagree but none of what you describe is what Indigene does: it's not doing art with AI. You're mixing the development tools with the actual mechanism of the system, which in no way uses LLM output for suggestions presented to users. Indigene doesn't design gardens and I'm not sure I ever want it to. It's a spot and condition-specific tool that lets unfamiliar people discover the most ecosystem-positive specimens they could plant. Sure, it helps focus your efforts, but it's not a whole garden design tool. Far from it.
Landscape designers who are familiar with natives don't cost "a few bucks". I've hired one and not even in a particularly expensive area (Central Florida) just a year ago. So I think we have a different target audience in mind. If you're ready to spend hundreds or thousands of dollars on native landscape design assistance, Indigene is unlikely to be the thing that pushes you to action. You've already been convinced.
On the other hand, encouraging people to iteratively rewild home gardens with natives with the the Doug Tallamy "Homegrownt National Park" (see https://indigene.app/homegrown) frame of mind isn't something that's hinged on whether a garden would earn landscape design trophies. A few native plants can already have an outsized impact on the native wildlife ecosystem in a given area.
I'd second your pitch for befriending folks that are also into native plants (or any plants) however. That can start with planting a handful yourself and offering seeds or cuttings to neighbors too.
Wow this worked really well for me. It suggested some plants I already have from the botanical garden recommendations, and a few new ones I've considered but haven't tried yet. Now I know how they would interact with the bugs and critters too!
I'm so glad! Anything you found to be missing? Or that could help you with the new plants you haven't tried?
If I may offer a suggestion, it would be great if I could exclude trees from the list of suggested plants - it's much easier for me to print wildflowers and things like that, than an actual tree, where I live.
Who doesn't love a tightly scoped request like this. Would an anti-filter mode do the trick for you?
Click twice on a filter to flip it to exclude instead? I can noodle on it for sure.
Did you try the print stylesheet? I did some work on it a few months ago but I really like the idea of being able to print out instructions, especially for the propagation technique pages.
I haven't yet - I just took a quick look, and then had to get back to work :/
And I think the click-to-exclude would be fine!
It's in as of 1.1 yesterday, really good suggestion thanks! https://indigene.app/release-notes/1.1/
In restoring old farm ground in Michigan, please send claude robot to kill the autumn olive. I can give it work for years and years.
https://woodyinvasives.org/
Other super fun fact, do you know there are standards for needle punctures? hexarmor is expensive but its about the only thing I've found that will stop autumn olive woody thorns. They will punch through canvas and leather. Probably also useful for locusts.
Welp you're in luck because my friend Rikki Ocampos who's been helping with Indigene told me yesterday about CISMAs: > CISMAs are organizations for different management areas, they’re part of the Florida Invasive Species Council.
https://www.floridainvasives.org/for-homeowners/
I told her guidance on "unplanting" invasives is definitely something I want to tackle but it can be tricky to navigate so the first update is this PR (https://github.com/olivierlacan/indigene/pull/247) providing better guidance on whether, when, how to "unplant" invasives and whose permission you might need.
Definitely curious about any and all invasive management resources so thanks fo the https://woodyinvasives.org link, I don't think I was aware of that one yet!
Invasive worldwide are going to be such a wrench. The simple fact is there's no going backwards, and accelerating forward (with importing biocontrols) takes a lot of time and research to make sure you're not making the problem worse.
In the middle we're simply delaying. And each plant x biome has slightly different advice. We had a hard winter last year, and the damage done to autumn olive here in West Michigan was excellent. If we had consistently cold winters it wouldn't be a problem this far north, alas climate is going the other direction. If we had a native fungus like we have for our swamp alders, or a disease to prevent sprouting, the alders and the ash would out compete it.
As it stands the "best" way to remove large trees is via tractor due to its shallow roots, with dry summer tricloyr sprays on the root edge resprouts. (and never, ever backhoe it out, the disjointed root system splits and no longer spreads the triclopyr). And we have to worry about the permaculturists introducing things like seabuckthorn, another Elaeagnaceae that will probably be worse.
No zip code input, no zoom out. It's either scroll repeatedly or allow you to get my location so that I can check some location in California. No thanks.
"Search by Town" allows zip code. "Pick a region" lists
- Pacific Northwest
- Southern California
- Central California Coast
It's broken. I entered a zip code (sacramento) and it said it needs more detail but can't offer a map any closer than pittsburg to narrow. Pick by region is a dead end if that's not your region.
