[R] error propagation

[Manuel Gutierrez]

Are there any functions to do error propagation in R?
 I have done a search with little success. Any
pointers to read about this topic are greatly
welcomed.
My specific problem is: I use a linear model (lm) to
predict the biomass of an individual in a population,
then I add up the biomass of all individuals to
calculate the total mass of the population. I want to
calculate the error in the final estimate of the total
biomass.
Also, I have another case where I use a nls model
instead of the lm, but in nls se.fit is currently
ignored as stated in the help page. Is there an
alternative way to calculate the errors in the
predictions from nls?
Thanks,
Manuel