Could iNaturalist better help observers use this as a forum for good plant ID (and other kingdoms), and less a place to just post their pretty flower pictures?

While I regularly make suggestions to observers, “for future reference, if you want a better ID on this group, you might include these things”, I wonder if there is a way, or better way, for iNaturalist to encourage people to give identifiers more of the identifying features of the plant species they post observations of, and use iNaturalist less as a forum for posting their pretty flower photos.

I was just studying the distinctions between Ceanothus velutinus, an evergreen shrub, and C. sanguineus, its deciduous cousin. There are 321 observations of C. sanguineus. The vast majority highlighted the flowers of the plants. I was looking to see what C. sanguineus looked like in twig, that is in the winter, with no leaves, or flowers. I then did a search for observations of the plant between December and March, when they would be in twig. There wasn’t a single observation of that species for that period. An excessive focus on the flowers of plant species is standard within iNaturalist plant observations. Along similar lines, I long ago, realized that far too many books on plants were “wildflower” books, and not sufficiently focused on the plant species, their identification, and their stories. It is sad to see so many people so narrowly focused on the reproductive organs of plants, rather on the plant species as a whole.

I believe that people posting observations on iNaturalist are effectively asking us identifiers to offer our identification of them, usually either as a confirmation of their ID, making it “research grade”, as a correction of their ID, potentially to refine their ID from say genus, to species, or as a first ID. As an identifier, whose forte is plants, I am constantly irritated by the overwhelming number of observations of plants that feature the flowers, and little, or nothing, else, that I am effectively being asked to identify, to species, if possible. Then far too many of the flowers only feature the face of the flower, so we can’t even see the other identifying features of the flower, such as the sepals, leafy bracts, the shape of any tube the flower might have, or the length, and character of their stalks, features that might be seen from the side or the back of the flower, or from further away. I came to call these photos, of the face of the flower, with no accompanying views that show more parts of the flower, and plant, “throat shots”, cursing that I have effectively been asked to identify another species from another “throat shot” alone. In general I’ve found that, if I’m lucky, these “throat shots” alone might allow me to get to genus, and most often no better. I came up with the theory that part of the reason people focus so much on the pretty flowers, with their beauty generally best shown from their face, is that flowers were evolved with their pretty colors, and faces, to attract pollinators to come to the faces of those flowers, and humans ended up also being drawn to those pretty flowers, and their faces. It is as if we have been tricked into giving the bulk of our attention to the face of the plant’s flowers, but without the benefits the pollinators get from being drawn to them.

I’ve tried to teach individual observers that distinguishing features of a plant may occur in the look of the whole plant, in the many parts of a plant, as well as the habitat of the plant, so that more views of a plant, featuring the whole, more different parts, and habitat, as well as notes in the “Notes” section, describing habitat, hopefully including the associated species growing with it, and possibly features we can’t see in their photos, like smells, the more likely I, and others, may be able to accurately identify it to species, or to a more specific level than we otherwise might.

Another frequent problem I have is that not enough people understand that, to varying degrees, leaves change their characters, as they go from non-flowering basal leaves, on herbaceous plants, up into flowering areas, so that the most useful, and consistent, leaf views are those of basal leaves (if they are still intact, or, if not, the lowest leaves on a stalk, that are still intact), so basal leaf views are often very important for my being able to identify which species the observation is. Leaves of woody plants also often change character as they go from non-flowering areas, into flowering areas.

That said, there maybe a limit to how many photos an identifier will want to look at, so, after some limited number, more photos aren’t always better. One way to both conserve the number of photos, and often to get more useful views, showing both perspective, and more parts per shot, is 45 degree angles away from straight down, or away from straight at, something, like 45 degrees away from straight into the face of a flower, or 45 degrees away from straight down at a whole plant, showing both front, and side, or top, and side.

I’d like to think iNaturalist could improve their job of helping observers make observations that better focus on including more identifying features.

