Wikipedia talk:WikiProject Birds

0 views
WikiProject Birds
General information
Main project page talk
Naming and capitalization
 → Article requests
 → Spoken Article requests talk
 → Photo requests talk
 → Attention needed talk
 → New articles talk
Project portal talk
Project banner talk
Project category talk
Departments
Assessment talk
Collaboration talk
Featured topics talk
Outreach talk
Peer review talk
Country lists talk
Bird articles by size talk
Hot articles talk
Popular pages talk
Task forces
Domestic pigeon task force talk
Poultry task force talk
edit · changes

List of non-native birds in Great Britain

[edit]

Found this page List of non-native birds in Great Britain, almost entirely unsourced, and much of it worthless speculation, and very out-of-date. A complete mix-up of established feral populations, casual breeding by non-established species, and pure escapes with no hint of breeding. Delete the whole thing (my preference)? Or is any of it worth salvaging? - MPF (talk) 01:12, 2 December 2025 (UTC)[reply]

Anyone? - MPF (talk) 11:07, 12 December 2025 (UTC)[reply] Off it. Craigthebirder (talk) 20:15, 13 December 2025 (UTC)[reply] I saw this when you posted it almost two weeks ago and have been so busy that I never got around to replying until now, apologies! Despite the current issues with the article, I think there is value in a list of this nature, if someone is willing to put the time and effort in to improve it. While it only has two (rather poor) inline citations, the sources provided as external links are actually quite useful. The term "non-native" in this context might seem a bit arbitrary, but it is a term that appears to be used extensively in the context of British ornithology/conservation biology - they even have a non-native species secretariat! The British Ornithologists Union has a useful category system for species on their official British birds list, with non-natives mostly falling under categories C (partly - categories C2 and C3 refer to translocated but apparently native birds and birds re-introduced through conservation respectively), D (somewhat tentatively - birds listed in category D are essentially birds of doubtful natural origin who are pending placement in category A or E), and E. The British list is freely available on the BOU website, and there are a number of relevant articles on non-native birds in the UK in the BOU's journal Ibis. For non-natives recorded or reasonably expected to be breeding in the UK, the Rare Breeding Birds Panel has been recording information on rare and non-native birds breeding in the UK for over 50 years, with all of their reports publicly available on their website. I think this is a topic that satisfies WP:LISTCRIT and could actually be a very interesting and useful list if it were to be reworked based primarily on the BOU's British list categories C1/4/5/6 and E. Ethmostigmus 🌿 (talk | contribs) 05:58, 14 December 2025 (UTC)[reply] @Ethmostigmus - sounds a good idea. I'll add it to the long mental list of things to do, but probably fairly low priority . . . MPF (talk) 23:15, 14 December 2025 (UTC)[reply] See experimental AI generated list at User:Snowmanradio/Sandboxes/List of non-native birds in Great Britain. Users are welcome to edit the sandbox page and make comments n the talk page. After copy editing, several users will need to approve it (on the talk page) before it is moved to a Wiki article page. Snowman (talk) 12:22, 13 January 2026 (UTC)[reply]

Left/right flipping of pics for taxoboxen

[edit]

Is there any markup which can be used to make a right-facing bird in a photo, flip to face left, so it looks 'into the page' in a taxobox? - MPF (talk) 23:19, 14 December 2025 (UTC)[reply]

