Plot the results of a node test on the phylogeny

Draws the tree with the associations found by node_test() at each node, as in Figure 3 of Leibold et al. (2010). Blue circles are predictors and red triangles are biogeographic events, each with the names of the variables. Solid symbols remain significant after the false discovery rate, and open symbols are significant only as isolated tests.

Usage

plot_node_test(
  x,
  tree,
  test = c("max", "site", "joint"),
  type = c("phylogram", "fan"),
  alpha = 0.05,
  col = c("#0072B2", "#D55E00"),
  cex = 0.8,
  legend = TRUE,
  ...
)

Arguments

x

Result of node_test().

tree

The phylo object that gave the contrasts of node_contrasts().

test

Permutation test to show: "max" for the larger p-value of sites and species, "site" for sites alone, or "joint" for the permutation of the article. See node_test() for what each one tests.

type

"phylogram", as in the article, or "fan", for very large trees, with the names of the variables along the radii so that they stay clear of the species.

alpha

Significance level.

col

Colours of predictors and biogeographic events.

cex

Size of the text.

legend

Add a legend?

...

Other arguments of ape::plot.phylo().

Value

Invisibly, a data frame with one row per symbol, with node, set, variable, p, p_adjusted and corrected, which is TRUE for solid symbols.

Details

The article does not say which tests form a family for the correction. Here, the correction of Benjamini & Hochberg (1995) uses all nodes and variables of each set, so predictors and events are corrected separately.

As the smallest p-value is 1 / (n_perm + 1), solid symbols need many permutations. With few species, test = "max" may leave no symbols at all (see node_test()).

References

Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B, 57, 289–300. doi:10.1111/j.2517-6161.1995.tb02031.x

Leibold, M. A., Economo, E. P., & Peres-Neto, P. (2010). Metacommunity phylogenetics: separating the roles of environmental filters and historical biogeography. Ecology Letters, 13, 1290–1299. doi:10.1111/j.1461-0248.2010.01523.x

Examples

nodes <- node_contrasts(daphniids$tree)
set.seed(1)
out <- node_test(daphniids$comm, nodes, daphniids$env, daphniids$biogeography,
                 n_perm = 49)
plot_node_test(out, daphniids$tree, test = "site")

# \donttest{
# solid symbols need more permutations
out <- node_test(daphniids$comm, nodes, daphniids$env, daphniids$biogeography,
                 n_perm = 999)
plot_node_test(out, daphniids$tree, test = "site")

# }