Kia Ora from Wellington New Zealand. I’m working on a conservation/restoration project over an area of 3Ha and I’m really interested if anyone has a good way to map individual trees, potentially using inaturalist as a way to record what tree species are found where within our project area (with a high level of accuracy). We do have the benefit of having: https://wtree.landcareresearch.co.nz/ cover our area so we can measure ~how many trees are present but their is no really easy way to enter what each species is. Does anyone have any ideas or ways they found worked to achieve this goal? to clarify I want to be able to map every tree above 3m and what species each is. Is this possible and what way is best? -Thanks so much in advance -Dan
2 Likes
Yes, if you or another person on your project wants to take the time to photograph and upload every individual tree, then you certainly could. Bear in mind that if any of the trees were planted by a person, you will need to to mark them cultivated. If you don’t have the time to photograph them individually (sounds like lot of trees!) you could make observations without photos, which would make the observations “casual” grade automatically. They would still be useful to you, but not really useful for anyone else using the iNaturalist dataset, so this wouldn’t be preferred. If you need help identifying the tree species, then photographs are must.
If you are asking whether anyone knows of a better tool to meet your goal other than iNaturalist… I once worked with a GIS map of many individual trees, which was far inferior to using iNaturalist, at least from my perspective on that project. Creating and editing the map required a lot more know-how and it still turned out pretty sub-par. The search function was not good, refusing to display more than 5 results for any inquiry; we had too many trees to load on one layer and had to split them into two layers; when a person needed to go through every tree to enter the ID, there was no way to distinguish which trees were already completed without clicking each one individually, and of course there was no built-in taxonomic database to automatically complete the species names with correct spelling… so at least the way we were doing it, it was frustrating, although it’s very possible we simply didn’t know the best ways to do it.
2 Likes
unless you have a specific reason to put this data into iNaturalist, i don’t see why you would do so. if you’re just trying to map trees, you have more options and more flexibility doing it in your own setup.
the hardest part is collecting and structuring the data. once that’s done, you can visualize it easily any number of ways. just for example: https://forum.inaturalist.org/t/mapping-of-excel-data/30531/12
4 Likes
awesome thanks for the input, the purpose of such would be to track our outcomes i.e so we can say for certain our effort has led to a decrease of 30% of introduced trees, or increase in biodiversity of large native species over time. I was also hoping have the species displayed we could recognise problem areas or areas where there is exclusively one species.
that’s going to be easier to achieve outside of iNaturalist. nothing you’re saying here suggests to me that you need to or should put your data in iNaturalist.
1 Like
The issue I’m having is finding a way to upload the data inaturalist to me seemed like the easiest option as myself and others in our restoration group are familiar with it and we’re all very inexperienced when it comes to GIS. Is there an alternative or an easy way to learn a program/software? I’ve tried excel but tbh it is “ridiculously confusing” for some of our group trying to get it to do what we want with no add-ins infact it may be impossible but again very inexperienced so don’t know.
I was thinking I could add something unique to the description of the observations or add them to a small project in order to only count the trees we wanted in the area. Also again its mainly for mapping native and wild introduced/pest trees in order to effectively manage them.
Can you put ID tags on the trees?
We use this project
https://www.inaturalist.org/projects/alien-early-detection-rapid-response-s-afr
I flag up what I see when out hiking, and the team gets to them later.
1 Like
we could, do you know if there is a reference as to how that should be done? i.e the correct protocol when doing so?
A quick method could be to use Google My Maps, which we could describe as a very simplified GIS, but this requires:
-
ability to orient yourself on the orthophoto, that is to know precisely where you are at that particular moment - the “ball” of the GPS usually moves, so you can use it to know in general where you are, but not for the position of a single tree;
-
ability to identify the species of each tree, since Google My Maps, as well as other GIS apps, does not have the computer vision of iNaturalist.
there’s data collection and there’s visualization / analysis. look at the thread that i noted earlier. look at the example that i provided to visualize the street trees. since the source data is just a table with coordinates and other fields delivered as CSV, there are lots of different ways to easily visualize that data. if you can’t figure out how to visualize things in Excel, use another tool. if you still want to use Excel to collect the data even if you’re not using it to visualize the data, that’s perfectly fine. but there’s no reason to use a single tool for both data collection and visualization unless you think that’s the best way to do it.
there are some platforms that are good for both data collection and visualization, but ultimately what you choose to use depends on your needs, comfort, and budget.