I asked for such a function to be made years ago, but it was denied. All that can be done is flip image templates without captions, otherwise the captions are also mirrored too. FunkMonk (talk) 06:30, 15 December 2025 (UTC)[reply] Template:Image flip works fine with captions. If we are allowed to use it to flip real photographs in taxoboxes, I don't know. --Jens Lallensack (talk) 08:47, 15 December 2025 (UTC)[reply] Huh, weird no one pointed out back then[1] that the functionality to not flip captions already works. FunkMonk (talk) 09:27, 15 December 2025 (UTC)[reply] @FunkMonk @Jens Lallensack Thanks! That looks like it is designed to work with [[File:Xxxxxx.jpg|thumb]] markup though, and wouldn't work in a taxobox? - MPF (talk) 14:02, 15 December 2025 (UTC)[reply] It works in a taxonbox with a syntax like this: image = } But again, it's kind of a hack and I'm not sure if we are encouraged to do this to begin with. --Jens Lallensack (talk) 14:21, 15 December 2025 (UTC)[reply] I just tested in my sandbox, and it seems to work on desktop, mobile, and the app without side effects. --Jens Lallensack (talk) 14:28, 15 December 2025 (UTC)[reply] @Jens Lallensack thanks! I don't see why we shouldn't use it; it is within the remit of the derivatives that creative commons licensing allows, and for bilaterally symmetrical birds, flipping an image left-right doesn't affect reality. Just don't use it for Wrybill or crossbills, or photos where there's text or a location-identifiable object in the background. I'll try it out now on Roseate Tern (since that's one of my own pics) - MPF (talk) 20:20, 15 December 2025 (UTC)[reply] @Jens Lallensack done; looks OK to me 👍 - MPF (talk) 20:23, 15 December 2025 (UTC)[reply] What about when there are background features like buildings or landscapes, where it would be a falsification to mirror right to left facing. There must be guardrails. Snowman (talk) 15:47, 12 January 2026 (UTC)[reply] I have doubts that left to rightr facing switching this is authentic. It is not allowed in the press in the UK, as far as I am aware. Snowman (talk) 15:49, 12 January 2026 (UTC)[reply] Rethink: I think flipping images should not be done. Snowman (talk) 10:37, 13 January 2026 (UTC)[reply] One issue is that if a user clicks on the image to view it, it will not be flipped anymore, certainly causing confusion as nobody expects this ("wtf, this is not the image I clicked on!"). --Jens Lallensack (talk) 10:40, 13 January 2026 (UTC)[reply] I can confirm this odd behaviour aftter clicking on the flipped image on the Roseate Turn page. I think the original non-flipped image should be returned to the infobox and the flipped copy deleted from Commons to reduce confusion. Snowman (talk) 12:47, 13 January 2026 (UTC)[reply] @Snowmanradio - it doesn't need any deletions at Commons, just removing the "image flip" markup from the page here. If it's the consensus not to use image flips, I'll remove it - MPF (talk) 11:40, 14 January 2026 (UTC)[reply] There are more subtle issues too; while the birds might be completely symmetric, some plants that might also be in the photo are not, for example vines can coil left or right, depending on species. I feel that the negatives outweigh the positives, and would also vote for removing the image flips. --Jens Lallensack (talk) 11:54, 14 January 2026 (UTC)[reply] I have reverted the flipping of the roseate tern infobox image. Snowman (talk) 13:10, 14 January 2026 (UTC)[reply]

Automation project:

[edit]

It was about over 15 years ago when I asked WP:Bird writers for ideas on tasks where automation would help.

Back then, after a suggestion, I designed an automated method for checking all the WP:Bird taxoboxes (over 10,000), which I would say was a success. That was before AI. With AI getting better, I wonder about emproving selected articles with AI. I do not know how it will work out, but I think it is worth a try. I think it would need these components to negate any doubts about AI generated or part-generated material:

  1. Users giving suggestions on WP Birds pages that need improving with AI assistance.
  2. Sandboxes for the selected artices with their talk pages. Probably a max of three bird articles at a time depending on participation.
  3. Reviewers to put the AI generated articles into shape, discuss the topic on the talk page, and to suggest reliable sourses for the AI to look at.
  4. Votes to show when is a consensus to move the sandbox into the main Wiki space. I guess any article without this approval would not be acceptable to the wiki community.

Any comments? Snowman (talk) 15:44, 12 January 2026 (UTC)[reply]

