Seeking Advice on using BioGeoBEARS with Incomplete Continental Data-reg.

[SK]

Hi all, I am  currently a postdoctoral in India, focussing on subterranean taxa called  Bathynellacea, which is distributed across the continents. Despite numerous species being described from the South American (Neotropical) region, there is currently no molecular data available from that continent. In discussing this with colleagues, I have received conflicting opinions, some argue that BioGeoBEARS may not be appropriate under such conditions, while others suggest it is still worth pursuing and encouraged me to proceed. This has left me uncertain, especially as there are no published studies on this group using BioGeoBEARS that I could consult. I would greatly appreciate your expert opinion on whether it is methodologically sound to use BioGeoBEARS in such a scenario, and if there are any precautions or alternative approaches I should consider. Thank you in advance for your time and guidance.

Regards

[Nick Matzke]

Hi all, I am  currently a postdoctoral in India, focussing on subterranean taxa called  Bathynellacea, which is distributed across the continents. Despite numerous species being described from the South American (Neotropical) region, there is currently no molecular data available from that continent. In discussing this with colleagues, I have received conflicting opinions, some argue that BioGeoBEARS may not be appropriate under such conditions, while others suggest it is still worth pursuing and encouraged me to proceed. This has left me uncertain, especially as there are no published studies on this group using BioGeoBEARS that I could consult. I would greatly appreciate your expert opinion on whether it is methodologically sound to use BioGeoBEARS in such a scenario, and if there are any precautions or alternative approaches I should consider. Thank you in advance for your time and guidance.

Hi! Without molecular data, presumably it is hard to get molecular branchlengths or time branchlengths.  If you can get time branchlengths some other way (morphological branchlengths?), then you can run BioGeoBEARS in the usual way. 

If you just have a cladogram, as long as it is a binary tree (no polytomies), you can physically run BioGeoBEARS on it if you put in some fake branchlengths, e.g. make all branchlengths have length 1.  However, the question is, what is the interpretation of what you are doing?  Parameters "d" and "e" would become "per branch rates" rather than "per unit time rates", but the branches in real life would be different lengths. Perhaps it resembles a parsimony analysis (BUT I add immediately, this is not identical to a parsimony analysis). 

Cheers,

Nick

On Wed, May 28, 2025 at 4:44 PM SK <shabu.b...@gmail.com> wrote:

Hi all, I am  currently a postdoctoral in India, focussing on subterranean taxa called  Bathynellacea, which is distributed across the continents. Despite numerous species being described from the South American (Neotropical) region, there is currently no molecular data available from that continent. In discussing this with colleagues, I have received conflicting opinions, some argue that BioGeoBEARS may not be appropriate under such conditions, while others suggest it is still worth pursuing and encouraged me to proceed. This has left me uncertain, especially as there are no published studies on this group using BioGeoBEARS that I could consult. I would greatly appreciate your expert opinion on whether it is methodologically sound to use BioGeoBEARS in such a scenario, and if there are any precautions or alternative approaches I should consider. Thank you in advance for your time and guidance.

Regards

--
You received this message because you are subscribed to the Google Groups "BioGeoBEARS" group.
To unsubscribe from this group and stop receiving emails from it, send an email to biogeobears...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/biogeobears/e641fd65-521f-4c6b-9d68-9c111aaf0ee4n%40googlegroups.com.