Hello, I joined iNat because I was told I would get help IDing things I find in my backyard. (Hasn’t exactly worked out that way, but I am not complaining.) Now that I am on here, I feel like getting “research grade” on my observations is like scoring points in a video game! I mostly post observations of birds and insects/arachnids/mollusks, etc. in my backyard. I was hoping to learn species names, but what I have learned is that many small forms of life in my backyard are really difficult to ID to the species. So, hoping to keep my motivation up, I am asking you knowledgable folks here on the forum - What things are reasonably easy to ID to the species level? I will probably still post everything, but I was hoping to have this be a learning process.
It depends a lot on what local community you have around you, birds, mammals, good photos of fish, large insects, most plants if you photograph them correctly are among easiest taxa to id. If all you have is a smartphone, then plants the easiest, if you have camera for birds, birds are the best ided group on the site.
You can check topics like these:https://forum.inaturalist.org/t/observing-identifying-wildlife-wiki/15332
If I didn’t miscount you got species level IDs on 26 of your 35 observations. That’s a great ratio, if you ask me and it might even go up over winter when IDers have more time, so no reason to be disappointed.
Your question is still valid and interesting of course.
Many common species that you’re likely to run across in your backyard are fairly easy to ID - IF you can take good photos of all the right things. For example, Eastern White Pine is a very common plant where I live (New England in the USA). If I post just a photo of the trunk of a white pine, it very likely won’t get confirmed by somebody else. The same is usually true if I post a photo of the whole tree from a distance, even though white pines have a somewhat distinctive silhouette.
But if I take a reasonably clear photo of the number of needles in a bundle (five needles per clump for Eastern White Pine), I can be reasonably sure that an identifier will agree with me and the observation will make it to Research Grade. Just to make sure, I often include a photo of a cone or a branch as well.
So what I do is take good photos of species I already know and post those, because I’m interested in documenting what species are where. Those observations usually make it to Research Grade. I also often take photos of species I don’t know, because I want to go home, consult my field guides, and try to learn the new-to-me species. Those sorts of observations often don’t make it to Research Grade, because I don’t know enough to take photos of, say, the hairs or lack of hairs on the stem of a plant.
I think there are two things you can do to get more Research Grade observations:
-
Learn how to identify species before you post observations, so you know what you need to photograph. You can learn from good old-fashioned field guide books, or online ID resources, or classes, or going out with people who know more than you do, or poking around in iNaturalist.
-
When you post something you don’t know, make a comment to the effect that you’re hoping to learn how to identify this species. With luck, an identifier will come along and comment back something to the effect of “you can’t tell this moth from that moth without dissecting the genitalia; best to leave this observation at genus” or “a photo of the calyx of the flower helps a lot to pin this genus down to species.” Of course, then you might need to go learn the meaning of the word calyx!
I hope this helps. I’m going to go see if I can find your observations in iNat and make some helpful-I-hope comments.
And I’m back - you only have eight observations that needs IDs! That’s not bad at all. Plus, they are all arthropods - bees, flies, bugs, spiders - and arthropods are both somewhat hard to ID and there aren’t that many people who know how to ID them. I certainly don’t know enough to help out - sorry! - but I will say that your photos are in general really pretty good. Maybe an expert would need close-up photos of the arthropods to be able to ID them down to genus or species? I really don’t know.
Agree with all the other comments, and just have one additional point to make.
There can be plenty of value in the harder to ID observations (even if they can’t be IDed to species). They still help you learn for one thing (IDing future observations, even to family, can really help IDers find your observations). And sometimes things that haven’t been ided to species for years will get IDed when the right person comes along! If a species is rarer or harder to ID, then an observation can help it get into the CV model, which can be very useful too.
I get that, as I am also quite well triggered by gamifying things. Especially in the beginning it might be more motivating to have lots of feedback and many RG observations (although I really agree with what has been said… you have a good ratio going on, so no need to be disappointed). To increase your RG-boost, you could try to have a look, what species are common in your area and try to specifically search and photograph those. Those will got to RG quite easily I guess. Otherwise, many spiders (especially the ones without net) are often difficult to ID to species. Butterflies are usually very well known and as has been said, of course birds. It´s amazing what the birdies here can do with even the blurriest picture! There is also quite a few beetle enthusiast, but you often need good pictures from different angles (e.g. from strait above, some details…).
Maybe at some point your source of motivation might shift, e.g. getting a kick out of collecting as many different species as possible (overall or of a certain type), getting rare or so far undocumented species in your pool and so on… this usually requires photographing a lot of unknowns and one has to be prepared that many will stay above species level and some might not be looked at by someone for many months… but when then something is IDed after a year or so it is all the more exciting :-)
In my experience, birds and common plants are the easiest to get species IDs on. More than 2/3 of my bird observations are currently Research Grade (actually it’s 1455/1818 but that’s a confusing fraction) and my total count is hovering just below 2/3. For plants, most of them are fairly easy to ID if you photograph the right characters. Leaves, whole plant, and flowers are a good start, but some species need closeups of the stem or basal rosettes for a species ID to be possible. However, it often takes a while for my plant observations to get species IDs because of the shortage of specialized identifiers. As far as arthropods, they’re often difficult to ID beyond genus even with good photos. Many spiders can’t be ID’d to species without microscopic examination of the genitalia, and I’m sure a lot of insects are similar. Some of the easiest insects to ID are butterflies, especially if you can get both dorsal and ventral (top and bottom of wings) photos. Good luck!
