💾 Archived View for magda.cities.yesterweb.org › gemlog › 2024-06-19.gmi captured on 2024-06-20 at 11:55:28. Gemini links have been rewritten to link to archived content
-=-=-=-=-=-=-
Wait a minute – "against citizen science"? Some may consider this odd, others outright contradictory, as my lousy observations technically fall under the umbrella of "amateur science". My observation's "Resources & Notes" page, which easily may fly under the radar, actually states that I do not contribute to any citizen science project but Flora Incognita (and largel only because this app automatically transmits my scans to Ilmenau and Jena upon confirmation).
There is no denial on my part that so-called "citizen science" not only sounds intruging but also may provide insights professionalized science is unable to make (or even misleads intentionally, remember the replication crisis and phenomenona such as "p-hacking"). The way Flora Incognita approaches plant identification and tracking even used to be a solid compromise between plant data and user interaction; all scanned plants are sent to a public database while users remain fully anonymous to third parties, and FI, when users decide to create a profile, only keeps Email addresses and an auto-generated name of its users.
I regarded FI as the perfect app concept-wise because it still lacked quite a bunch of wild plants at the time I downloaded and started to use it, though this limitation was justified and thus acceptable. FI started off with a rather small set of plants, nearly all of those based on scans made in the Ilmenau and Jena areas. Some information on specific plants and virtually all information on pollinators were – and sadly still are – superficial at best, non-existent or false at worst. still, my opinion did not drastically change until FI introduced "badges" and justified this kind of "gamification" by running out of funds if FI doesn't "grow" (in terms of users and features) – Patrick Mäder mentioned this reason in a developer's comment in reponse to one user review on Apple's App Store that criticized the radically new UI, so good luck finding it.
But if funds are a concern, why doesn't he make his app "open source"? The German government recently announced to sponsor The Gnome Project, after all, and this project certainly doesn't lack sponsorship (in fact it's backed by Red Hat and, by extension, IBM, among other "big players"). Why badges for completed challenges and a sudden inclusion of non-European wild species and domesticated/breeded plants? The Max Planck Institute for Biogeochemistry (Jena) appears to fully embrace gamification without any sort of critical analysis whatsoever, despite a considerable amount of international literature highlighting significant drawbacks caused by this method in other areas of human lives¹. So far, studies analyzing "gamified" citizen science projects and their short- and longterm effects on users are virtually non-existent; few studies deal with the quality of data collected by general citizen science projects that do not rely on gamification.
At least to me, it appears that the vast majority of users eventually move on and stop contributing, regardless of gamification.
Visit tagfalter-thueringen.de, switch to "Statistik" ("statistics"), maybe even do a search query and select a bunch of very common species, and you'll quickly come across a drastic drop in 2023 in all graphs. Since this project does not provide anything one would find in a typical research paper, including a methods section, elaborations and possible data limitations, a layman accessing this site may assume that 2023 was a catastrophic year for butterflies.
Fortunately, I can confirm that this is not the case, however it did take me several manual search queries to discover that the amount of contributors saw a drastic decline. Individual reasons of each former contributor's sudden decision to turn their backs on TT are unknown to me; only one former contributor switched to Lepiforum and now uses this project's forums exclusively. But even the contributions of Gerd Kuna, the sole person managing the database and one of the few contributors providing new data on a regular basis, has begun to submit less observations, now excluding any area outside of southern Thuringia (perhaps justified, he isn't getting any younger and may not be as mobile as he used to be anymore).
As far as I can tell, no contributor is below the age of 30 and this project is largely unknown even among the lepidoptera crowd. The available data largely covers very specific areas in southern Thuringia, surprisingly almost all areas harbor a high amount of species not observable across the rest of the state (unrepresentative "hotspots"). Data from other areas are nearly as sparse as randomly conducted investigations by state-sponsored scientists and orgs. Despite hoping that some younger people may discover it, contribute and introduce some necessary changes like the abolishment of mandatory "real names", TT is a project that started off and will end in obscurity without most people ever learning about its existence.
In contrast, the US-based iNaturalist is not exclusive to butterflies, provides a smartphone/tablet application and a "sister app" called "Seek", which relies on gamification. Even one of my former schoolmates used it for a while, with an emphasis on "for a while". She provided a single observation of a praying mantis on 20 September, 2020 and last used iNaturalist on 22 June, 2021, never following anybody and never assisting other users at identification. Abusing iNat's search abilities and entering various locations in my district yielded disappointing results.
