Modelling vegetation types

[Weverton Carlos]

I'm using Maxent to estimate the suitability area for some vegetation types (rainforest, dry forest, savanna, etc.)

I have occurrences of thereabout 15 diffent species of each vegetation type. 

I'm not sure how I should organize that ocurrence samples: should I consider all species from the same vegetation type as being a single specie in CSV file or should I model each specie of the same vegetation type separately (and join the results after)?

For example, if I consider all species from the same vegetation type as a single specie, the CSV input will be:

Specie, Long, Lat

Rainforest,-50.5468010,-25.9716870

Rainforest,-48.9486110,-25.6069440

Rainforest,-52.5217000,-25.3550000

Rainforest,-49.7931000,-25.4929000

Rainforest,-50.3962900,-25.8549270

...

On the other hand, if I model each specie separately:

Specie, Long, Lat

Asterostigma reticulatum,-50.5468010,-25.9716870

Aulonemia cincta,-48.9486110,-25.6069440

Bernardia confertifolia,-52.5217000,-25.3550000

Bernardia confertifolia,-49.7931000,-25.4929000

Croton polygonoides,-50.3962900,-25.8549270

Any suggestion? 

Weverton C. Trindade.

State University of Ponta Grossa (Paraná, Brazil)

[Adam Smith]

Hi Weverton,

This honestly sounds like an empirical question--I can think of papers that have done it either way.  I am guessing the "correct' way depends on how much your species are endemic to each ecoregion.

Best,
Adam

[Sanjo Jose]

I'm also trying to check the suitability of different vegetation types using MaxENT.  If we go for modeling each species separately it won't yield our desired output.  I think all species of the same vegetation types can be compiled together to come under a type name and we can model it. The only thing is that you should select the dominant tree species in an area. We can check the accuracy of the current scenario comparing with the existing vegetation map.

Regards,

Sanjo

JRF, Centre for Climate Change and Adaptation Research,

Anna University,

Chennai, India 

ᐧ

Sanjo Jose V,

Junior Research Fellow,

Centre for Climate Change and Adaptation Research,

Anna University,

Chennai - 600025

Mob: 9400824393

Research Gate

--
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+unsubscribe@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.

[Brunno Gmail]

Ola Weverton,

Peço licença para te responder em portugues.

Com relacao a sua primeira opcao, modelas todas as especies como sendo um unico vegtation type, caso voce tem muitas observacoes em cada vegetation type,  pode ser que esse metodo gere um resultado similar ao de gerar pontos aleatorios no background dos vegetation types e usar esses pontos nos modelos de nicho.

Com relacao a opcao de se modelar cada especie separadamente e depois juntar os modelos, isso vai depender de se voce tem pontos de ocorrencia suficientes p cada espécie,  e se vc tem especies suficientes p representar o pool total de especies em um vegetation type. 

Conceitualmente, a diferenca entre as duas formas de molagem que voce porpos no email é bem sutil. 

Brunno Oliveira

Wildlife Ecology and Conservation

University of Florida, Gainesville, FL, USA

Web oliveirabrunno.wordpress.com/

Sent from my iPad

--

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.