Some iNatters would like to filter out and UNsee Dead obs, so I do annotate them as I come across them.
https://forum.inaturalist.org/t/implement-photo-blur-on-observations-annotated-as-dead/21214 declined
https://forum.inaturalist.org/t/no-ability-to-report-disturbing-images/49843/12
I agree to a point, but as long as people watch notifications and are ready to withdraw if wrong I think even guesses on older observations can be helpful. I wouldnāt suggest it on new observations that others are highly likely to see and (hopefully) identify, but if youāre digging through older high-level Needs ID, even wild guesses can sometimes get someone looking at it who can actually ID, whereas a comment rarely seems to me to get anywhere.
Having said that, if youāre wrong and donāt withdraw, youāve made things worse by requiring three correct IDs rather than two, so I would never suggest guessing if youāre not willing to follow up.
I would also say that I avoid species-level IDs unless Iām certain because too many observers agree blindly.
If you donāt stick your neck out and try an ID where you are not sure - you deprive yourself of an unfolding conversation. The learning curve is good. I can wait ā¦ as we accumulate a row of Not That, or That, or That Either. We are not sure. And I will withdraw if / when I become the dissenting Pre-Maverick.
Here we plant people say your āseedā is looking at me!
Thatās a great idea! Iāve just spent some time doing just that.
While most of the things I encountered were plants and fungi I couldnāt identify any closer that that sometimes you do actually stumble upon observations which can easily be identified much more narrowly.
I wonder if we (the experienced IDers) should start and support a forum topic titled something like: āNew to Identifying? Ask Your Questions Here!ā Or even a project. Or a series of projects tailored to different regions. With an initial comment something like: āDo you want to help with identifying, but are afraid to start? This forum thread/project is for you - weāll help you figure out how to start.ā
Because I suspect most of us commenting on this thread have been identifying for a while, so weāre all preaching to the choir, so to speak. Iād like to hear from people who want a little mentoring getting started.
If it has been dead more than a couple of weeks, it could make the seasonality chart misleading.
One thing every person using iNaturalist data should do is check outliers. Check records that are out of the expected location or out of the expected time. I learned this long before iNaturalist but itās especially true here, where outliers can accumulate so easily.
As in this case:
https://forum.inaturalist.org/t/does-the-beetle-donacia-crassipes-occur-in-north-america/43326
All gone now.
FYI, you donāt need the ā&iconic_taxa=unknownā in the search if you have ā&identified=falseā. My unknowns url is just https://www.inaturalist.org/observations/identify?identified=false
The reason someone might be afraid to start could be
No learning can occur under such a condition. Whoever learned how to walk, or bike, or skate without falling down? Whoever mastered a language without speaking it poorly? Whoever became a chess master without losing a lot of chess matches?
The same researchers who complain about āpoor qualityā IDs would not have such a large dataset to work with if people didnāt also use that dataset to learn on. Instead of having to correct the species IDs on lots of Chironomids, they would have to find the Chironomids in the huge mass of āInsectsā that people did feel certain enough to take any further.
Yes, many of the photos are low quality and most of others have some obstacle (they are larvae, etc.) but thanks for trying - one in 15 is progress!
I agree, and some of them turn out to be interesting puzzles! Like the population of Smilax glauca in Washington State, thousands of miles from the main range.
Thatās a good point. I suppose Iāve been lucky to learn many species before iNat came along. Iāve certainly made my share of mistakes in IDing, but at least Iāve had field guides, classes, and naturalist friends to teach me off-iNat. Iāve just returned from two weeks in Mexico and as I sort through about 2600 photos for uploading, Iām frustrated by the lack of a field guide to the flora. My hats are off to all the botanical identifiers in Mexico who have learned species somehow, but not through field guides, apparently.
I donāt think ācertaintyā is a useful criterion for thinking about IDs because it is not a black and white sort of thing where you are either certain or you are not, and also because certainty (which is a subjective feeling) does not necessarily correlate with either knowledge or correctness.
I may feel uncertain about something because I am aware of my lack of experience ā and yet the ID may still turn out to be correct.
I may feel certain about something because I donāt know enough to realize that the situation is more complicated than it appears at first glance ā and thus I might end up suggesting something that is completely wrong.
I think a better way to think about whether you should make a particular ID is whether you feel that you could justify it if asked. And by ājustifyā I mean more concrete reasons than āI compared a few pictures and the CV suggestions look OK to meā. This isnāt to say that one needs to consult technical scientific literature in every case, but rather there should be some reflection about āwhy do I believe it is thisā and ācould there be other possibilities I need to considerā.
In this context, I think it is also important to recognize that iNat data only provides a very incomplete picture for many taxa; in other words, just because only a few species in a genus or family are represented in iNat observations in a particular region, it does not necessarily mean that there are no other plausible possibilities.
regarding annotations, a useful thing to keep in mind is that if an insect has wings, itās adult
(some exceptions apply - looking at you, mayflies!)
Being certain about things can be unhelpful. Iāve come across many cases of people being certain about IDs they were quite wrong about. I think a better criterion is āonly ID if you know what the distinguishing features areā.
I have recently used the Sort By Random filter and I find it to be a great balance. You chip away at the old ones, but get fresh (possibly easier) ones as well.
This project:
https://www.inaturalist.org/projects/sa-marine-mortality-events-2025
Tracking deaths due to a toxic algal bloom. Animals have to be annotated as dead to be included in the data.
I am mostly looking at plants, so I go by Date Observed Ascending. Flower obs collected together, buds and fruit in their batches. Duplicate obs, or Should be One but single pictures are scattered across many obs. All of that context is lost if Random (not for my mindset ;~))
I donāt usually speculate on what others do with the dataā¦ but phenology is first to mind, is it end of season, or early emergence? Are these dying naturally or is something causing it, like a different chemical, or improper use (I live in farm/forest area). Are the plants affected as well as insects? Are things dying from preventable causes? Are the deaths or early emergence caused by weather shifts? Is my little corner of life part of a new trend or the part of a wider trend? As landowners in Michigan, weāve been asked to inspect our trees for invasive pests and tree damage so we donāt lose the forests, so dead/dying trees need recorded. Iāve asked observers to report some Iāve IDed for them who were unaware of the reporting and the damage potential.
Roadkill and other deceased animals are great resources of many observations of flies and beetles and such. But many find the hosts disgusting, and seeing such things randomly can be disturbing.