Sticking with the location of my former schoolmate, one recent observation is from a new user that joined in May, 2024; the second result is from another user active since 2020, lists over 300+ observations and nearly 500 identifications – he isn't a "local" by any means and his observations are scattered across Central Europe and southern France. Overall, this chosen location lists ten observations from ten observers spanning from May, 2018 to May, 2024.
To be fair, this location is a small village next to an Autobahn that is dominated by conventional agriculture. So what about locations closer to, let's say, Jena?
I chose the village Isserstedt situated just seven kilometers northwest to central Jena. 20 observations from 10 observers within just five years overall. Nine observations are up to a month old and originate from a single user.
What about Jena itself? 14,652 observations, 472 observers spanning from September 2012 to 18 June, 2024. While this may be far more data than for Isserstedt, Jena is quite big and a quite popular city among scientists. One user alone provided nearly all observations on 18 June and his profile reveals that, while there are a lot of observations in and around Jena, the second biggest "circle" on the map highlights Protugal, the third biggest the Dresden area. Most other observations are scattered across Europe and extend to places such as Lebanon.
So despite relying on gamification, it does not encourage laymen to use iNaturalist or longer than five minutes. This is further reflected on the photo quality of nearly half of the observations from locations close to my village and Isserstedt. Instead, such apps are attractive among those already highly enthusiastic about flora and fauna that may regard badges and achievments as a mere bonus. a very tiny minority, however, may only be focused on collecting badges and thus put a lower emphasis on learning more about nature and engage with other citizen scientists. And ou really won't need a research paper to notice a stark bias favoring large cities and tourist locations. They largely collect data for locations that already were well represented in records published by professionals, there are few to even no gains for regions ignored by professional scientists.
Back to Flora Incognita and what the decline of TT and iNat's stark biases mean for this project. Quckly skimming user profiles and locations is nearly to entirely impossible and FI does not publish any data on how their app is "growing". What I can say, however, that the app has been silently changing a handful of my own observations to empty, "unidentified" entries. The new AI powering the identifcation now misses more often than prior to the UI change and the new UI itself, pardon my French, runs like ass on my ancient iPad mini, with its new autofocus ALWAYS being off in rather hilarious ways. Because I tested this new app on my iPad and was incredibly disappointed by it, I still use the last version with the app's old UI on my iPhone. While no longer being updated, I still can access the database, which is continiously updated, and access the new AI. The client itself largely is a wrapper, after all.
But because of this, even scans made with my iPhone are not immune to observations suddently being rendered invalid. So far, I can't make out a certain pattern that may explain this behavior; it truly happens randomly and no other user so far has pointed this out. I also can't tell whether this is exclusive to user's own observations or may even extend to the maps of each individual species.
But hey, at least I now get notifications about new badges users can claim and stories about how taxonomists actually being fed up with breeders breeding new thymes² and stories about cannabis just because Germany somewhat legalized it. Meanwhile a lot of plant profiles, mainly those of species less often encountered in Ilemnau and Jena, still offer little to no information on pollination, its usage possibilities, possible toxicity and descriptions that are not limited to one or two short sentences. FI hardly encouages me to voluntarily contribute, in fact I only contribute because each scan is submitted automatically – and then gets randomly wiped by FI itself.
Citizen science sounds neat in theory, in practice it just adds more noise to the already huge pile of noise generated by professionals. In most cases, its effects are limited to short-term curouslity by layman that never spent more time and ressources on learning and engaging beyond the average length of a TikTok video. Citizen Science projects are just another specific way of throwing something at a wall and propagating whatever sticks at said wall in a positive light and in the most superficial way possible.
Don't tell my state, which now proudly claims that we offer 55k plant and animal species without providing a list of all species, that Flora Incognita no longer discriminates against domesticated plants and breeds.
---
¹
"Gamification: Is it game over?" – BBC.com (18 November 2014)
² Instead of rather being glad to never run out of work to do and thus never be at the risk of losing their jobs, FI published a story in which it was stated that taxonomists classifying new breeds of thymes are overworked and pretty much tired of their very job. I should've kept a screenshot of this because such stories disappear after a certain amount of new stories but even my trusted zoologist from Austria shook his head after seeing this on his own device.