A few weeks ago while on a hike, I encountered a fungal infection of multiple conifers that I couldn’t identify, with needles seemingly fused together into mats. I couldn’t find it in lists of common conifer infections, and the computer vision wasn’t even able to ID it as a fungus. With the help of @guidingguida, I determined that it was brown felt blight, herpotrichia juniperi/neopeckia coulteri, which grows at high elevations on conifers buried in the snow pack. After determining that, I combed through observations of ‘unknown’ and unclassified fungi in several mountainous regions and found several more examples.
Before I figured out the ID, I took hundreds of pictures of different infected conifers in several completely separated sites, basically a survey of almost every infected individual I saw from the trail. The photos are at a variety of angles, distances, and stages of infection. So far, I have posted just a few representative observations. I’m not sure of the etiquette of whether and how I should post these. If I split every infected tree into its own observation, it would probably be enough that it would kick herpotrichia juniperi over the threshold to be included in the computer vision, helping people in the future ID it and preventing them from getting stuck in ‘unknown’. However, it would also wildly skew the frequency and location of occurrence statistics, because it would represent the vast majority of the total observations of the species.
Another alternative is I could post all the pictures in just a few observations, perhaps divided by the general site and specific host ID, but not specific host plant. Or I could keep just posting a few representative observations at the sites I observe it.
Good question! I don’t think there is an agreed-upon standard here. The only thing I would say no to is putting observations from different sites together. But otherwise you could put a few from one site together, or you could do one sighting for each infected tree (but that gets tedious). It’s really up to you! If they’re the same species at the same location and on the same date, there’s nothing wrong with putting them together and making a note that it’s several trees. On the other hand though if they’re different individuals, there’s nothing wrong with making separate observations for them all.
One thing I could see being interesting in making each separate infected individual its own observation is that the geotags create somewhat of a local cross section at one point in time of the severity as a function of altitude, and because its growth is highly dependent on snow pack depth that might be interesting for tracking the effects of climate change.
iNat is not designed to accurately represent/record abundance of species across space and time, so this shouldn’t be a concern/reason for you not to post them. All 131 observations of the native snail Sauroconcha sheai are from me, and all from the same tiny reserve I’m surveying, even though they do have a (slightly) bigger range.
I may be wrong, but I think I remember reading that iNat requires not just a certain minimum number of observations but also a threshold number of distinct observers before including a species in the computer model. No doubt somebody more up-to-date on the computer model policy can confirm or otherwise.
"Lastly, in recent models, a taxon must have at least 100 verifiable observations and at least 50 with a community ID to be included in training (actually, that’s really verifiable + would-be-verifiable-if-not-captive, because we want to train on images of captive/cultivated records too). That’s quite different from the criteria for our first three training sets, which were filtered by the number of photographers. "