The review is important. Articles produced by LLM without adequate review can be speedy deleted on that basis alone, so any AI improvements should be able to pass Zerogpt or similar Jimfbleak - talk to me? 16:43, 12 January 2026 (UTC)[reply] Yes. This would depend on several interested users activly reviewing, editing, and finally voting on suitabitity as an article. This documentation will be kept on the article's talk page. I think the Wiki will have to come to terms with AI generated material soon. I think this is what we can do right now. Is Zerogpt jargon? What exactly does ZeroGPT mean? Snowman (talk) 17:18, 12 January 2026 (UTC)[reply] I'd not trust a large language model to keep species straight in its "mind", especially given the inconsistent ways species are named. I imagine an AI could be useful if it could review existing articles and tell us what it thinks is wrong in them, then we could check and see if it is right and make the changes. That would be useful. There are over 10,000 species and many articles more than that, I'm sure there are errors we should find and correct. ZeroGPT is one of many AI systems which claim to be able to detect AI. I'd trust it as far as I could understand its programming. That is, not at all. SchreiberBike | ⌨  17:36, 12 January 2026 (UTC)[reply] To me, looking for mistakes in a large batch of wiki articles probably is possilbe for AI, but it might not be easy to make this consistent with Wiki standards and rules. Snowman (talk) 18:35, 12 January 2026 (UTC)[reply] Essentailly, my idea is to use AI openly for AI-assisted improvment of selected articles. I might be interested in experimenting with AI article reviews for good article nominations. I think a small team of wiki editors and AI can do this now. Snowman (talk) I'm not sure if you're aware of this, but LLM usage in the context of Wikipedia is a highly controversial topic that has spawned a lot of very heated arguments, and many editors have very strong opinions about it. WP:NEWLLM is a recently adopted guideline which states Large language models should not be used to generate new Wikipedia articles from scratch, and several proposals have been made for additional guidelines of a similar nature. I have personally had to do cleanup (we even have an AI cleanup WikiProject!) on a number of articles that were primarily generated by LLMs, and even with human oversight I have seen many, many errors slip through. You seem reasonable and aware of the shortcomings of this technology, but please exercise the utmost caution when working in mainspace. Ethmostigmus 🌿 (talk | contribs) 03:52, 14 January 2026 (UTC)[reply] I don't think it is a good idea to let an AI write articles. The main issue is hallucinations; such a text would have to be checked against the sources claim by claim, implication by implication, as the AI can make unexpected errors that a human would never make. Just look at the La Isla Bonita debacle, an AI article that passed FAC because it looked great, but has serious sourcing hallucinations that a normal spot check didn't detect. Properly checking that article requires so much editor time (and who wants to check AI-created content in the first place?), and just writing it from-scratch might be the easier (and more enjoyable) approach. --Jens Lallensack (talk) 10:45, 13 January 2026 (UTC)[reply] By way of an example; see the above discusssion, about the list of non-native birds in the UK. I have made an AI generated list at User:Snowmanradio/Sandboxes/List of non-native birds in Great Britain. Users are welcome to edit the sandbox page and make comments n the talk page. After copy editing, several users will need to approve it (on the talk page) before it is moved to a Wiki article page. Snowman (talk) 12:32, 13 January 2026 (UTC)[reply] I guess a lot of unskilled use of AI has got AI a bad reputation here. WP:Birds has agreed (see topic above) that List of non-native birds in Great Britain needs a tidy up, becuase the existing page is a bit of a mess, so I have experimented and made an AI improved version in a sandbox at User:Snowmanradio/Sandboxes/List of non-native birds in Great Britain. I inputed the current bird list artilce at the start of the process. I put it through several AI revisions in a canvas (not merely input -> output as one simple prompt). After thatl, I did some manual edits, which you can see on the sandboxed page history. It has not been made entirely from scratch by AI, so it should not automatically banned. I have had a quick look at the guidelines, and is seems that very little has been agreed on the use of AI, so far. Have users made a judgement without looking at the sandboxed ariticle? Snowman (talk) Sorry but did you even look at the main source (the British Bird List)? How many species are in that E list and you just list six? The Red-winged Laughingthrush is not even in the British Bird List as far as I can see, I guess because the population went extinct. That's the hallucination problem I was talking about. Our current article is in bad shape, but it does not have this mistake (it has the thrush, but it does not claim it's in E* or that the British List has it). Sorry but the draft looks useless to me, and not an improvement over the old one. --Jens Lallensack (talk) 14:04, 14 January 