[R-sig-eco] NA in species scores using capscale in vegan

[Jari Oksanen]

On 07/11/2013, at 12:05 PM, Duarte Viana wrote:

> Hello all,
> 
> I am doing a CAP (or dbRDA) in vegan, using the function "capscale", where
> the response is a dissimilarity matrix (Sorensen's distance index) and the
> explanatory matrix consists of 19 variables. The code runs apparently fine,
> but the species scores appear as NA (site scores are properly given).
> 
> Does anyone knows what might be the problem?
> 
Duarte,

Did you supply species data? Species scores cannot be estimated without information on species. If your response is a dissimilarity, there is no information about the original species. You can supply the original community data (which has the information on species) as argument 'comm' in capscale().

Cheers, Jari Oksanen