2
$\begingroup$

I am working with a whole genus of land snails, trying to predict its distribution in the coming decades and how it might get affected by climate change. I have already worked with MaxENT and the R package Biomod2 (I have tried out all the important algorithms within it as well). I was wondering if any of you can recommend another option thats reliable.

$\endgroup$
3
  • $\begingroup$ Anything with the negative binomial distribution will always work. $\endgroup$
    – M__
    Mar 1, 2023 at 17:40
  • $\begingroup$ It might be helpful to know what ecological data you are working from. $\endgroup$ Mar 5, 2023 at 0:41
  • 1
    $\begingroup$ Not sure if that's what you're asking but I'm working with presence only occurrence data, climate data from worldclim and I also possess phylogenetic data but I have yet to use them in such work. Also @M__ id like it if you were a little more specific and maybe give some reasoning behind your response. $\endgroup$
    – Nickmofoe
    Mar 6, 2023 at 18:02

1 Answer 1

0
$\begingroup$

Spatial species distribution always follow a negative binomial distribution at any point of most other factors e.g. climate. When a species enters a human monoculture type environment - then deviation from the negative binomial can occur within that monoculture. The clumping will be approximated via trigonometry: 'cause the clumping can be approximated via a circle, very clumped - small circle, more dispersed bigger circle.

The parameters of the negative binomial are likely to change in relationship to climate and thats what I'd model if I was doing this work. The 'clumping' (which is what the negative binomial models) is likely to tightened as global warming increases. In practical terms the snails start drying out so they have to clump in the remaining pockets of moisture (this might be more true of slugs than snails ... at a guess). Alot of invertebrates that cannot or don't really fly - are prone to desiccation.

Personally, I go to empirical stats rather than use a package. However, the packages you are using will follow these two principles. So you can do it from first principles. Its fairly simple ecology its not like phylogenetics that have been through decades of development.

If you've DNA from the snails then the phylogenetic Bayesian spatial models around Beast provide a powerful basis for the investigation. That would be very strongly recommended if you've DNA sequence from a good proportion of the snails.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.