2 Likes
One benefit in mapping an area with iNaturalist is that other people can use the data you collect in the future. Collecting such a detailed inventory of 3 Ha and not sharing it in an accessible setting puts the data in a scientific cloister. Other people can come along and add to your project on iNaturlist by recording things other than trees, and by making a record of the place over time. For example, a detailed tree inventory combined, with moths and smaller plants.
4 Likes
Could you download the tree-map (or any other sufficiently detailed satellite map) and just label the trees on there?
While it may be a bit more manual work than other options, that seems to me to be the least complicated approach. And I think it would save time compared to making an iNat observation for every tree.
You can still make an iNat project, just looking at species diversity, and make a journal post with the map. That would still give you the iNat benefits without having to have 1000 observations of a single species which may get tiring for you and for IDers.
1 Like
The big question I have for you is how discrete are your trees? Is it easy to say where one tree begins and the other ends, or do you have forest/wood areas where the canopy of individual trees are intermeshed and overlap? GPS signals get degraded under trees, and trees in a forest are much closer together than the error from the GPS, so being able to reproduce the exact or even relative location of a single tree in a forest is very difficult, and getting the forest of points to be more than a mess of nonsense can be quite frustrating. You may need to use laser survey equipment, like a Total Station, etc.. As iNaturalist depends on mobile device GPS, I am worried you will end up putting a lot of effort into mapping a lot of trees that you can’t find again.
If your trees are discrete I would say your best option is to contact Dr Jan Schindler, Manaaki Whenua - Landcare Research. Dr Schindler might be interested in thinking through how best to map the trees, as it could be used as ground truth data to calibrate the AI model that was used to produced the tree canopy data in the map you link to. An example would be that Dr Schindler might be able to add a show my location icon to the map for you, so you can see your blue dot on the map.
If you click on each canopy polygon in the Urban Tree Explorer they all have IDs. As others have noted you could mark each tree with that ID. If you try to do that with say 20 to 50 trees you will get an idea of how huge of a task mapping 3 hectares of trees will be, and whether or not the AI data is any good for that task, especially if you have forest.
-Thayer
1 Like
What if he just marked the invasives to be removed? Everything still gets documented, and he leaves behind easy to find tags for later removal.
I’ve often used Avenza, a phone app, for things like this. You’ll just have to prepare a geolocated map of the area to upload into it as a PDF, but from there it’s pretty straightforward to put down marks on the map as you’re walking around, and you can either just make the name of each point the species or add another data field for that. It also allows you to take a photo as you create the marker or add one later.
It’s still a GPS based method so the things someone said above will apply regarding accuracy and precision if you need it, but otherwise I highly recommend checking out Avenza and its features. It’s also free, unless you need to have more than 3 maps active at one time.
The State of Wisconsin Department of Natural Resources, Division of Forestry offers the following list of free digital data collection applications for Tree Inventories - which essentially sounds like what you are looking for. Check them out and see if any of them might work for you (iNaturalist is an option listed …)
- Mark
1 Like
I’m running a similar project at the moment and trying to map about 3hec of ancient woodland with individual trees numbered so that their condition can be assessed year to year. For mapping I set out a 100m transect line through the woods, starting from a fixed point, with the line running north-south as exactly as I could get it. I then mapped each tree with a distance perpendicular to the line using a laser tape measure. By covering 50m either side of the line I end up covering a 100m quadrat.
My main worry is repeatability. Someone has to be able to easily replicate my results and find the correct tree on the ground. I think tree tagging is almost unavoidable. You only need to tag a few key trees and provide GPS points for the start and end of the transect.
I’m drawing my map up on AutoCAD as QGIS is still giving me a headache. I’ve got 20 years experience using CAD but GIS is still too frustrating to use for me. As someone else suggested, Google My Maps is a useful tool for survey notes and basic maps. I use it a lot for surveys and it’s fine as long as you have landmarks to refer to on the ground.
1 Like
Before you get too into this project using iNaturalist, I suggest you get a better idea of how accurate and precise your GPS readings will be at your site. I would have at least 3 people go out and map the same 5 trees but at different times of day or different days. Then take a look at whether the data points are congruent enough to be useful. The overall quality of GPS coverage at the site, the way it varies through the day, and even the receiver quality of different phones can all make a reading pretty far off.
5 Likes