Recently had to fix a similar issue where a zipcode from Florida landed in Spain (https://github.com/olivierlacan/indigene/issues/219).
That's what I get for trying to limit myself to client-side APIs but I'll take your repro and try to address it. Here's the tracking issue if you're interested: https://github.com/olivierlacan/indigene/issues/251
Edit: should be addressed by this change if you'd like to give another try: https://github.com/olivierlacan/indigene/pull/252
Yeah, that's not great. You missed "Don't want to use your location? Search by town or pick a region instead." because it's lower on the page (see edit below for fix).
I specifically limited map levels to reduce the app's size and that's a big downside.
I don't get your location, your browser does since everything is client-side by the reverse geocoding API definitely does for the map (and you can't take my word for it) so let me see if I can make this less annoying.
No point of trying to be privacy-conscious if the alternative is a pain in the ass, no argument there.
Edit: should be resolved via https://github.com/olivierlacan/indigene/pull/253
That's pretty nice. I had to use mediterranean France as the closest analogue to North Atlantic Spain/Portugal, but the recommendations look about in line with our native flora.
However, I don't think any of the stated goals lined up with my use case here - in this specific region, we're trying to deal with rapidly warming/drying summers, and I'd be looking for recommendations to help with soil water retention, shade, etc.
Thanks! There's actually a "Soak up rain" and "Stop erosion" goal for the spot-centered default flow to find natives, but it seems like your need is a bit more heatwave resiliency focused?
The alternatives (get it? https://indigene.app/alternatives) section focuses a bit more on the swapping non-natives or turf grasses that do poorly or require intense irrigation during heatwaves, but I really like the idea of adaptation planting.
Generally most natives are inherently better suited for heat waves and warming/drying summers since they tend to need little to no watering once established, but I agree it'd be nice to get a subset of plants that can create shade and help create a bit of an oasis.
I think in areas experiencing rapid climate change we're having to go a little further - augmenting with plants from more desert-like climates to help our natives out as the summers go from traditionally rainy -> drought
as a native plant nerd and GIS dev i love the idea
however
i think for an experienced developer using claude the app itself would be easy to build
it is the underlying domain knowledge of "what plant actually grows where" is the actual hard part and im not getting the impression that your system is reliable in that area yet
Sure. But this same argument goes for every open source project ever built. At some point some convergence is in our best interest one some things, or we all burn tokens recreating the same thing until the planet is mostly data centers.
You'll note I'm not reinventing things, most of this project is connecting dots between open science and already amazing projects like iNaturalist. But if you're familiar with it, you might also be frustrated with its extremely simplistic approach to identifying natives.
The reason I suspect is why your last point exists, except I've spent an inordinate amount of time and effort ensuring that Indigene doesn't claim a plant is native to a region (vs. a whole continent) when it's not. That makes it less comprehensive than other resources, but unless you can point to specific issues it doesn't make it less reliable.
I can't make you care, but if you might this is where I'd look if I were skeptical:
https://indigene.app/sources
https://indigene.app/confidence
https://github.com/olivierlacan/indigene/blob/main/DATA_SOUR...
https://indigene.app/plants/penstemon-digitalis (example of a status conflict)
https://github.com/olivierlacan/indigene/blob/main/app/scrip... (which is what does this status checek)
And if you care to call me out on any GIS mistakes (given my own lack of expertise in that area) this where you're likely to find mistakes: https://github.com/olivierlacan/indigene/blob/main/app/scrip...
First, really cool project and love what you're doing.
I'm working on something VERY similar, but focused on a particular country, so I'm in the weeds (hah) on this and here's my feedback:
* I absolutely do not trust the data provenance here, and reading through all your links doubly so. You cannot trust claude to do analysis. Not only does it misattribute often, but fake sources and poor interpretations seep in. Without url sources tied to each claim it's inevitable that it rots. You'll have no way to keep track of claims across plants, and over time things slip.
I did see your data structs in github, but a lot of it is just LLM summaries of what it read. This isn't bad (and at some point necessary), but Claude in particularly is notorious on tasks like this, that don't pass human evaluations.
HOWEVER, that said, it's probably okay because it's not a life or death subject. Claims like "full sun" aren't very usable as is, and these can just serve as pointers for people to go research plants on their own.
----
Excited to see where you get to!
