[R-sig-eco] adonis once again

[Jens Oldeland]

Dear list members,

I have a question again on adonis and whether the interpretation is 
correct. I must admit that I am a big fan of ANOSIM, I think it is 
pretty simple to explain, in contrast to adonis.

I have plant species dataset based on a literature review. I have 
calculated the morisita distance, because there were different numbers 
of plots in different habitats. Now, I used adonis in the following way:

adonis(formula = reho.dist ~ env$RICHNESS, strata = env$orgTab)

               Df SumsOfSqs MeanSqs F.Model      R2 Pr(>F)
env$RICHNESS   1    5.9068  5.9068  193.43 0.49416  0.001 ***
Residuals    198    6.0465  0.0305         0.50584
Total        199   11.9533                 1.00000
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

reho.dist is the distance matrix (morisita)
richness is the plant species richness per plot
strata are the numbers of the original tables hence coding for the 
publications.

do I interpret it correctly when I say; "the relationship between the 
distances between plots and plant species richness differs between the 
different strata" i.e. richness differs between strata??


likewise, then I can say that AREA is not different between the strata? 
(which should be the case)


adonis(formula = reho.dist ~ env$AREA, strata = env$orgTab)

            Df SumsOfSqs MeanSqs F.Model      R2 Pr(>F)
env$AREA    1    0.5007 0.50070  8.6565 0.04189   0.75
Residuals 198   11.4526 0.05784         0.95811
Total     199   11.9533                 1.00000


so finally, richness is not affected by area in this particular case?

sorry for all the text.

best regards,
Jens

http://www.biologie.uni-hamburg.de/bzf/fbda005/fbda005_e.htm