Dear list,
I am on the initial stages of trying to analyze community data, to determine
the geographic scale where most variation in species composition occurs (and
how much variation there is at other scales). This was originally inspired by
the analysis in Underwood and Chapman (1996), where they conduct a hierarchical
analysis of variance of the abundance of single species with quadrats,
transects and shores arranged in a nested design (more specifically, they set:
factor 1, 500 -m intervals, 3 levels; factor 2, 100-m, 3 levels, and so on).
My data is multivariate so I can't use the exact same method. I'm thinking that
adonis() might work for my case, but there are two issues that have me confused:
First: what should I put on the right hand side of the formula?
Second: how I should arrange my data so that I can use something like "distance
interval" as strata, since this would be a distance matrix in itself (rather
than a vector)? In other words, if sample A and sample B are 10 m apart, and
they are 100 m away from sample C, the pair A-B would be in one stratum, but
the pairs A-C and B-C would be in another.
Perhaps I'm going down the wrong path all together?? This is all rather new for
me, and I will appreciate any pointers!
(In case you are wondering why I'm not using a Mantel correlogram, it is
because there are gaps of different sizes between my samples, so I would not be
able to generate a continuous function of spatial dependence versus distance).
Many thanks in advance!
Genoveva Gonzalez Mirelis, Scientist
Institute of Marine Research
Nordnesgaten 50
5005 Bergen, Norway
New phone number!! +47 55906504
[[alternative HTML version deleted]]
_______________________________________________
R-sig-ecology mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