2026 (UTC)[reply] Do you really expect about 50 or 60 birds listed just becuase one or two (or a low number) of each species escaped. I could simply add that the BOU list includes may sightings of escapies, but they are low numbetrs and not listed in the Wiki list. The ostridge is listed by the BOU, I think. Snowman (talk) 19:56, 14 January 2026 (UTC)[reply] For a section titled "Category E: Escapes and casual breeders", I indeed expect that all species in Category E are listed. Otherwise it is just a mistake (and again, a mistake that our current article does not have). --Jens Lallensack (talk) 20:00, 14 January 2026 (UTC)[reply] To me, it is outside the scope of the Wiki to document every bird (or small numbers) of a bird species that has escaped. Anyway, this is not an AI error, I could have easily asked the AI to add the full list, but it was my choice not to included specifically add it in. The scope of the Wiki is the duscssuion point here, not the role of the AI. Snowman (talk) 20:46, 14 January 2026 (UTC)[reply] I'm also curious as to why the AI has created a MOS:THISISALIST opening not in the current page. Is the general corpus of our lists dominated by that style of writing? CMD (talk) 17:38, 14 January 2026 (UTC)[reply] Thank you for pointing this MOS error out. I see this as a teethinhg proplem. I could stop the AI making this mistake in future. In this experimental situation, I have not showed the AI much MOS to work by, but I could do in the future. At the present time, I see this MOS error as my error, because I should have corrected the AI output manually myself. I have not been doing much editing for a few years, and I missed it. Snowman (talk) 19:45, 14 January 2026 (UTC)[reply] Re-think: the article is about the official list of bird species in the UK (cat E), so it seems natural to say so in the introduction. Snowman (talk) 20:56, 14 January 2026 (UTC)[reply] Not really looking at it as an error than a question. Knowing (or having a good guess as to) why an llm produces particular output may help refine things. They often create new reference fields for example, presumably in a desire to present more information linked to the existing fields. CMD (talk) 14:56, 15 January 2026 (UTC)[reply] Red-winged Laughingthrush is not anywhere on the British List, as the British List only covers Great Britain; the R-wLs were on the Isle of Man (where they have now died out, it is true), which has its own independent Manx list. For clarity, Northern Ireland isn't included either, nor are the Channel Islands. - MPF (talk) 18:14, 14 January 2026 (UTC)[reply] I think the AI got it right about the E* status of the red-winged laughing thrush on the Isle of Man. Look at the BOU source that the AI provided in the in-line ref (currently ref 22 in the sandboxed page), and you will see this note at the end of the BOU webpage:
"NOTE The following species has been recorded from the Isle of Man, but not since 2005: Order Passeriformes Family Leiothrichidae Red-winged Laughingthrush Trochalopteron formosum E* 6,8,12" "The British List: Category E Species" (PDF). British Ornithologists' Union. Retrieved 13 January 2026. Where exactly? I still can't find it. --Jens Lallensack (talk) 19:46, 14 January 2026 (UTC)[reply] I have fixed the link. The link was an AI error, but I have fixed it now. You can see it as a note at the bottom of the list, but above the references on the BOU 10th edition pdf. Snowman (talk) 20:33, 14 January 2026 (UTC)[reply] Ok, but that's the 2023 version of the list, not the current one. Hence, even if the AI had used the correct citation, it would still be a mistake (nothing in the draft says that different list versions have been mixed, and obviously we shouldn't do that in any case, and the reader of course assumes that the current list is cited, especially if the link says "retrieved in 2026"). Anyways, the AI selected a tiny fraction of the species without any explanation or criteria whatsoever (the old article has over 40 entries, the draft just 17). I stand by my opinion that we are better of, both in terms of quality and speed, if we improve the article the conventional way, and leave it as that. --Jens Lallensack (talk) 20:56, 14 January 2026 (UTC)[reply] I default to not letting any AI write anything intended for mainspace, either articles or lists. That could change, but I've seen so much bad stuff I don't trust it until it has proved itself. I've heard people say that it seems to sometimes get better, then it absorbs another source of data, or changes something, and gets worse without any reason the user can see. As something to check our work, I'd love to have access to that, even if it is not perfect (so long as it's pretty good). I understand professors have satisfactorily used AI to check papers for accuracy of facts and quality of writing. I think we could use that. SchreiberBike | ⌨  16:44, 14 January 2026 (UTC)[reply] Everyone is entitled to there own views and opinions. Snowman (talk) 19:56, 14 January 2026 (UTC)[reply] The complexity and nuance of this example you've posted is a perfect demonstration of why AI isn't any good for this. We've gotten this far with just people who care working n it. No need to defer the hallucinating robots. Sabine's Sunbird talk 18:39, 15 January 2026 (UTC)[reply]

