search

vjjan91 / nilgiris-resurvey-project

Historical resurvey of bird communities across the Nilgiri hills of the Western Ghats
GNU General Public License v3.0
0 stars 1 forks source link

review this script #10

Open vjjan91 opened 2 months ago

vjjan91 commented 2 months ago

https://github.com/vjjan91/nilgiris-resurvey-project/blob/916ead6085b23d12d5650c17f63404c6fa7fdc95/code/xx_relative-abundance-envtChange.Rmd#L6C1-L6C10

@pratikunterwegs I have currently removed this script from the main project and left it in another folder. I am not very confident if this should be a part of the main text, along with the beta regressions and rank abundances etc. What do you think? Is this an important part of the story or do we have enough stats to provide a convincing story?

pratikunterwegs commented 2 months ago

Hi @vjjan91 - looking at this, are you not including the results from this script in paragraph 2 of the Results? Line beginning: Relative abundance comparisons for generalist species across the three time periods revealed - what does this refer to?

vjjan91 commented 2 months ago

The results you are referring to are included in the last script 08_relative-abundance.

This script was initially to run a glm which examined delta abundance ~ delta climate + delta landCover + random effects of site and species

But I wasn't completely convinced if this needs to still be included in the current way we frame the story or do you think it's important?

pratikunterwegs commented 2 months ago

Having read through the ms, I think I would possibly include this analysis (and script), as it goes towards building the case for abundance changes being related to landscape changes. What do others like Morgan think?

vjjan91 commented 2 months ago

I see your point - the only qualm I have with this modeling approach is that none of the models seem to converge/suggest that there is indeed an effect of change in grassland area/change in climate on change in abundance; ie. declines in bird pop. Would you be able to review this script once? I think it is fairly well annotated and follows your initial approach to preparing the predictors for the glm.