sdm_sim: Simulate single species dispersal dynamics using the BAM framework.
Source:R/sdm_sim.R
sdm_sim.Rdsdm_sim: Simulate single species dispersal dynamics using the BAM framework.
Usage
sdm_sim(
set_A,
set_M,
initial_points,
nsteps,
stochastic_dispersal = TRUE,
disp_prop2_suitability = TRUE,
disper_prop = 0.5,
progress_bar = TRUE
)Arguments
- set_A
A setA object returned by the function
model2sparse- set_M
A setM object containing the adjacency matrix of the study area. See
adj_mat- initial_points
A sparse vector returned by the function
occs2sparse- nsteps
Number of steps to run the simulation
- stochastic_dispersal
Logical. If dispersal depends on a probability of visiting neighbor cells (Moore neighborhood).
- disp_prop2_suitability
Logical. If probability of dispersal is proportional to the suitability of reachable cells. The proportional value must be declared in the parameter `disper_prop`.
- disper_prop
Probability of dispersal to reachable cells.
- progress_bar
Show progress bar
Value
An object of class bam with simulation results.
The simulation are stored in the sdm_sim slot (a list of sparse matrices).
Details
The model is cellular automata where the occupied area of a species at time \(t+1\) is estimated by the multiplication of two binary matrices: one matrix represents movements (M), another abiotic -niche- tolerances (A) (Soberon and Osorio-Olvera, 2022). $$\mathbf{G}_j(t+1) =\mathbf{A}_j(t)\mathbf{M}_j \mathbf{G}_j(t)$$ The equation describes a very simple process: To find the occupied patches in \(t+1\) start with those occupied at time \(t\) denoted by \(\mathbf{G}_j(t)\), allow the individuals to disperse among adjacent patches, as defined by \(\mathbf{M}_j\), then remove individuals from patches that are unsuitable, as defined by \(\mathbf{A}_j(t)\).
References
Soberón J, Osorio-Olvera L (2022). “A Dynamic Theory of the Area of Distribution.” doi:10.48550/ARXIV.2212.06308 , https://arxiv.org/abs/2212.06308. .
Examples
# \donttest{
model_path <- system.file("extdata/Lepus_californicus_cont.tif",
package = "bamm")
model <- raster::raster(model_path)
sparse_mod <- bamm::model2sparse(model,threshold=0.05)
adj_mod <- bamm::adj_mat(sparse_mod,ngbs=1)
occs_lep_cal <- data.frame(longitude = c(-110.08880,
-98.89638),
latitude = c(30.43455,
25.19919))
occs_sparse <- bamm::occs2sparse(modelsparse = sparse_mod,
occs = occs_lep_cal)
sdm_lep_cal <- bamm::sdm_sim(set_A = sparse_mod,
set_M = adj_mod,
initial_points = occs_sparse,
nsteps = 10,
stochastic_dispersal = TRUE,
disp_prop2_suitability=TRUE,
disper_prop=0.5,
progress_bar=TRUE)
#>
|
| | 0%
|
|====== | 9%
|
|============= | 18%
|
|=================== | 27%
|
|========================= | 36%
|
|================================ | 45%
|
|====================================== | 55%
|
|============================================= | 64%
|
|=================================================== | 73%
|
|========================================================= | 82%
|
|================================================================ | 91%
|
|======================================================================| 100%
sim_res <- bamm::sim2Raster(sdm_lep_cal)
raster::plot(sim_res)
# }