Access the dataset on Dryad

This dataset is from the paper titled "Evolution of interspecific variation in marine larval dispersal kernels: the role of larval navigation ability", by: E. Lou Schlatter, Allison K. Shaw, Colleen T. Webb, Peter Buston, published in: Oikos, and is based on the model from the paper titled "The evolution of marine larval dispersal kernels in spatially structured habitats: analytical models, individual-based simulations, and comparisons with empirical estimates" by: Allison K Shaw, Cassidy C D'Aloia, Peter M Buston, published in: The American Naturalist.

Contents

KernelEvolution_Oikos.zip is a compressed file containing the following folders, files, and variables:

1. Folder files_runsims: contains Matlab code (.m and .mlx files) for running model simulation data used for Figures 1-2 in the paper. Contains files:
   • 1_Preparation.mlx: runs function to create simulation environment: 2 x 512 habitable grid, surrounded by uninhabitable space.
   • 2_Trajectories.mlx: runs functions to simulate evolutionary trajectories from 4 different initial displacement strategies, and for 3 levels of larval navigation (nmax=0, 1, and 2).
   • 3_NK_surfaces.mlx: generates a set of displacement strategies representing the entire range of possible distribution shapes. Then, for each element of the set, runs function to calculate nonviable habitat and kin competition costs in a population where all individuals share that strategy.
   • fn_create_env_bounded.m: function to create environment.
   • fn_create_initial_pop.m: function to generate population data that can be used to initialize a simulation.
   • fn_IBM_dispersal.m: function to run a simulation of the individual-based model.
   • fn_many_costben_sims.m: runs the function to calculate nonviable habitat cost and kin competition many times, once for each of a set of displacement strategies.
   • fn_one_costben_sim.m: function to calculate nonviable habitat and kin competition costs for a population in which all individuals share the same displacement strategy, by simulating dispersal of a single generation of offspring.
   • NK_kernel_set.mat: data file containing the set of displacement strategies generated in 3_NK_surfaces.mlx. Contains variable:
     • kernels: matrix of kernels (each row a kernel, each column a distance probability) representing the entire range of possible distribution shapes, used in 3_NK_surfaces.mlx.
2. Folder output_environments: contains Matlab data file (.mat file) for the simulation environment, with variables:
   • b = number of offspring per site in the environment
   • bvec = vector of offspring capacity of each viable site, 1...S
   • dists = array of distances between each site. Width of one site = one unit of distance.
   • E = 2D matrix of patches, marked as viable (1) and nonviable (0)
   • patches = vector of birth rates at each site
   • S = number of actual viable sites in the environment
   • sx = number of sites in the x-dimension of the environment
   • sy = number of sites in the y-dimension of the environment
   • via_ID = list of viable patches in E, marked by their index number
   • xcoord = list of x-coordinates for all patches in E
   • ycoord = list of y-coordinates for all patches in E
3. Folder output_simulations: contains Matlab data files (.mat files) for each individual-based simulation. Contents:
   • Folder NK_surfaces (data on kin competition and nonviable habitat costs for all displacement strategies in NK_kernel_set.mat. One file for each value of nmax = 0, 1, and 2.) Variables:
     • kernels = matrix of kernels used in simulation (each row a kernel, each column a distance probability)
     • out_fitness, out_kincost, out_mortcost = population-wide averages of fitness, kin competition cost, and nonviable habitat cost for each kernel
     • dist_fitness, dist_kincost, dist_mortcost = distance-specific values of fitness, kin competition cost, and nonviable habitat cost for each kernel
   • Folder trajectories (data on evolutionary trajectories for 4 initial displacement strategies x 3 values of nmax = 12 total simulations). Variables:
     • Output variables used in plotting:
       • dtime_avg = population mean of dispersal parameters at each timestep
       • kincomp_cost = kin competition cost at each distance and timestep
       • mortality_cost = nonviable habitat cost at each distance and timestep
     • Saved simulation parameters:
       • gflag = whether (1) or not (0) to show plots during simulation
       • eflag = which environment to use: 1=unbounded, 2=bounded, 3=reef, 4=heterogeneous
       • sx = number of sites in the x-dimension of the environment
       • sy = number of sites in the y-dimension of the environment
       • nbins_env = max number of dispersal bins the env can support (must be >= 2)
       • nbins = how many dispersal bins to actually use
       • nmax = maximum larval navigation distance (behavior)
       • G = number of total generations to simulate
       • b = offspring produced per individual
       • p = probability of surviving dispersal
       • del = fraction of dispersal probability to move during mutation
       • pop_init = initial population array (e.g., from previous simulation)
       • saveto_filepath = where to save output files
4. Folder files_plotsims: contains Matlab code (file 4_Plots.mlx) for plotting Figures 1-2 in the paper, using the above .mat files.