9 Likes

There’re many topics about this, but most of such observations come from new users, you can’t do anything with how they initially look at plants, and most people have no idea about plant diversity too, for many it’s also enough to get family id. iNat is a place to connect with nature and even though notes, etc. are good you can’t force many people into it. Just get with an idea people don’t know plants and don’t know anything about leaves or even flowers.
To ease your way you can create a journal post that describes the best way of making plant observations, so you can just link it ithout spending time describing the same thing over and over. (e.g. https://www.inaturalist.org/projects/flora-of-russia/journal/37806-kak-snimat-chto-snimat-uchimsya-u-klassikov)

11 Likes

I think your original question is one of those false dichotomies – iNat is a community science platform, with all the attendant diversity of desires and priorities entailed in that description.

Having spent a lot of time teaching people to identify mushrooms in person, and simultaneously seeing many of them make the jump to iNat my advice is this:

  1. Don’t worry too much about the observations you can’t make any headway on. There will always be low-quality and junk observations and that is simply a sign that the community science platform is working. The barrier to contribution has been sufficiently lowered such that we have a chance of approximating good spatial and temporal coverage of biodiversity. Platforms that make excessive demands of ‘data quality’ from their users lose that, and in my opinion are seriously hampered by it.

  2. The best way to solicit better observations is to do direct community education (make YouTube videos, iNat Journal posts, Facebook posts, Instagram posts), and then simply … wait. I have found that naturalist communities have an amazing tendency to getter better over time, especially learning by example (although it may be slow). The quality of observations has really risen among the mushroom community as people see what other folks are doing and attempt to imitate high quality observations in their own practice.

18 Likes

I share your frustrations and often try to get photos of multiple parts or angles of a plant or other organism for ID purposes, but my sense is that almost nobody looks beyond the first photo.

I then have the dilemma of which photo to put as the primary one. Do I choose the one that shows all/most of the plant, with leaves and stem and flower etc… or do I choose the most striking one that shows only the flower full-frame? My strong sense is that I get more IDs/comments/interest more quickly on observations where I choose an eye-catching colourful first photo.

At the end of the day, this is a site for untrained people offering their time and effort voluntarily, so anything that makes the process of posting or identifying less thrilling and/or more laborious will reduce the amount of interaction, unfortunately.

7 Likes

The short answer is neither. iNaturalist’s principle (and very important) function is to promote learning about biodiversity and the things that are part of it. The main process for achieving those ends on iNaturalist is collection and identification of images. Learning which images work for identification is part of the process.

If you see an observation of a flower that needs supporting images of leaf or stem or other features to make a proper ID it is best to leave a comment to that effect. Next time the observer will know.

The accumulation of good identifications is an objective, and important, but it is subsidiary to the learning objective. Achieving both outcomes involves tradeoffs and pretty flower pictures that can’t be identified to species are one of them.

8 Likes

Initially my post put the question about whether iNaturalist could do better at the end. I have now put it at the beginning. I regularly suggest to individuals how to improve their observations, so I could better identify them. I then edited the title to clarify that question.

3 Likes

I had never tried using journal posts to link my repeated suggestions to people on how they might improve their observations for better ID’s, without writing them over, and over. So that is a good suggestion! I still wonder if there is a way for iNaturalist to better encourage their users to offer better quality observations for better quality ID’s

I am not asking that iNaturalist make any additional demands of users for higher quality observations, I am asking if iNaturalist might be able to do a better job in encouraging better quality observations for better species ID.

Your link looks nice. Do you know if there is a way I can translate it into English?

You can use automatic translation or if you want I can translate it, but not today (will be on a train soon).

3 Likes

I’m not asking you to translate it. I just didn’t see how to apply an automatic translation to it.

Depending on browser you’re using, in chrome right click on the page and choose “translate to x” where it uses the default language you’re using.

2 Likes

Thank you! I found the translator, and translated it!

