PERMANOVA effect size for each factor with adonis2

[siobhanschenk]

Hi,
Similarly to issue #704, I'm wondering if it would be possible to show the effect size for each term in a PERMANOVA with adonis2. I can see the individual terms with by="margin" but I would like to see the per-factor effect size without needing to run a marginal PERMANOVA.

I was able to see individual effect sizes before without by="margin" but not anymore. Not sure if this is R version thing (I'm running 4.4.1 currently) or a change with vegan itself (I'm running 2.6-8).

Thanks!

Current output with marginal PERMANOVA (for context if needed)

adonis2(formula = dune ~ Management + A1, data = dune.env, permutations = 99, by = "margin")

#            Df SumOfSqs      R2      F Pr(>F)   
# Management  3   1.1865 0.27600 2.4828   0.01 
# A1          1   0.4409 0.10256 2.7676   0.02 
# Residual   15   2.3895 0.55583                 
# Total      19   4.2990 1.00000                

[jarioksa]

In vegan prior to 2.6-8, adonis2 default was by = "terms". adonis2 (or adonis) never had a default by = "margin" , but the tests were sequential.

If you want to have statistics by = "margin", you must request tests by = "margin". There is no way of getting marginal effects without running an analysis using marginal effects. There never was.

See the release NEWS for vegan 2.6-8:

• adonis2() now defaults to running an omnibus test of the model (by = NULL) instead of a sequential test of model terms (by = "terms"). This makes adonis2() more consistent with the default for related ordination methods. See #677.

[siobhanschenk]

Thanks, adding by = "terms" gets me the output I wanted.

Including the output here for anyone reading this later.

data(dune)
data(dune.env)
adonis2(dune ~ Management*A1, data=dune.env, permutations=99)

adonis2(formula = dune ~ Management * A1, data = dune.env, permutations = 99)
Df SumOfSqs R2 F Pr(>F)
Model 7 2.4987 0.58122 2.3792 0.01
Residual 12 1.8004 0.41878
Total 19 4.2990 1.00000

adonis2(dune ~ Management*A1, data=dune.env, permutations=99, by="terms")

adonis2(formula = dune ~ Management * A1, data = dune.env, permutations = 99, by = "terms")
Df SumOfSqs R2 F Pr(>F)
Management 3 1.4686 0.34161 3.2629 0.01
A1 1 0.4409 0.10256 2.9387 0.03
Management:A1 3 0.5892 0.13705 1.3090 0.21
Residual 12 1.8004 0.41878
Total 19 4.2990 1.00000