I see several threads discussing how to upload observations of groups of organisms, but can’t find an answer to a certain sub-category of that: How to handle photo sessions of groups of animals in which the individuals frequently mix? I read on old forum threads that the ideal for data quality is 1 observation per individual, or a single observation of the whole group as a reasonable alternative. How about opportunistically targeting individuals within a group over a period of time?
To be clear, I’m only talking about reasonable quality, detailed photos that are distinctly identifiable to species, and/or short bursts showing distinct behaviors. I’m not trying to shoehorn in all the trash on my SD card!
Current example: I recently spent about 30 minutes photographing a flock of about 300 waterfowl, nearly all 1 species except for singles of a couple others. The flock was restricted to a small (0.25 hA or so) patch of open water and was a closed population while I was there. Since the flock was quite active it was impossible to track individuals of the main species for long. During the 30 minute session I took occasional wide angle photos of portions of the flock, but mostly took narrower shots of individuals opportunistically. I now have many photos I could upload, but I’m not sure how best to divide them up since I can’t tell how accurately my sampling depicted the range of individuals. My best guess is to make a separate observation each time there’s a notable shift in the camera’s field of view or time log, but that seems a little extraneous and feels string-y.
I personally usually just make one obs per species per day especially if I stay in a single area as I don’t want to waste identifiers time by uploading 300 low resolution crops of mallards. I don’t really think iNat is a great database for measuring abundance, but I do try to include photos of full flocks sometimes in my observations (if they are a mixed species flock I sometiems have the first photo be a crop to the individual I am interested in and then include the flock photo as a secondary one). If you want to record useful abundance data I think a platform like eBird is going to be much better suited for that.
The only benefit of this is boosting your observation numbers. It does nothing for count as iNat is not geared for that.
When doing identifications, I get frustrated when I have to ID the same species at the same time/location multiple times.
My take on this point is:
If you have changed location from one patch of open water to another in a different area, then by all means, upload a 2nd observation of that species as it is a totally new site.
0.25Ha is basically 50m50m. A very small area to justify multiple observations.
Different days, weeks or months, by all means.
I would justify it if it was a male, and female separate observation, or adult and juvenile. Because the annotations only allow you to stipulate one of each.
The concurrence across all the threads I’ve found is that multiple similar observations are absolutely fine for the data perspective. I’m trying to work out what’s best for the data quality. Personal statistics and feelings are not important.
Sorry, I’m not sure I understand what you mean. It sounds like you’re saying personal preferences of identifiers equate to data quality, but that wouldn’t really make sense. It’s understandable to a point that observations of taxa that are difficult or have few capable identifiers may get skipped for inconvenience, but that’s not at all the case with birds. Also, wouldn’t it be poor practice for an identifier to mark an observation they can’t be bothered with as Reviewed?
I think that makes perfect sense! Identifiers are human and have a finite amount of time to spend identifying. Any hour spend combing through 300 mallard photos at the same lake on the same day is an hour not spending identifying other obserbations, and given that I think iNaturalist is basically never going to be used for abundance data for birds it’s not clear what is gained by this approach.
no! identifiers are self-motivated volunteers with no obligations to identify any particular observations whatsoever. there’s nothing stopping anyone from only IDing aesthetically pleasing photos their friends take of penguins if that’s what they wish to do. Someone may choose to spend their free time clicking yes that’s a mallard over and over again but I don’t think anyone owes it to you.
if you mouseover, iNat explains that ‘Reviewed’ applies only to the person who clicked that. It has zero effect on anyone else - except the observer who will miss out on that identifier’s ID, and the ripples of that missing ID. It is not ‘cannot be bothered’ - no identifier is under any obligation to ID ALL The Birds.
I used to upload them all separately and it was such a pain to manage. For anyone else reading: if you’re on the web uploader, you can just select multiple photos and drag them on top of each other to merge them into one observation instantly. Saves so much time when dealing with a big flock!
Adding multiple observations of the same species at the same location on the same date does nothing to improve data quality. I’m both an identifier and a data user. I do everything I can to filter out what I consider to be duplicates. They’re a huge waste of my time throughout the data pipeline, from identification to end use.
I’m considering creating my own tools for filtering/consolidating observation data in advance of the identification stage. If I can come up with a way to do it such that it saves me time, I may cease to do identifications on iNat entirely - I’ll just do everything in my own system. If that comes to pass, there will be one less active identifier on iNat.
At a minimum, I filter out observers who post a lot of duplicates. So they’re already not benefiting from my ID efforts. I also filter out folks who are sloppy about specifying accurate locations.
It’s no fun drinking from a fire hose, and my time is not theirs to waste.
‘poor practice for an identifier’ ? in my book is - you have 7.6K obs. Presuming you always upload with your own ID to start with you need 1.25 as many IDs from others = 9.5 say 10K IDs are what you ‘owe’ back to other iNatters. You have made 2.1K IDs for others. Poor practice … If you usually upload without an ID then 7.6K obs, allowing for the ones that need discussion = 19K which may take you some time. Will be much appreciated. iNat has 503K Unknowns waiting for you. And then there are the Needs ID 109 million.
I try not to upload too many of the same species from a single location, unless…
Sometimes I upload the same species or even the same plants at different times of the year to create a record of how they look at different times of the year.
The species can vary in how they look (some plants that are the same species can sometimes look different when just casually observing them and I like to show the different forms in a single location).
Or, like recently, I uploaded two observations of birds from the same location that I thought were different birds until I went to ID them and found out they were the same species. I’m not great with bird IDs. This actually happened two times recently. This has happened with bees, too. I think they are different only to learn they are the same when I upload them and receive an ID.
Or sometimes it is just a species I love and then I might get a little carried away with sharing several different plants from a larger location…especially if in different stages of growth.
I don’t think this matters (i.e., not mandatory), but I do ID way more than I upload on iNaturalist.
I’d suggest making one or two observations from a flock, like maybe one of each sex if you have those photos, or one of an adult and one of a juvenile, or maybe a photo that documents interesting behavior.
Yes, eBird is great for recording abundance. iNat’s not made for it.
That’s what I’ve been aiming for, although it’s still several uploads due to 4 age/sex classes and a few behaviors. I’ve also been separately uploading individuals I’m sure are different by plumage. Is that still reasonable?
I think it’s fine to make several uploads. The main reason I prefer seeing this as an identifier is that I run into cases all the time where someone mistakes several species in a group as being “all the same species”. If all their photos are in one observation, it’s impossible to make a record of all the different species present. If they upload them separately, you can make different IDs on different observations and get all the species represented.
I realize I work with moths rather than birds though, and while the odds of the group of “little moths on my porch light” representing 5 extremely similar species is high, the odds of 40 mallards in a flock secretly being 5 different cryptic species is much lower. But personally I’d much rather have to click “agree” on 20 photos of the same species from the same person at the same time and place than see one observation with a hodgepodge of different species that the observer mistook as being the same. I know I’m not great with plants, and I’ve had times where I upload 20 of what I believe to be the same plant species on the same walk, only to find out later from an expert that I had some other species mixed in there that I was unaware of.
So in short, 20 photos of the same species from the same time and place might not seem useful, but I’ve made enough misidentifications in my life that I’m never surprised when photo # 17 out of 20 turns out to actually be something else.