(Feel free to point out typos. Or keep them. My arthritic hands are hurting.)
I've already voiced my highly-negative opinion on the "infamous Krefeld Study" a while ago and there is no point in rephrashing everything I said back then. For the sake of fostering quick understanding, especially among those who did and do not want to read my sloppy rant, here are the key points:
malaise traps even in potentially vulnerable environments, arbitrary choice of 96 legally protected "flatland" locations, arbitrary sample collection within chosen locations with 59% being underrepresented compared to one location that was sampled for times, overrepresentation of the data collected in 1989 and 2014, primary focus on "biomass measures" with species being determined AFTER the study was published and thus exact species not being made public, pure incompetence regarding the usage of statistics by leaving out null hypothesis testing among other questionable "homebrewed" models), blind application of Ellenberg's "Zeigerwerte" despite Ellenberg himself discouraging such practices
trends are predicted even though the data wasn't collected and used for statistical trend analysis (see statistical imcompetence above), data is applied to the entirety of Germany despite a clear bias towards western German states such as NRW and Lower Saxony, using the "Discussion" section to analyze a different study from the UK and reuse the largely-untested statistical model introduced by Krefeld earlier, excluding "landscape" and "climate change" as driving factors even though land changes and local climate changes were not recorded, merely implying pesticide usage to be the primary driving force despite not conducting any toxicological or even soil tests, overall loaded language and mere speculations
Now that we got this out of the way, we already can conclude that its "follow-up" study will be even worse than that, given the much shorter time span (2017 up until the early 2020's), even more randomly selected locations now across the entire country, and, surprising more importantly, a response to the "replication study" from Bavaria that was published in early March, 2025.
The funny thing about the response to Bavaria is that it has been published in Nature and thus got put behind its paywall. One researcher behind this response admitted his pity and how the group will be unable to make it Open Access for six months[1], however...
NO ONE FORCED THEM TO PUBLISH THEIR RESPONSE IN NATURE. In fact, they DID publish their original study in PLOS One, an Open Access journal, so why, all out of the sudden, decide to approach Nature and even pay a good amount of money to even get published in the first place? One can only speculate but just browsing through the public profiles of one researcher behind Krefeld will, at the very least, leave a bad taste in your mouth if you value scientific integrity over the high-dosage "moraline" that came with the study.
We're staying on this particular researcher's profile to get an idea what this group's actual motives behind their "Krefeld Studies" may be.
Thomas Hörren is one of the twelve researchers behind the original study. He joined ruhr.social on 30 October, 2022 and has published, as of 16 March, 2025, only 30 posts overall, three of which are reposts from other accounts, putting the total of his own posts at 27. Excluding responses, three posts deal with the promotion of books he (co-)authored, two posts are part of a thread where he introduces himself to the Fediverse, one post (his very first) promotes a pre-print he co-authored (but apparently never went beyond its pre-print status), one post was "reboosted" by himself, three post explain two particular insects, one discusses a particular jellyfish, one deals with urban plants, one is a microscopic shot with little information, one promotes a "networking course" for endometriosis patients, four post promote three podcast episodes in which Hörren took part, one post shows Hörren taking part in the children's show "Tigerentenclub"...
How many times have I mentioned "promote" so far? Five? Blimey, his account is overwhelmingly filled with some form of promotional content. Out of all 30 posts, only five are of the educational type (actually explaining stuff, excluding the pre-print), 14 are entirely self-promotional (including the pre-print).
46,7% of his personal account is some type of advertising, 16,7% is educational. Now let's include the replies.
One is a reponse to the response published in Nature where he mentions the group's inability to make their response open access, three are replies to some users asking questions about his books, one is a personal opinion on "national parks", one merely deals with how he's not going to add content warnings to his posts (a typical Fediverse drama topic). Six posts are replies to other users, four of which are connected to his PR posts.
Now adding those to the overall napkin numbers, this pushes the advertising content above 50%, at a whopping 66,7%. I'm not going to calculate the percentage of his more personal posts but you should be able to tell that this, alongside the promotional posts, decrease the amount of purely-educational posts significantly.
Now you may claim that there's nothing wrong with self-promotion on someone's own social media account; that's what "all" researchers do. If that's the case, why does he only talk about insects without somehow drawing attention to himself in only three cases overall? He is an entomologist, he should be babbling about insects like there's no tomorrow and perhaps even mention how each is dependent on specific plants and biotopes. There is NONE of the latter, in fact when he does mention plants, it's unrelated to insects.
