Strategy on invasion ecology modelling

[Dimitris Poursanidis]

Dear group,

I have a question that several of us have already face or will during the career.

Am working on invasion ecology - particularly on species Native in Indo-Pacific and invasive in the Mediterranean.

species modelling, spread rates and niche analysis (Broennimann approach and Petitepiere)

The example below show that i have presence only data from both areas.

what strategy is "correct" for this modelling work ?

Use all data and train the model ?  10-fold model 

Use only data from Native range and allow for model extrapolation (the enviro data cover both areas) - test on invasive area data.

Use only data from Native range and project at Mediterranean performing also MESS analysis ? 10-fold method

Literature provide examples with all above approaches but no clear method is preferred.

I set this question also for brainstorming and opinions exchange

D.

[Sam Veloz]

Somewhat depends on what the goals are for your study. I would say that the combined presence data will give you a fuller sample of the environmental conditions where the species occurs than the data from either region. Selecting the appropriate background data is a non-trivial matter.

Cheers,

Sam

Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.
To post to this group, send email to max...@googlegroups.com.
Visit this group at https://groups.google.com/group/maxent.
For more options, visit https://groups.google.com/d/optout.

[Jamie M. Kass]

I agree with Sam if the motive is indeed to develop a “niche model”, as the species fundamental niche does include the climate at all the areas it has successfully invaded. A model trained on just the native range will be missing these parts of environmental space occupied by the invaded range that surely belong to the fundamental niche, and would then be unable to give accurate predictions in any of the invaded ranges. I know there are lots of papers on this, but will perhaps post some later.

Jamie Kass
PhD Candidate
City College of NY