Fast Spatial Species Accumulation Curves with C++ Performance
The spacc package computes species accumulation curves and diversity metrics. It supports both classical random-order SACs (method = "random") and spatially-explicit methods that respect the spatial arrangement of sites using nearest-neighbor algorithms. All methods use a C++ backend (Rcpp/RcppParallel) for speed.
Quick Start
library(spacc)
coords <- data.frame(x = runif(100), y = runif(100))
species <- matrix(rpois(100 * 50, 2), nrow = 100)
# Spatial accumulation curve
sac <- spacc(species, coords, method = "knn", n_seeds = 50)
plot(sac)
# Diversity partitioning
part <- diversityPartition(species, q = c(0, 1, 2))
print(part)Statement of Need
Species accumulation curves are fundamental tools for understanding sampling completeness and comparing biodiversity. spacc computes both classical random-order SACs and spatially-explicit SACs that respect the spatial arrangement of sites. Comparing the two reveals how spatial structure affects observed diversity. Applications include:
- biodiversity assessment for spatially structured surveys,
- community ecology and scale-dependent diversity patterns,
- invasion modeling with wavefront and directional methods,
- conservation planning to identify high-diversity areas.
Performance
spacc automatically selects the fastest backend for nearest-neighbor queries:
| Backend | Distance | Method | When |
|---|---|---|---|
| k-d tree (nanoflann) | Euclidean | O(log n) per query | >500 sites (auto) |
| Ball tree (custom) | Haversine | O(log n) per query | >500 sites + distance = "haversine"
|
| Exact (brute-force) | Any | O(n) per query | ≤500 sites (auto) |
Override with backend = "kdtree" or backend = "exact". The ball tree uses haversine distance natively — no coordinate projection needed for geographic data.
Features
Accumulation Curves
-
spacc(): Spatial accumulation with 7 methods (knn,kncn,random,radius,gaussian,cone,collector), grouped accumulation (groups), and spatiotemporal distance (time,w_space,w_time) -
spaccHill(): Hill number accumulation (q = 0, 1, 2) -
spaccBeta(): Beta diversity with turnover/nestedness partitioning (Baselga 2010) -
spaccCoverage(): Coverage-based rarefaction (Chao & Jost 2012) -
spaccPhylo(): Phylogenetic diversity accumulation (MPD, MNTD) -
spaccFunc(): Functional diversity accumulation (FDis, FRic) -
wavefront(): Expanding radius accumulation for invasion modeling -
distanceDecay(): Species richness vs distance from focal points
Diversity
-
alphaDiversity(): Per-site Hill numbers (local diversity) -
gammaDiversity(): Pooled Hill numbers (regional diversity) -
diversityPartition(): Alpha-beta-gamma decomposition (Jost 2007) -
spaccMetrics(): Per-site accumulation metrics (slope, AUC, half-richness)
All diversity and accumulation functions support spatial mapping via map = TRUE, plot(x, type = "map"), and as_sf() for sf/GIS integration.
Analysis
-
extrapolate(): Fit asymptotic models (5 options) -
compare(): Statistical comparison between curves -
rarefy(): Rarefaction to common sampling effort -
subsample(): Spatial subsampling of sites -
coleman(),mao_tau(),spatialRarefaction(): Analytical (non-simulation) methods -
distances(): Pre-compute distance matrices
Usage Examples
Spatial Accumulation
coords <- data.frame(x = runif(100), y = runif(100))
species <- matrix(rbinom(100 * 50, 1, 0.3), nrow = 100)
sac <- spacc(species, coords, method = "knn", n_seeds = 100)
plot(sac)
# Compare spatial vs random
sac_random <- spacc(species, coords, method = "random", n_seeds = 100)
comp <- compare(sac, sac_random)
plot(comp)Grouped Accumulation
# Compare native vs alien species (same spatial ordering)
status <- ifelse(grepl("alien", colnames(species)), "alien", "native")
sac_grouped <- spacc(species, coords, groups = status, seed = 42)
plot(sac_grouped) # Overlaid curves per group
plot(sac_grouped, facet = TRUE) # Faceted panels
# Manual grouping with c()
sac_a <- spacc(species_a, coords, seed = 42)
sac_b <- spacc(species_b, coords, seed = 42)
plot(c(native = sac_a, alien = sac_b))Spatiotemporal Accumulation
# Sites sampled across years — weighted space-time distance
sac_st <- spacc(species, coords, method = "knn",
time = site_years, w_space = 1, w_time = 0.5)
plot(sac_st)
# Emphasize temporal proximity over spatial
sac_st2 <- spacc(species, coords, method = "knn",
time = site_years, w_space = 0.2, w_time = 1)Diversity Partitioning
species <- matrix(rpois(100 * 50, 2), nrow = 100)
part <- diversityPartition(species, q = c(0, 1, 2))
part
#> Alpha-Beta-Gamma Diversity Partitioning
#> 100 sites, 50 species
#>
#> q alpha beta gamma
#> 0 12.45 4.01 50
#> 1 8.23 3.12 25.68
#> 2 6.54 2.87 18.77
# Per-site alpha with spatial mapping
alpha <- alphaDiversity(species, q = c(0, 1, 2), coords = coords)
plot(alpha, type = "map", q = 0)Coverage-Based Rarefaction
cov <- spaccCoverage(species, coords, n_seeds = 50, map = TRUE)
plot(cov, xaxis = "coverage")
plot(cov, type = "map", metric = "final_coverage")Phylogenetic & Functional Diversity
library(ape)
tree <- rtree(50)
colnames(species) <- tree$tip.label
phylo <- spaccPhylo(species, coords, tree,
metric = c("mpd", "mntd"), map = TRUE)
plot(phylo, type = "map", metric = "mpd")
traits <- matrix(rnorm(50 * 3), nrow = 50)
rownames(traits) <- colnames(species)
func <- spaccFunc(species, coords, traits,
metric = c("fdis", "fric"), map = TRUE)
plot(func, type = "map", metric = "fdis")Spatial Mapping with sf
# Convert any diversity measure to sf for GIS integration
metrics <- spaccMetrics(species, coords,
metrics = c("slope_10", "half_richness", "auc"))
metrics_sf <- as_sf(metrics, crs = 32631)
# Works for all diversity types
hill_sf <- as_sf(hill, crs = 32631)
beta_sf <- as_sf(beta, crs = 32631)Support
“Software is like sex: it’s better when it’s free.” — Linus Torvalds
I’m a PhD student who builds R packages in my free time because I believe good tools should be free and open. I started these projects for my own work and figured others might find them useful too.
If this package saved you some time, buying me a coffee is a nice way to say thanks. It helps with my coffee addiction.