It doesn't help that he gave an interview to the German newspaper "DER STANDARD" where its headline states that "we don't understand nearly every ecosystem". Now recall the conclusions that were drawn in the original "Krefeld Study" – personally I'm feeling scammed by Hörren and, by extension, his colleagues. They pretty much admit that they don't know shit and don't gain any insight from their own studies YET continue to insist on their own flawed conclusions and promote their books.
The perhaps most ironic part of the respose to the Bavarian study is that its preview alone attempts to debunk it by claiming that the Bavarian study was conducted in areas "manipulated" to improve the abundance of insects.
Wait a minute, Krefeld picked randomly selected areas that are "legally protected" and thus are are largely managed, as well. Only in the case of Krefeld, it is unclear how their chosen protected areas are protected in detail and what even is supposed to be protected and maintained in the first place, as those largely focusing on landscapes, especially those trying to protect "steppe landscapes", considering relatively high amounts of insects undesirable (at least this appeared to be the case in my FFH territory "Bog Hill" since planners advocated for intentional overgrazing).
But whilst voicing this in an interview with Riffreporter[2], Hörren claims that the Bavarian study is invalid compared to the yet-to-be-published Krefeld 2, not the original study which Bavaria attacked. Hörren also admits, likely without noticing, that Krefeld 1 never was "representative for the whole country", claiming that Krefeld 2 now includes 600 different spots across Germany and will continue to rely on biomass measures with malaise traps.
That's cheap projecting, plain and simple. Krefeld did the exact same stuff but now attacks Bavaria, which mostly re-analyzed the Krefeld data and added additional weather data AND spots + weather data in selected areas in Bavaria, for doing the same things and provide some crucial data Krefeld lacked.
Since this is giving me a migraine, I wanna finish this lousy rant by citing a comment from the original study:
"Looking at the data underlying this paper makes me feel uncomfortable the ease by which conclusions are drawn from weak data.The data were not collected for trends analysis; only one trap was used per location; only a few locations were resampled more often. All years with only 1 or 2 locations can hardly considered as a serious estimate for insect abundance, the exposure time greatly varies by years neglecting the impact of insect phenology and natural fluctuations in activity, the years 1989 and 2014 are strongly overepresented and more kinds of biases makes the data set very unbalanced. Qualitatively, the data may indicate a signal of insect biomass decline, but too many hard conclusions about the dramatic decline and speculation about the causes are disputable. I hope more ecologists, statisticians and entomologist will critically look at this paper to judge its significance and to ask for better data."
Ironically, right after I studied the comments, which are far more critical than German researchers AND our press that overwhelmingly and thus blindly support the Krefeld Entomological Society, I came across a pre-print on bioRxiv which takes a closer look at how biologists work with statistics and document their own methods. Take it with a grain of salt but it appears to be ridiculously easy to come across biological studies not even meeting formal requirements.
Now one could argue that applying statistics within certain biological disciplines is, at the very best, useless but straight up "cargo-cult science" in the worst cases. I can only speak for myself but I do agree that throwing statistics into the study of ecosystems neither adds anything relevant to the discussions nor helps flora and fauna in practice. Please remember that "statistical significance" seldom translates to "practical relevance" due to statistics largely being meant to make sense within themselves. They're largely isolated from reality, hence I do NOT rely on them. All I do is keep track of the development of the butterfly and odonata compositions in and around my village – I'm not gonna be obnoxious enough to claim that my village represents my state or even the entire country since I can already tell that locations nearby, which are similar to mine, demonstrate small but crucial diferences (e.g. missing species within either area, nutrient preferences of plants and animals, varying recklessness of one or two farmers etc.). I also won't use the collected data to make predictions that aren't as obvious as a severely damaged or completely destroyed biotope and even then I remain cautious and give this area the benefit of the doubt for a year. I'm keeping it humble and follow Feyerabend's "anything goes" philosophy as it makes more sense to me, my skills and my resources than "following protocols" and risk accidentally misleading people.
Maybe Krefeld could need some humbling but then again the head of the project is also in a leading position at NABU[3]. NABU is known to accept bribes[4]. Fuck them.
---
[1]
[2]
[3]
[4]
[5]
---
Edit (12 July, 2025): I wrongfully attributed the newspaper article to the Süddeutsche Zeitung. This has been changed to DER STANDARD and now also includes a proper source.