Thanks, Rebecca! :-)

Anyway, after seven(!) desc rejections, which is by far my record, we made it through in @oikosjournal.bsky.social. Thanks, Oikos, for appreciating the resolution over extent! ❤️

nsojournals.onlinelibrary.wiley.com/doi/epdf/10....

The strength of our data is, however, not supposed to be its spatial extent, but its spatial resolution and related sampling density of 900 sites covering a lot of the fine-scale variation of this mountain. Which is a very rare thing. But apparently not easily sold to journals.

Up to now, it wasn't very difficult. But now, the process of getting this published has started. From basically every journal, we got quite nice and often positive constructive feedback, but concerns about the size of the study area, which is only one mountain.

Second, we tried to narrow this diversity down, keeping on mind all to colinearity in these data. It's dificult to put it in 250 words, so just check our paper. ;-)

Eventually we foud out that only 6 variables are enough to model even a quite large number of species. Out of 250!

First, we looked at how they correlate with species distributions, comparing them with Bioclim variables most commonly used for this purpose.

And, a first surprise, they were not doing well at all. That's not a criticism of Bioclim, but for microclimate-based models, they are probably not the best.

Even though we only used soil temperature from our Schrankogel loggers, we reached approximately 250 temperature variables.

The story started sometime in 2023 when I wanted to test if microclimate temperatures are indeed multifaceted as Christian Körner and Erika Hiltbrunner emphasized in their 90 ways to describe plant temperature. doi.org/10.1016/j.pp...

literally

Ah, and actually discovered one more cause of invalid external pointers in the way. terra rasters are non-exportable. stackoverflow.com/questions/76...

Oh, by comparing with your code, I found my bug. What a ridiculous one. 🤦‍♂️ Within the arguments of `runauto`, I was calling my own function defined outside the foreach. So it was that one that wasn't loaded on cores. Thanks a lot, man.

Thanks mate for coming back to this! Yes, I used .packages = ... argument in foreach.

From a package (github.com/ilyamaclean/..., microclima::runauto).