Next question: What about insects eating plants in parks or gardens? Up to now I haven’t bothered to upload the plant, because it’s casual. If the web page only uses RG observations, there will be a lot of interactions lost. Not only “pests”, but also something like those Monarch caterpillars: https://www.inaturalist.org/observations/99402526 in a park in Benalmádena.
2 Likes
I understand that this is just something that’s meant to interact with this particular iNaturalist project and may never go beyond that, so here are just some random thoughts.
Requiring that both eater and food are represented by iNaturalist observations (and RG at that) is in my opinion an enormous hurdle that doesn’t have to be. There are fields dealing with food interactions (predator taxon, prey taxon, feeding on, and many others) that require that a valid taxon is specified, but not that an observation exists. One could survey such fields and vastly expand the scope of useful observations.
Sure, there are bound to be more misidentifications than with the RG observation and one would have to think about this in terms of probability and not so much as the one-to-one correspondence in which the results are currently being displayed.
Along these lines, as a user I would be less interested in “which specific observation illustrates” any given interaction, although I can imagine that one may be interested. What I’d hope to get out of this could be a presentation of results like cercis is being eaten by cervidae in 80 % of the observations, and by megachile in 15 %, and others in 5 %. Or something along those lines.
Sounds like a pretty massive effort though.
1 Like
All really useful thoughts, thanks for taking the time to share them.
-
Defining “eat”: I honestly don’t know the right way to draw a box around this. The simplest project would just be focused on macro-animals eating things. The most inclusive, expansive project would include face microbes “eating” humans, etc. The benefit of the former approach is that it simplifies the network and makes the relationships more intuitive. The benefit of the latter approach is that it showcases the mind-boggling interrelatedness of the natural world. Given the nascent stage of the project, I think I’m OK with not making the rules, for now. So far, most observations are of the former, which definitely qualifies as “eating” no matter how expansive or not we decide to make the project. My current leaning is to include decomposition as well, where it’s clear who the partners are in the interaction (maitake mushroom on a maple log, etc.).
-
Common names: given how many people have requested this in this thread, this will definitely be my next step. For the first version it was just easier to code by common names, and I thought would be easier for a general public that is less familiar with scientific names.
-
Including the wasp/cicada interaction: I definitely see your point and I’ve spent some time thinking it over. I was worrying that opening the door to observations where the photo isn’t direct evidence of feeding, even if it’s 90% accurate, could reduce the utility of the project to ecologists who wouldn’t want to manually filter observations in the way you describe for a macro-level study. That said, many of the observations in the database aren’t necessarily direct evidence of feeding either. A snake with a frog in its mouth didn’t necessarily eat the frog if the frog ended up escaping after the photo was taken. Hmm. Maybe the best comparison for the wasp is a photo of a cheetah with a prey mammal in its jaws that it might not end of subduing, or might not eat itself and might be bringing back to its young. If we accept the cheetah photo, I actually don’t see a reason not to accept the wasp photo. I think you convinced me. The project is now open to wasp-cicada interactions!
3 Likes
All great points, thanks for your thoughts here.
-
Requiring both partners to be IDed separately and to RG: Yes, this is absolutely a big hurdle that limits the data quantity. However, there is an existing trophic interaction database (GLOBI: https://www.globalbioticinteractions.org/) that uses iNaturalist observations in the more expansive way you describe. I think Who Eats Whom could serve as a useful complement to the efforts of GLOBI that focuses on interactions that are confirmed by the iNat community rather than just the observer, and thus, perhaps, more accurate.
-
Showing summary-level data (“Hognose snakes eat 80% amphibians, 10% mammals…” etc.). This is an excellent point and on the shot list of my next steps. I just have to figure out how to code it! My first thought is a graph at the top showing a breakdown of the diet of the selected species broken down by iNat’s “iconic taxon name” with options to expand into finer-level taxonomic categories. But I’m definitely open to suggestions regarding what would be most useful to users.
1 Like
Oh, I think insects eating plants in parks/gardens absolutely qualify. In fact, one group of people who I think could really benefit from the database is community gardeners trying to better understand local pollinator and pest interactions in their gardens. If you just upload that same photo of the monarchs and specify in a comment that the photo is for the plant then we can get a RG observation of the plant and get the observation in the database!
That was the point I was trying to make - maybe it was not clear enough. Since the plant is “captive” it won’t reach RG, but stay casual.
1 Like
Next step is definitely to add scientific names! I’m sorry the search bar is horrible. One problem is that it loads names in the search bar that don’t actually have any interactions yet and so you get a blank screen. My web design skills have yet to catch up with my enthusiasm for making this project, but I’m working on it!
1 Like
Oh I see what you mean. Dang, this thread is really productive for coming up with things I didn’t think about. I’ll have to come up with a separate coding workflow for dealing with captive/cultivated. Maybe like, “if >2/3 of people agree on the ID, then include it.”
2 Likes
Amending this in response to the below thread: on further reflection I think there is no real difference between a wasp holding a paralyzed cicada and a cheetah holding an antelope that it might bring back to its cubs. I have changed my mind and think the wasp-cicada interaction is permissible!
Thank you for reminding us that iNaturalist does not need to be all things to all people.
1 Like
Random night thought - not totally serious, but maybe something to think about. ;-)
So you have included “present tense” and “future tense” eating - what about “past tense”? Like this scat of an otter, that has eaten red swamp crayfish or this rest of a pigeon, which was eaten by a peregrine falcon.
Owl pellets with distinctive bones of rodents would also qualify…
2 Likes
So would worm castings, and, by extension, soil itself. But going there would not be useful to the premise of the project.
Oh wow, those are great observations, and a good point. I think past-tense is fine if the iNat community is able to ID the eater from the scat etc., as seems to be the case with those two! I agree with @jasonhernandez74 that worm castings is probably a step too far.
1 Like
I’ve run into a new problem. Today I have observed three species of birds feeding on kakis from one tree in a garden. Great! I really enjoyed watching them and thought: great for the “who eats whom”-project. But when I started uploading I realised I should have three different observations of the tree (the organism being eaten). Photos were taken just minutes apart. Three photos of the same organism in five minutes is not allowed by iNat-standards. I suppose this will also occur with pollinators on a flower bed (well, there you could argue that is is always a different flower). Do I have to go back tomorrow and hope to find another one of the birds?
Three observations in five minutes isn’t recommended, but it isn’t outright forbidden either. I would suggest making the observations, and being explicit about why you are doing so and linking to the others in comments. I think this would be ok.
Depending on how the observation fields work (I think this is why you would make three, to have the observation fields different for each?), you could also make the observation ID for the eating organisms (not the eaten) in which case, three different birds observations in the same tree is no problem.
I don’t immediately see a problem with uploading three bird observations but just a single tree observation. I think you can link the three bird observations to the same tree observation without a problem. If that doesn’t work though let me know!
It’s no problem to link the three birds to the same tree - but I can’t link the tree back to the three different birds, because I can only put in one “partner observation”.
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.