Thank you for the feedback, I'd love to see your thing and what angle you're focusing on. I see what you mean, but I've never trusted Claude to do analysis. And I'm still only trusting it to help me build acceptance criteria that I can validate against a multitude of region-specific sources for different data points:
- host plants (e.g. European caterpillar counts via Gaytán et al: https://onlinelibrary.wiley.com/doi/10.1002/ece3.73004)
- food sources
- pollen interactions
- broadly documented wildlife & plant dependencies
- etc
What is your skepticism based on? If you've identified a gap I'd be very obviously interested to close it. But if you think I'm allowing Claude to editorialize without verification, then I can assure you I'm not. I've definitely caught it trying to skip steps and "leave out" verification steps, especially for new region coverage where we have insufficient data for a clear cut boolean answer.
And like everything, I keep expanding the rings of what I cross-reference for each region. Hopefully the more keen eyes on this project will mean folks can spot fallacies and holes. But to be frank, I'm in this for the long run so I very mulch welcome it :-)
PS: https://github.com/olivierlacan/indigene/blob/main/DATA_SOUR... might be a good one to focus on if you hadn't picked it up yet
> i think for an experienced developer using claude the app itself would be easy to build
Maybe. But if everyone has to vibe code this themselves, it would be a hugely inefficient endeavour. Better to share these projects and help others!
It works well in the sense that it suggested to plant all the wild plants that I see daily in the woods and nature around me.. on the other hand I guess I could have come to that conclusion myself :D .
That seemed to be my story as well, which is good. I'm low on nannyberry and ninebarks, and it missed tamarack and poison sumac which grow together, and sweet gum which is endangered, but overall good.
For people with degraded ecology though this is a great resource.
That actually may be more due to the fact that I've tried to ensure we have sufficient cross-referencing before we claim things are native to a specific region.
Curious which regions you looked into.
The intro video looks very well done. How did you go about making it?
This one deserves a little build-focused post on the blog I referenced in the OP. But in summary, it was a frankly shocking roughly 60-minute session with a surprisingly narrow prompt in Claude Code.
I didn't use a cloud environment which is what I typically do on this project. I set it up to use OpenRouter to delegate to appropriate voice model for the narration (which still blows my mind, especially the French version as a native speaker). That cost barely $2 of credits.
The art direction was simply to remain coherent with both the tone and voice of the project but the visual identity and existing design and illustration elements.
All told I think the final cost was around $30 (at API rates) for the both the US English version and the French version. It was much cheaper to French, because it produced reusable HTML, CSS, and JS assets to animate the scenes in the video.
As someone who's done a lot of tedious video editing work and (lousy) motion graphics in the past for some educational content, it both terrified me (for all the hard-working folks who do this for a living) and made me realize how accessible producing this kind of video content is. It's a bit more palatable given the focus of this project, but I know many content producers, video editors & audio engineers (even voice actors!) who likely wouldn't find this exciting at all.
Given this project has no profit motive, I haven't paid artists (or programmers to make this hit home) to contribute to the project but if people find Indigene helpful (outside of me and a few friends) it's something I'd much rather do. I've done that in the past for Keep a Changelog for instance.
I couldn't find an option to change the soil type, since it guessed mine very incorrectly.
Once you got to results you mean? I just made the numbered steps at the top of the spot result pages clickable so you can use them to navigate back to soil but that may not be obvious. So thanks for the feedback!
Should be addressed soon: https://github.com/olivierlacan/indigene/pull/262
Amazing! Very impressive, I like how the app guided me towards the results page
Very nice! My wife is very much into this, and I'll share it with her.
Thank you for that! Curious what feedback you get from her.
That is cooool.
Love it!
Biggest obstacle for me is the giant invasive Ailanthus tree in my back yard. Apparently its roots produce allelopathic chemicals, and I’m finding that some native plants are unaffected (spicebush, wood fern) but others can’t cope at all (asters, cup plant).
I can’t find much if any info on this, so it’s been an ongoing experiment.
Have you considered killing the alianthus tree? I know loosing a big tree is a tough sell, but tree of heaven is one of our worst invasives in the eastern US. Every year that tree is producing wind-born seeds that mean it is going to be causing huge problems for anyone attempting to plant natives anywhere in your neighborhood and actively degrading any natural spaces in the area. You’d do the environment a lot more good just by getting rid of that tree than via anything you plant under it.
Also as an aside, it’s possible that the problem for your cup plant and asters is shade, not chemicals. There are a lot of species of aster so I can’t say for sure there, but cup plant is a prairie species that absolutely wants full sun.
Either way though, please, please, kill your tree of heaven.
Source: My wife does this for a living.
[flagged]
[flagged]
[dead]