Constraining extrapolation by distance (of continuous suitability from recorded sites)

[ah3881]

Does anyone know of a way of telling Maxent outputs to constrain the
projection of suitable areas to those connected to known recorded
areas. I.e If species A is recorded repeatedly in areas of East
Mainland tropical Southeast Asia is it possible to tell the model not
to extrapolate unconnected suitable areas in Borneo/ Bhutan, etc?
Thanks in advance

Alice

[Niels]

Dear Alice,

You'll have to include an additional layer where Borneo has a
different numerical/categorical value as Bhutan,

Niels

[David Galbraith]

There's a broader issue of species' dispersal probability and rates with respect to habitat suitability, landscape configuration and connectivity, and proximity to propagules/juveniles...  No?  I'm not sure if Alice is studying a threatened/endangered raptor, or an invasive algae.  Regardless, I'm trying to come up with ways to transform suitability indeces such that they can be 'cost rasters' in a cost-distance analysis.  Barriers to dispersal could then be weighted with an extremely high 'cost' to species dispersal, and highly suitable areas could approach a 'cost' of 0.

-Dave

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To post to this group, send email to max...@googlegroups.com.
To unsubscribe from this group, send email to maxent+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/maxent?hl=en.

[David Le Maitre]

Alice & Dave

 

You can constrain it anyway you like, for example you can include a categorical variable which has missing values except in the areas you want to focus on. The issue is your rationale for doing this, you need a logical reason for doing so whether that is biogeographical, ecological or simply the area. Another constraint can be dispersal rates over a given time period - essentially a form distance buffer around your known localities. Jane Elith eta al. discuss an develop an example of this type of approach in their excellent paper on using Maxent (http://onlinelibrary.wiley.com/doi/10.1111/j.2041-210X.2010.00036.x/abstract)

 

David

>>> David Galbraith <davidmg...@gmail.com> 2011/10/13 02:50 PM >>>

--
This message is subject to the CSIR's copyright terms and conditions, e-mail legal notice, and implemented Open Document Format (ODF) standard.
The full disclaimer details can be found at http://www.csir.co.za/disclaimer.html.

This message has been scanned for viruses and dangerous content by MailScanner,
and is believed to be clean.