Many of the replies mention using annotations on the “front end” - ie. for filtering observations they want to look at. I will provide a concrete example of a life stage annotation being used at the “back end” (where data is being extracted from iNat).
For an annotation to be useful at the back end, it becomes a bit of a chicken/egg scenario. They’re only useful if they are filled in consistently. To my mind, if a data user knows (or suspects) that the annotations are not being set consistently, then there isn’t much point in trying to use them.
As someone who “harvests” data from iNat on an annual basis (for inclusion in the Ontario Butterfly Atlas Database), the “life stage” annotation is important. But in order to rely upon it, I have to ensure that it is filled in consistently. More people are setting this annotation now than in the past, but there are still many instances where it is not set. Because most Butterfly observations are of adults, and I don’t want to have to set the annotation to “adult” on all of them (there are simply not enough hours in the day), I have adopted the convention that any butterfly observation that doesn’t have the life-stage annotation set is defaulted to “adult” when it goes into my database. I try to set the life stage on all egg/larva/pupa observations as I come across them, but at the end of the year, before I download the iNat observations for that year, I go through all observations for each life stage value and ensure that it is set correctly, and I check that there are no “egg/larva/pupae” observations without an life-stage annotation. In cases where someone has set the life stage incorrectly, I add a the “Insect Life Stage” field which my software will use to override the life-stage annotation.
Yeah, it’s a lot of work, but like I said, if you are going to make use of something like an annotation, you have to make sure it’s set correctly. My life would be easier if more people set the life stage annotation, but I’d still have to visually scan hundreds of pages of thumbnails to make sure they are all set correctly.
Perhaps there is someone else out there using the same observations who doesn’t want to follow my convention of defaulting observations without a life stage annotation to adult. They are welcome to go through the 10’s of thousands of observations without a life stage annotation and add one.
In theory, I could make use of some of the other annotations. The “Sex” annotation would be one candidate. For many butterfly species, determining the sex is fairly straightforward (but there are many exceptions where it is near impossible to tell from the typical iNat photo). But unlike the life stage, knowing the sex is not critical, and it can be omitted where it is indeterminate. From a “work flow” point of view, taking this annotation into account might turn into a nightmare, since every adult has a sex, and there’s no “default” one can adopt if it is not set. I suppose one could do a year end audit of all observations that have the Sex annotation set, and visually scan all the thumbnails for any errors. If only a handful of observations have it set, that might not be too much work. One could then scan all the observations that don’t have the annotation set, and set it for any that can be easily determined from the photo, but with ~50k observations to review every year, that’s a daunting task. Difficult to justify for a non-critical piece of information.