1 Like

Thanks for taking time to identify iNat observations. I wouldn’t worry too much about getting something to species if it’s not possible. Honestly some of the best things to teach a budding naturalist are what’s needed to make an ID at any level, and that it’s not always possible to get to species, which is OK and even fundamental, IMO. Doing what you can to help and then moving on is fine.

I mostly identify California snakes, and many people don’t get the close-up head shots one needs to differentiate garter snakes here, either because the snake slithers away, they’re scared, or they don’t know they need that photo. But if I can tell them what to look for next time, that the evidence in their observation can only be used to confidently get to genus, and that that they did indeed see a garter snake, I like to think I’ve helped them understand what they saw and how to improve their experience in the future. Getting to research grade is great, but so is just teaching and helping people on their journey to better understand and appreciate the life around them. A journey we’re all on, since I think most of us start observing taxa outside of our comfort zone once we start iNatting. If I can tell someone how to differentiate a bug from a beetle or that the snake they saw isn’t dangerous, I think that’s a net good.

Eventually it would be cool if, once the computer vision IDs something as a plant or a fungus, we could show a pop-up to say “Try to get a shot of the side of the flower” or “Make sure to photograph the gills” or “why are you that close to a cobra?” but I don’t think we’ll get there for a while. I think some people will take the time to take those extra shots, for others it will be too big of an ask.

Until then, we have the community approach, which does take time and effort but like @fffffffff and @leptonia said, can be effective although it might take time. I recently profiled nathantaylor, who’s created a ton of amazing resources for people to use when photographing or identifying sandmats, and I think has helped create a culture of looking for and making better observations of those taxa. Obviously we have our own levels of how much we can or want to engage in this way, but I iNat has shown that it can be done.

8 Likes

I find it very rewarding to leave comments saying that it’s in genus or family X, and next time take photos that show these parts of the plant, to get it to species. I get a surprising number of people thanking me for the feedback. I’ll check back in on them a month or so later and indeed, their photos are getting better for ID purposes.

I try to focus on meeting people where they’re at, and enjoy watching/facilitating their enjoyment.

Which is not to negate your focus on feedback that doesn’t depend on identifiers having the time/interest to do this. Just my perspective on growing our community.

10 Likes

I have now read that linked article on what views to include for best identification of different groups, at least in Russia. ( https://www.inaturalist.org/projects/flora-of-russia/journal/37806-kak-snimat-chto-snimat-uchimsya-u-klassikov ) It seemed like a useful series of suggestions for the different groups covered, for at least that country. I wonder if there is a way for iNaturalist to offer more links to observers to explain general ways they might improve the quality of observations for ID’s of more general groups, or more specific groups, (ideally in higher quality of English, or other languages, than offered by automatic translators)

1 Like

Great suggestion! I started writing journals for similar Penstemon species to highlight the features that make a species ID possible. I have a long way to go with the other posts, but this one is complete. https://www.inaturalist.org/journal/egordon88/55949-penstemon-similar-species-part-7-large-round-flowers

2 Likes

I too have been pleased to see observers incorporate my suggestions on how, next time, to get a better ID of that species, or group, incorporate them in future observations, allowing me to identify them better!

And look at all the responses! In fact I’d say growing community is essential for getting better id results, this project is not loading now as it’s too big, but look up the posts it has, making regular news and informing users, including tips and “how to” is very cool and also it allows non-botanists like me to get involved and connected with professionals, I was invited on iNat because of this project and I got interested in plants in 2015 after like second summer practice had shown me some new things, iNat helped me with many taxa and now I at least remember what to photograph in plants. So, maybe everything you see is that gradual process going on?.)

2 Likes

Since I started using the Newcomb’s wildflower guide, I took a minimum of three photos when possible. Occasionally, I still find out I missed an important angle . Other times when I could not get all the shots I needed was due to unexpected blurry shots, rain and biting bugs.

5 Likes