Ordination of community data — ord

Fits an unconstrained or constrained ordination to long community data and returns its scores as tidy tibbles. The ordination itself is computed by the vegan package, so results are identical to vegan's; tidydiversity adds the data handling, site covariates, tidy output and publication-quality plots (autoplot()).

Usage

ord_fit(
  data,
  site = site,
  species = species,
  abundance = NULL,
  method = c("pca", "ca", "dca", "pcoa", "nmds", "rda", "cca", "dbrda"),
  formula = NULL,
  env = NULL,
  transform = NULL,
  distance = "bray",
  scaling = c("species", "sites", "symmetric"),
  k = 2,
  permutations = 999,
  ...
)

Arguments

data: A data frame in long format, with one row per record.
site, species: <tidy-select> The column holding site identifiers and the column holding species names.
abundance: <tidy-select> The column holding counts. If NULL (the default), a column called abundance is used when present, otherwise every row counts as one individual.
method: Ordination method; see Details.
formula: A one-sided formula of site variables, such as ~ elevation + season. Required for constrained methods; for unconstrained methods the variables are fitted passively.
env: A data frame of site variables with a column named like the site column. By default, site variables are taken from data: every column that is constant within each site.
transform: Transformation of the community matrix before the analysis: "hellinger" (the default for PCA and RDA, which otherwise overweight abundant species; Legendre and Gallagher 2001), "chord", "log" (log(x) + 1 for x > 0), "sqrt", "pa" (presence-absence) or "none" (the default for the other methods).
distance: Dissimilarity for "pcoa", "nmds" and "dbrda", passed to vegan::vegdist(), such as "bray" (the default) or "jaccard".
scaling: For PCA, CA, PCoA and the constrained methods, which scores to scale (ter Braak and Smilauer 2012): "species" (the default, vegan scaling 2) preserves the correlations between species and variables; "sites" (scaling 1) preserves the distances between sites; "symmetric" (scaling 3) is a compromise.
k: For "nmds", the number of dimensions.
permutations: Number of permutations for the tests of passive variables.
...: Passed to the vegan function that fits the ordination.

Value

A td_ordination object: a list with tibbles sites (site scores joined to the site variables), species, vectors (environmental arrows), centroids (factor levels) and axes (eigenvalues and proportion of variation explained), plus stress for NMDS, envfit for passive variables and fit, the vegan object. Use tidy() and glance() to extract them and ord_test() for permutation tests.

Details

Unconstrained methods summarise the main gradients in species composition: principal components analysis ("pca"), correspondence analysis ("ca"), detrended correspondence analysis ("dca"), principal coordinates analysis ("pcoa") of a dissimilarity, and non-metric multidimensional scaling ("nmds"). Constrained methods find the gradients explained by environmental variables: redundancy analysis ("rda"), canonical correspondence analysis ("cca") and distance-based RDA ("dbrda"). Add Condition(x) to the formula for a partial analysis that removes the effect of x first.

For an unconstrained method, a formula fits the variables passively with vegan::envfit(), so they appear as supplementary arrows and centroids (as in CANOCO) without changing the ordination.

References

Legendre, P. and Gallagher, E. D. (2001). Ecologically meaningful transformations for ordination of species data. Oecologia, 129, 271-280. doi:10.1007/s004420100716

ter Braak, C. J. F. and Smilauer, P. (2012). Canoco reference manual and user's guide: software for ordination (version 5.0). Microcomputer Power, Ithaca, USA.

Oksanen, J. et al. vegan: Community Ecology Package. https://CRAN.R-project.org/package=vegan

Examples

# Tree-nesting ants: one sample per plot and subplot
ants <- png_tree_ants
ants$plot <- paste(ants$site, ants$subplot)

# Constrained: how much of the composition does elevation explain?
ord <- ord_fit(ants,
  site = plot, abundance = nests, method = "rda", formula = ~elevation
)
ord
#> RDA of 5 sites and 169 species
#> Transformation: hellinger
#> Variation explained: RDA1 63.3%, PC1 28.2%, PC2 4.5%
#> Constrained: 63.3% of the total
ord_test(ord)
#> 'nperm' >= set of all permutations: complete enumeration.
#> Set of permutations < 'minperm'. Generating entire set.
#> # A tibble: 2 x 5
#>   term        df variance statistic p.value
#>   <chr>    <dbl>    <dbl>     <dbl>   <dbl>
#> 1 Model        1    0.475      5.17  0.0333
#> 2 Residual     3    0.276     NA    NA     

# Unconstrained NMDS with elevation fitted passively
nmds <- ord_fit(ants,
  site = plot, abundance = nests, method = "nmds", formula = ~elevation
)
#> Warning: stress is (nearly) zero: you may have insufficient data
#> Warning: skipping half-change scaling: too few points below threshold
#> 'nperm' >= set of all permutations: complete enumeration.
#> Set of permutations < 'minperm'. Generating entire set.
#> 'nperm' >= set of all permutations: complete enumeration.
#> Set of permutations < 'minperm'. Generating entire set.
tidy(nmds)
#> # A tibble: 5 x 6
#>   plot      axis1   axis2 site   elevation subplot
#>   <chr>     <dbl>   <dbl> <chr>      <int> <chr>  
#> 1 Wanang B -0.393  0.250  Wanang       200 B      
#> 2 Wanang A -0.429 -0.0484 Wanang       200 A      
#> 3 Numba A  -0.117 -0.336  Numba        900 A      
#> 4 Yawan A   0.398  0.237  Yawan       1800 A      
#> 5 Yawan C   0.542 -0.103  Yawan       1800 C