I think the easiest species to identify tend to be the ones with a complicated colour pattern, as it is less likely there will be two species with the same pattern. For example, there are lots of pale brown warblers that are difficult to separate, but the finches tend to have strikingly original colours in this country (Britain).
It isn’t a perfect rule. Some colourful species are quite variable, such as ladybird beetles and some moths, and species id may depend more on tiny morphological differences than on pattern. And some dull coloured species are nevertheless easy if they are taxonomically isolated. For example there is only one tadpole shrimp (Notostraca) in Britain so you can recognise it without considering colour.
I disagree that common species are easier to identify. A common species may have a dozen close relatives that can’t be separated from it without a microscope. For example, bluebottle flies in Britain: two common species that are somewhat difficult to separate, plus a few rarities. Dandelions (Taraxacum) are an extreme example of the common being difficult.
Most of the latter three are often difficult to identify. Often they need dissection or specific features to be present. I mostly do Noctuid moths, and some are very difficult. One other factor is the number of unidentified observations. Canadian Nocuidae have many thousand of observations that need to be identified or confirmed. Identifiers are often few and far between!
A couple of things about learning. Practice, practice, practice. There lots of online resources (depending on where you live) and do your best to refine your identifications. It will take time, and you may be wrong, but it’s really the only way to learn a taxon. Books can help for general features in invertebrates - learning how to differentiate groups of flies is a valuable starting point, as is knowing what features make a fly a fly! Many flies can look like bees, for example. Browsing field guides is a great way to learn stuff. Oh, and ask questions (write down pertinent information - I’ve learned that the hard way).
Also, keep in mind that all your observations are helping to document local biodiversity. Species ID might not matter as much as knowing an invertebrate group exists in your region. Good luck, and feel free to contact me. I’m always willing to help, but have a somewhat narrow focus.
Thanks! I know I have lots of room to improve!
Thanks very much for this comment! What is the CV model>
Thanks so much for your comments and encouragement!
Thanks for that advice! I will try it.
Thanks for your suggestions and encouragement!
Thanks for your comments and advice!
Thanks so much for your comments and offer of help! I’m not afraid of IDing things wrong. It’s just not very emotionally satisfying when I am not sure of an ID and no one else comments on it. I don’t feel like I am learning anything that way. Thanks again!
Computer vision is undergoing learning cycles where new photos are added to it so it can remember new taxa, each completed cycle is a new model.
The CV (Computer Vision) is what generates ID suggestions when you create an observation. A species will only be included in the training set if it has at least 100 RG observations (I think; I can’t remember the exact requirement), so observing uncommon or little-known species could allow those species to be included in the next CV model and appear as suggestions. Of course, that would also require identifiers to be able to confirm the species so it becomes RG.
A few more notes to add to all the excellent advice above:
-
Because on iNat there is so much interest in wild spaces and native species, critters in backyards are often overlooked. This is a cultural issue and not an expression of the value of your observations. The best way to use iNat is to find the best ways to ask questions. Find top identifiers in your region and reach out to them. Browse your library’s field guide section and find the best ones for you. Use iNat’s Explore feature and search taxa that are similar to the ones you are curious about to see what other people post, how they ID, what pictures they take, etc. Since focusing across taxa in a small area requires you to be a generalist, maybe pick one group to focus on and learn that first. It’s much more satisfying to get competent in a field than forever fumble in all of them.
-
Identification takes time. Last I checked, the average Research Grade/Needs ID ratio is about 50%. This seems to be true no matter how you slice it-by user, region, whatever, it’s usually close to 50/50. This is a really curious phenomenon. Use iNat to record your observations, follow your joy in what you want to learn about and ID yourself, and leave the rest alone. People will eventually get to it.
*Recognize your wins. Really study the species that you can identify, and you will start seeing it all over the place. Learn lookalikes and you will constantly be scanning for the things you know. Eventually, you will start picking out the things you don’t know from the things you do, and those become the next subjects to learn!
I would agree that it’s a popular idea outside of iNat, but actually most data comes from spaces close to people’s homes and actual wild places (or close to that) get much less attention, so if there’s an ability - travel there while something is still there, in ten years it can be someone’s house, there could be easy to id plants that nobody have seen around yet.
Thanks for the explanation!
Thanks for the info!
Thanks so much for this advice! Everyone is so helpful here. I had no idea!
Thanks! Always so great to hear different points of view.
Thanks for recognizing that. I am not fond of killing things.
Dandelions are only as difficult as you make them. When every cultivar of bananas becomes a microspecies with its own binomial, and likewise every genet of breadfruit, then we can discuss dandelion diversity.
What do the San Francisco Bay Area, Puget-Willamette Trough, and Vancouver Island have in common? Two things:
- They all have a highly diverse native flora
- Most iNat users there upload the same twelve or so taxa of invasive weeds.