Hello Animes,

it depends on what you want. At least it makes sense to investigate only sites where the species principally occurs. I would not conduct a study on elephants in Europe, just as an example.

If you only look at sites where you find the species, you might be more interested in factors you suspected to positively affect abundance or density. If you have factors in mind that might also negatively affect the species, it would be important also to include sites where the species principally should occur but actually doesn't.

Principally, it would be more adequate to sample quadrats in an area where the species principally occurs, and then look at correlations of factors within the quadrats with numbers of salamanders.

Best regards

Armin

I basically want to investigate how environmental variable corelate with a species community. So what do u think? Is it ok?

Is it a study on a community of more than one salamander species? And maybe other groups with more species as well?

You could then try a canonical correspondence analysis which visualises the correlation of several species with several environmental variables at several sites. But this is more an explorative analysis.

Yeah

I am going to do it for all amphibian species.

Then you may try CCA (see above) which is a multivariate analysis (several species).

All the best

Armin