[R-sig-eco] Rarefaction trouble

[Stine Kramer Jacobsen]

Hi!


I am struggling with some data - hoping to find some answers. My main
question is regarding the output when using the rarefaction function,
rarefy().



I have a dataset of six locations (in rows) and 45 species of insects (in
columns), and I want to express the diversity of insects in these six
locations and clarify the differences between locations. Pretty
straightforward I guess.



I have rarefied the data by this code:

rarefy(rarefactionlocation, 5, se=TRUE, MARGIN=1)


And get an output like this:



       [,1]      [,2]      [,3]      [,4]      [,5]      [,6]

S  3.3946786 3.6740240 3.8372812 4.0785534 3.8305276 3.7252981

se 0.9639723 0.9112285 0.8641984 0.8093713 0.8859364 0.9109916

attr(,"Subsample")

[1] 5



I am also able to get a nice curve of the data, showing the six locations
in separate lines, though I have been unable to distinguish lines with
different colours - yet. I did this by:



raremax<-min(rowSums(rarefactionlocation))

rarecurve(rarefactionlocation, step = 5, sample = raremax, col = TRUE, cex
= 0.6)



A colleague of mine have used the rarefaction function by Jenna Jacobs (I
could not get it to work), and she received an output in an excel file with
SE and approximate confidence interval for each 5 individuals for each
location. By these she could find if there were significant differences
between the different locations.



My question is: How can I implement the SE from the output to calculate the
differences (significance) between locations? Should I implement it into a
new code?


Or would another function be more appropriate in this case? I have also
been reading about specaccum and accumpcomp, but rarefaction seems to be
right for my data and for what I want to achieve... I just don’t know how I
can tell the differences between locations.



Thanks!



Stine

	[[alternative HTML version deleted]]