Led by Viktoriia Radchuk, an international collaboration shows phenological shifts in response to changing temperatures allow populations to remain stable or even increase in numbers.
Support from #sDiv @idiv-research.bsky.social
Enjoying my low carbon footprint form of travel toward #ecobhvr2025, I am re-reading some accounts of environmental/societal destruction caused by the extraction of rare natural products that dazzle most of us...
My selection of illustrated stories:
- 1. mammoth ivory: www.rferl.org/a/the-mammot...
It turns out that since the 2 legs (Berlin -> Paris and Paris -> Montpellier) had not been purchased together, the SNCF could not put me up in an hotel at their own cost. They did, after a smile, gave me a free 1st class ticket for the next day, so I am now in a fast train towards Montpellier 🛤️
It looks like that the direct train between Berlin and Paris is so delayed that I will miss my connection to the night train... So I'll spent a night in Paris, exciting... The adventure continues. #SlowTravel & #SustainableTravel to #ecobhvr2025!!
Just boarded a train with my bike in Berlin. If all goes smoothly I should reach Montpellier within 21 hours, with plenty of time before the launch of #ecobhvr2025. Only one change in Paris... So, time to unpack the laptop and check my slides (perhaps after an episode of Wednesday #2...).
I have just read the Anti-Autocracy Handbook and I think everyone else should too:
sks.to/autocracy
Thanks to @lewan.bsky.social and others for putting this together!!
😱😱😱
I just wrote a tutorial about how to use #Rstats & RStudio using a container system (podman): gist.github.com/courtiol/ce0...
I did this for myself (a linux user), but it may be useful to others too.
Let me know what you think of it.
Do Spix's macaws really behave as they do in the animated films Rio? To find out, read our new preprint: doi.org/10.1101/2025...
Our paper on how climate change impacts bat hibernation is out🎉 It was a long journey paved with many desk rejections despite a huge amount of work put in collecting and analysing the data, but we are proud of the result and very happy with the one journal that did review it: doi.org/10.1111/ele....
Working in or with Marine Protected Areas?🌊
We need your help! In this video, Christine Ward-Paige, introduces our workforce survey:
youtu.be/M3Nm5KuyN_Q?...
This is an effort involving Blue Nature Alliance, MCI, IUCN, @rewildorg.bsky.social, Leibniz-IZW and others.
Please share!!
Our paper on ducks in the city is out! Using 16 yrs of data (>1000 breeding events) collected by NABU Berlin (berlin.nabu.de), we looked at why many ducks nest on buildings. The landscape doesn't matter much, it's more an individual choice and the appeal of the breeding site: doi.org/10.1093/jue/...
Yes, strict double blind requirements from the journal.
Anonymizing an #rstats 📦 for a submission to Methods in Eco Evo, I replaced names in DESCRIPTION, README, vignettes, doc... But calling packageDescription("pkg.tar.gz") revealed that the computer name is used as metadata 🤨
Solution: build your tar.gz in a terminal using R CMD build --user=anonymous
Done! I reworked the filters and reduced the rate of both false positives and false negatives. It should now catch your pics @matkempphotos.bsky.social
Na, to find wildlife pics I should make sure that #wildlifephotography is enough. I personally like the combo 🪶📷 as it is short and pertinent albeit bird biased.
To repost wildlife pics I simply detect messages with at least one element among "#photography #picture #pics #camera 📷 📸 🖼️" AND one among "#wildlife #bird #animal 🪶". I also try to remove messages with keywords that tend to be associated with false positives. These rules could be expanded!
What do you call usual? Using c_across? If so, lay runs much faster. Check the vignette or README on GitHub.
Landed in Budapest. About to spend a week with the Finnish national team of horseback archery! 🏹🐴🇫🇮
Why a growing population is not necessary a sign of its environment getting better?
Check out our latest preprint:
doi.org/10.1101/2024...
-> huge amount of work from former postdoc Liam Bailey, now data wizard for WWF.
It is also the result of the amazing data collected by hyena-project.com
I've just created a feed called "wildlife photography".
It catches posts with pictures that are about wildlife through a collection of hashtags that I will continuously refine to improve the feed.
It was easy thanks to skyfeed.app
Please pin/save and let me know what you think.
#wildlifephotography
I took that photo in Tokyo (Higashi-Yukigaya 2002). The birds sometimes install such nests on power lines of Japan Railways and need to be removed in order to avoid power outages.
If people wonder why {terra} and not {stars}. My main answer would be that all plotting functionalities of the #rstats 📦 IsoriX are based on {lattice} via {rasterVis}. The latter 📦 now handles {terra} objects very well. Had I switched to {stars}, all plotting functions would have had to be recoded.
Without {raster}, the #rstats 📦 IsoriX lost the ability to store objects it creates as binaries. So I had to implement S3 & S4 methods for that, building on {terra}'s wrapping functionalities. I also discovered that `raster::aggregate` and `terra::aggregate` have different default `na.rm` behaviour.
I thought that the transition would be trivial, but it turned out not to be. Without {sp}, IsoriX initially lost the ability to plot oceans, which are polygons with holes (i.e. polypaths). So we had to add this feature in #rstats 📦 {lattice} and {rasterVis}, which IsoriX uses.
More 👇
IsoriX v0.9.2 -- an #rstats 📦 for building isoscapes using mixed models and inferring the geographic origin of samples based on their isotopic ratios -- is now on CRAN.
The main change between v0.9.1/v0.9.2 and earlier versions is that it no longer imports {raster} & {sp}, but {terra}.
More 👇
You should not decide based on the effect sizes but on how well it fits. If you impose a continuous effect, it can indeed artificially increase effect sizes if the real effect is not linear. Then the predict functions can sample different values along the continuum, hence your problem. -> factor
