I have several photos of mating houseflies/flesh flies and even off of damselflies. I have uploaded a separate observation for the male and the female in the pair. But I would like to know whether uploading 2 observations, one for each of them, is actually right.
3 Likes
Yes, itâs OK to upload two copies of these images. Please annotate one to male and the other to female. That may head off anyone who thinks they are duplicates.
14 Likes
This has been discussed many times before on the forum. For example:
In short: thereâs no ârightâ way to do it, as such. Many people treat the mating act itself as the subject, and so only create one observation. Other people create separate observations for each individiual shown, regardless of the interactions between them. I have even seen it suggested that there should be three observations: one for each sex, and another for the mating act itself.
Whatever you decide to do, you should always also consider clarifying your intentions via the observation description and/or observation fields (ideally, including links to any associated observations).
3 Likes
Thereâs even an observation field for linked observations. So the two observations could easily be associated with each other.
4 Likes
I make two observations. Each obs. includes a cropped photo of one of the pair, and a photo of them both. I link them each to the other using the observation fields and also in a note. (The note usually also points out that I donât know which one is which sex, but identifiers often add that annotation later.)
I make sure that the first photo in each of the obs. is different, so that identifiers seeing them next to each other on a page when theyâre identifying wonât automatically think theyâre duplicates. (In other words, I use the photo of them âin the actâ as the last photo, not the first.)
6 Likes
I would suggest adding a note that says this observation for the thing on the left or right (as the case may be). Annotating as male and female assumes that both organisms in the observation are of different sex. Iâve seen quite a few observations in which two males were fighting and annotating as male and female would make it difficult to know which observation was for which organism. If you are at all uncertain, do not annotate as it is a tad more difficult to remove incorrect annotations.
1 Like
I agree, definitely two observations is ideal. I know with moths, when theyâre confused at a light, theyâll often pair up with moths of entirely different species/genera/families, and Iâve seen several cases where someone uploaded one observation of a âmating pairâ that included two different species. It would have been good to have one ob per individual in those cases. Iâm sure moths arenât the only ones this happens with sometimes.
6 Likes
Yeah, Iâm a big fan of the Interactions (linked) project and its array of linked interaction fields. Itâs super helpful when creating separate observations of a host and a parasite, especially for when I go back to one of the two observations later and I want to navigate to the other one. But for this use case itâs definitely useful too. The only downside is that observation fields arenât shown on the old / Android iNat app, so not everyone sees them.
1 Like
I leave this as one observation, annotate the Sex field to âCannot be determinedâ and add the observation to the Mating behaviour and Multiple Sexes projects. It would be a good idea to duplicate them and add all the same details except use different sexes, but from my experience that leads to incorrect IDs on one of the observations (sometimes) and better require there to be a âlinked observationsâ feature that forces the CID taxon on one observation to match the other especially if the same organism(s) was/were used in the observation (and overrides âopt out of CIDâ requests of observers if enough people agree the same organisms are present). Feature request for another thread if I have time.
Other people can do what they want. If mating behavior is present but all the organisms depicted are engaging in homosexual behavior itâs okay to annotate to all male or all female, so when observing donât assume the pair are different sexes if the species has a high incidence rate of homosexuality or bisexuality.
1 Like
Thanks for clarifying.
1 Like
Well, that just sent me on an interesting journey through Wikipedia! 
Yes! I had to use this to debunk a âhomosexuality is against natureâ argument a few times, so itâs pretty high in my bookmarks.
2 Likes
Definitely not the only insect species where this happens. Iâd like to point out that there is a project Interspecies Insect Mating that can be used in such cases (mating behavior/attempted mating between different species is acceptable for this project, even if they havenât managed to hook up).
Regarding the original question: I donât generally create observations for both individuals in a couple unless there is a particular reason to do so (such as a mixed-species pairing), but this is my personal preference. It is perfectly fine to do so if you choose.
I agree with suggestions above to indicate in a note which individual you are interested in (if youâre not sure about the sex, relative position can be useful â the one on top, the one on the left, etc) and cropping to highlight the individual of interest can be helpful, though not always feasible.
The appearance of two or more legitimately interesting taxa together in photos is actually incredibly common, and these interactions and associations are an important aspect of observations.
INaturalist does not currently attempt to capture this elegantly or implement the actual attachment of multiple taxa to observations, apart from âobservation fieldsâ or special manually curated projects.
Under the hood, this could be solved in the backend databases etc by including a list of 1:n putative separate taxa IDs represented in an observation record, rather than a single taxon record only.
It would involve some work to update the core iNat system to work this way, but arguably more of an unimplemented feature than an impossibility.
As for how the UI could be updated to render and handle the assignment, identification threads and display of multiple legit taxa per observation, that would be up to designersâŚ
I proposed this as a feature request recently, but was advised by @tiwane that it might have been a bit beyond a simple âfeatureâ
Doing this as a fundamental update to iNat would hugely expand the treatment and recording of species interactions, as well as likely greatly increase the number of species in general that get identified overall.
Iâm pretty sure the CV (and human observers/identifiers) already think in terms of multiple separate taxa possible per photo/observation, but the current iNat data representation does not (though it readily could, with some backend dev workâŚ)
1 Like