Inconsistent Wikidata associations for Circus (Q207796) and Circinae (Q3325926)

[edit]

Some Wikipedia sites merge the concepts of the genus Circus and its subfamily / common name harrier into a single article, e.g. en:Harrier (bird), fr:Busard and de:Weihen (Gattung), with redirects en:Circus (bird), fr:Circus (genre) and de:Circus (Gattung). These three entries all use the common name as a title to represent the genus. The problem is that some of them are attached to Circus (Q207796) (genus) and some to Circinae (Q3325926) (subfamily) as in the three entries above. This prevents consistent sitelinks between them despite being the same concept and articles. If the solution is to put these articles under the same wikidata item, it's hard to tell which one to choose by looking at the sitelinks under both wikidata items. A quick fix might be adding the above redirects as sitelinks in the wikidata item where they're missing. So I'm creating this topic here hoping that a more experienced editor can chime in as to the preferred solution (if any). I'm sure this has come up in the past regarding inconsistent article titles across Wikipedia sites. Gumgl (talk) 23:23, 13 January 2026 (UTC)[reply]

This, unfortunately, a common problem. Wikidata requires one-to-one relationships for sitelinks. Wikipedia articles can't be linked to two wikidata items when the scope covers both, and different language wikipedias use different conventions for monotypic taxa (i.e. placing article at parent or child). I don't think there is a satisfactory solution. Wikidata now allows linking of redirects so a redirect at Circinae could link to the Circinae wikidata item, but this doesn't help with the sitelinks. In this particular case, it should solve itself in time. Cincinae is no longer used as the harriers are nested deep in the hawks (part of tribe Accipitrini). So the harriers should link to the genus wikidata items if using updated taxonomy. But some wikipedias are slow to change.  —  Jts1882 | talk  08:21, 14 January 2026 (UTC)[reply] Ideally, taxon pages on wikidata would link to taxon names on wikipedias and sitelinks would follow redirects for when articles are at common names. However, that allows navigation from wikidata items to wikipedia articles but doesn't populate the langauage sitelinks at the wikipedia articles. I've changed the en-wiki site link at Circus (Q207796) to Harrier (bird) so sitelink menu is populated at the article here. It still won't pick up sitelinks for Wikipedia articles at Circinae, but gets about 40 links including the bigger languages.  —  Jts1882 | talk  09:15, 14 January 2026 (UTC)[reply] What a mess! In some cases, the pages are duplicates that should be merged, e.g. at Italian Wikipedia it:Circus (zoologia) (connected to Circinae (Q3325926)) and it:Albanella (zoologia) (connected to Circus (Q207796)) cover the same topic, in both cases the genus Circus and not the subfamily. If anyone knows some Italian and is active on it:wiki . . . ? - MPF (talk) 12:00, 14 January 2026 (UTC)[reply]

Requested move at Talk:Pipipi#Requested move 15 January 2026

[edit]

There is a requested move discussion at Talk:Pipipi#Requested move 15 January 2026 that may be of interest to members of this WikiProject. TarnishedPathtalk 05:10, 24 January 2026 (UTC)[reply]

Ground-nesting birds

[edit]

Is there a page for ground-nesting birds somewhere? I wanted to link to it in Eider. WhatamIdoing (talk) 19:04, 31 January 2026 (UTC)[reply]

There is Bird nest#Scrape. SchreiberBike | ⌨  21:06, 31 January 2026 (UTC)[reply] Thanks. I made a redirect to that section. WhatamIdoing (talk) 21:26, 1 February 2026 (UTC)[reply]

Inclusion criteria for List of organisms of Place

[edit]

There is a discussion at Wikipedia talk:WikiProject Tree of Life#Inclusion criteria for List of organisms of Place about what should be included in such lists. Please contribute there. Thank you. SchreiberBike | ⌨  03:41, 4 February 2026 (UTC)[reply]

This article is sourced from Wikipedia. Content is available under the Creative Commons Attribution-ShareAlike License.