10 Neutral Landscape Models (NLMs)

You will have to run these once to install since these packages are currently not available from CRAN:

remotes::install_github(“ropensci/NLMR”) devtools::install_github(“ropensci/landscapetools”)

library(raster)
library(NLMR)
library(ggplot2)
library(dplyr)
library(landscapetools)
library(purrr)
library(tibble)
library(gridExtra)

10.1 Random

  1. Random landscape raster drawn from a uniform distribution from 0 to 1

  • Null baseline for comparing spatial structure to

  • Simulating the influence of random disturbances or habitat patterns

nlm1 <- NLMR::nlm_random(100,100)
show_landscape(nlm1) + ggtitle("Random Landscape")

10.2 Gradient

  1. Planar gradient landscape raster drawn. Direction of gradient is random by default or can be set manually in degrees as I have here. Useful for modeling environmental gradients like elevation or temperature. You can also do an edge gradient where the peak is in the center of the raster.
nlm2 <- nlm_planargradient(ncol = 100, nrow = 100, direction = 90)
show_landscape(nlm2) + ggtitle("Planar Gradient")

10.3 Mosiac tessellation

  1. Generates a tesselated landscape. This creates a realistic looking patchy landscape with discrete regions:
nlm3 <- nlm_mosaictess(ncol = 100, nrow = 100, germs = 25)
show_landscape(nlm3) + ggtitle("Mosaic Tessellation")

10.4 Simulating across a gradient

  1. You can also simulate across some parameter values as in some of the ppt examples. This uses yet another simulation called the midpoint displacement neutral landscape model. This is a fractal based algorithm borrowed from generating random terrain in computer graphics/games.
simulate_landscape <- function(roughness, weighting){
  nlm_mpd(ncol = 99,
          nrow = 99,
          roughness = roughness,
          rescale = TRUE) %>%
    util_classify(weighting = weighting)
}

# Parameter combinations
param_df <- expand.grid(
  roughness = c(0.2, 0.9),
  weighting = I(list(c(0.2, 0.8), c(0.2, 0.3, 0.5)))
) %>%
  as_tibble()

# Generate landscapes
nlm_list <- param_df %>% pmap(simulate_landscape)


# Create individual plots
nlm_plots <- lapply(seq_along(nlm_list), function(i) {
  ggplot() +
    geom_raster(data = rasterToPoints(nlm_list[[i]]) %>% as.data.frame(),
                aes(x = x, y = y, fill = factor(layer))) +
    scale_fill_brewer(palette = "Set2") +
    theme_void() +
    ggtitle(paste("Roughness:", param_df$roughness[i],
                  "\nWeighting:", paste(param_df$weighting[[i]], collapse = ", ")))
})

# Arrange and display in a grid
grid.arrange(grobs = nlm_plots, ncol = 2)