[R] question regression trees

[Claudia Romero]

Hello,
This is my first time addressing such a big audience so apologies in 
advance in case I fail to formulate this question.

I am working with 13 species of trees, and the data I have are:
  1 continuous (phenolic concentration in xylem and in phloem) and 2 
categorical variables: lineage (3 subclades) and habitat (fire and non 
fire).

I am trying to see how species can be splitted 'objectively' based on 
these variables. I tried to do a regression tree using the rpart 
library, but repeatedly got the following answer, even when I tried to 
run it using ONLY the categorical variables:

 > plot(fit, compress=TRUE)
Error in plot.rpart(fit, compress = TRUE) :
         fit is not a tree, just a root

Can anyone please help me think about this?

Many thanks,
claudia romero