Passively crowdsourcing images online for measuring broad-scale fly (Diptera) floral interactions and biodiversity

While biases exist in crowdsourced data, we show that data from photographs collected through citizen science offers potentially valuable information for monitoring pollinator-flower interactions and augment our understanding of pollinator ecology in an era of global insect declines.
https://pollinationecology.org/index.php/jpe/article/view/724

2 Likes

Thanks for the great article.There is a lot of great material there - the essential role of flies as pollinators, the role of online photographs taken by citizen scientists as a source of data for the research, and the creative use of those photographs by the researchers.

I could not get the link to the article to work simply by clicking on it. Instead, I had to copy the url and paste it into the address bar of the browser. It was well worth the effort, though. :slight_smile:

This should work as a clickable link:

1 Like

Interesting that they specifically did not use iNat observations for their study, but relied instead on photo-sharing websites:

Although iNaturalist is becoming increasingly popular for biodiversity-related research, we did not use this database because it has limited functionality for labelling and searching interspecific interactions.

1 Like

I was intrigued by the justification that

we did not use this database because it has limited functionality for labelling and searching interspecific interactions.

I can only assume the authors were unaware of observation fields

3 Likes

I wondered the same thing.

However, I can also see that observation fields might not be a particularly streamlined way to get the sort of data they needed for their study – because of the plethora of different, similar fields, and because one has to be fairly familiar with iNat to figure out how to view observation fields in a way that allows one to sort by taxa. And a lot of people don’t use them at all, so this would represent only a small part of iNat’s total data set.

Also, as far as I know it is only possible to view all observations with a certain field, or observations with a certain specific value in that field, but there’s no way to sort for observations where the value in the field is (for example) any taxon within Asteraceae.

Now, I don’t know whether the methods used in the study were more efficient than extracting the relevant data from iNat observation fields.

I agree that getting data from observation fields downloaded is not immediately obvious. However, the authors weren’t looking to record the info in iNaturalist, just download photos of flies on flowers. I think they probably could have figured this out with a little searching and gotten a lot of observations suitable for their paper.

They did a lot of internet searching to get to 1275 images to use. I bet they could have gotten more images from iNat in a standardized, more efficient way by spending a small amount of time learning to use iNat. But, still ai cool idea!

On a related tangent, is there an easy way to see how many observations have values for each observation field? The observation field search is here:
https://www.inaturalist.org/observation_fields

But it doesn’t show number of observations using each field. If that info was available as a column on the search page, it could help newer users find the fields with the most data more easily. I could make that a feature request, but if the info is already easily available elsewhere, there might be no reason to.

3 Likes

yeah e.g. here are 16k observations already tagged with interactions they missed out on reviewing: https://www.inaturalist.org/observations?place_id=any&taxon_id=47822&field:interaction->visited%20flower%20of=

also related: the (hidden) “interactions” tab: https://www.inaturalist.org/taxa/55719-Eristalis-tenax?test=interactions (took a while to load for me)

4 Likes