Effects of compositional heterogeneity and spatial autocorrelation on richness and diversity in simulated landscapes

Tardanico, Joseph 1 ; Hovestadt, Thomas1

Published Jul 25, 2024 on Dryad. https://doi.org/10.5061/dryad.mcvdnck5d

Data files

Jul 25, 2024 version files 941.77 MB

README.md

14.86 KB
Tardanico_Hovestadt2023_data.7z

941.75 MB

Abstract

Landscape structure plays a key role in mediating a variety of ecological processes affecting biodiversity patterns, however, its precise effects and the mechanisms underpinning them remain unclear. While the effects of landscape structure have been extensively investigated both empirically, and theoretically from a metapopulation perspective, the effects of spatial structure at the landscape scale remain poorly explored from a metacommunity perspective. Here, we attempt to address this gap using a spatially explicit, individual-based metacommunity model to explore the effects of landscape compositional heterogeneity and per se spatial configuration on diversity at the landscape and patch level via their influence on long-term community assembly processes. Our model simulates communities composed of lineages of annual, asexual organisms living, reproducing, dispersing, and competing within grid-based, fractal landscapes which vary in their magnitude of spatial environmental heterogeneity and their degree of spatial environmental autocorrelation. Communities are additionally subject to temporal environmental fluctuation and external immigration, allowing for turnover in community composition. We found that compositional heterogeneity and spatial autocorrelation had differing effects on richness and diversity and the landscape and patch scales. We also note a slight negative effect of compositional heterogeneity on the median total landscape population size. Landscape-level diversity was driven by community dissimilarity at the patch level and increased with greater heterogeneity, while landscape richness was largely the result of the short-term accumulation of immigrants and decreased with greater compositional heterogeneity. Both richness and diversity decreased in variance with greater compositional heterogeneity, indicating a reduction in community turnover over time. Patch-level richness and diversity patterns appeared to be driven by overall landscape richness and local mass effects, resulting in maximum patch-level richness and diversity at moderate levels of compositional heterogeneity and high spatial autocorrelation.

This dataset contains the results of a simulation experiment using an individual-based metacommunity simulation model. This dataset contains output from simulation scenarios covering 16 different combinations of compositional heterogeneity and spatial autocorrelation parameters. This archive additionally contains the parameter files used for the different scenarios, the shell scripts used to run the scenarios, a PDF user guide describing the use of the simulation program and its output data, and the R script used to create the figures in the journal article associated with this repository.

Description of the data and file structure

Data files consist of .txt files containing comma-separated values and can be opened directly via a text editor or imported into any data analysis software capable of importing text or CSV files. Data files for each scenario contain combined data from 30 replicate runs of the simulation. Data files contain landscape aggregated statistics data over time (ending in _trend), data on individual organisms at the final timestep of the simulation, and patch aggregated data calculated from data on individual organisms for the final timestep of the simulation (ending in _patches). This archive additionally contains the parameter files used for the different scenarios, the landscape files used for each of the replicates, and the shell scripts used to run the scenarios. Landscape files are .txt files containing semicolon-separated values representing environmental attributes for a patch. Each replicate run uses two landscape files. The precise files used are specified in the shell scripts. Parameter files consist of Julia code for a dictionary containing parameter values used for a simulation scenario. Parameter files are used as inputs for the simulation program. The parameter files used for a given scenario are specified in their respective shell scripts.

Files and naming schemes

Output Data Files

Output data files are contained in the /data folder. Data files are subject to regular naming schemes. The merged prefix indicates that the file contains combined data from multiple replicates. All output data files in this repository contain combined data from 30 replicate simulation runs and are thus named as such. The S1 prefix is a deprecated prefix originally intended to specify certain scenarios. The G prefix is followed by the value of the G simulation parameter. For example, G0.3 would indicate that this file contains data from the G=0.3 scenario. The _T prefix is followed by the number of time steps in a simulation run. This number is 10,000 in all scenarios in this study. The _AT designation indicates the Hurst index of the landscapes used in the simulation set. For example, a file with AT1_AT1 uses landscape files with a Hurst index of 1. Files with the patches suffix contain data aggregated by patch. Files with trend contain landscape aggregated data for each time step of a simulation run. Output data files with no suffix contain a census of individual organisms at the final time step of the simulation run.

Parameter Files

Parameter files are contained in the /parameter files folder. Parameter files contain the parameters needed to run the simulation scenarios used in this study. These files are intended to be imported by the simulation program at runtime and can be specified in a shell script. The parameters are contained in the form of a Julia dictionary. The purpose of each parameter is specified in comments in the files themselves. This information, as well as a template parameter file, is available in the GitHub repository for the simulation program: https://github.com/jtardanico/TardanicoHovestadt2023_Landscapes

Parameter files are named ParaDict_test and are followed by the value of G for its respective scenario (e.g. _g0.3). The lowAC (low autocorrelation) indicates a Hurst index of 0. The absence of the lowAC designation indicates a Hurst index of 1. The _trend ending indicates a fluctuating mean T. As this is used in all scenarios, all parameter files have this designation.

Shell Scripts

Shell scripts are contained in the /shell scripts folder. Shell script files are used to run the simulation program. Shell script files contain commands that specify the Julia code files and any parameters and file inputs needed to run them. Shells script file names contain the G scenario which they run, e.g. **_G0.3 for G=0.3. Shell scripts run both Hurst index=0 and Hurst index=1 scenarios, running 30 replicates of each.

Other files

This repository additionally contains the R script Tardanico_Hovestadt2023_figures.R which contains code used to create the figures used in the journal article associated with this repository. This R script will set its working directory to the directory the R script file is located in and load data files from a relative subdirectory path.

Data variables for data on individual organisms:

Variable	Explanation
ID	Defunct lineage identifier, replaced with LineageID
Replicate	Replicate number
Timestep	The time step of data output
H_t	Hurst index for T attribute
H_h	Hurst index for H attribute
alpha	Niche breadth trade-off parameter
clim_scen	Climate scenario (DEFUNCT)
gradient	Strength of compositional heterogeneity (G)
x,y	The x and y indices of the patch an organism is in
T_opt	T niche optimum
H_opt	H niche optimum
T_sd	T niche breadth
H_sd	H niche breadth
disp_l	Dispersal probability (Pdisp)
disp_g	Probability of dispersing via global dispersal (Pglobal)
fert_max	Intrinsic reproduction rate (R0), maximum expected offspring
LineageID	Identifier for organism lineage/species
Origin_x, Origin_y	Lineage patch of first appearance
Origin_time	Lineage time of first appearance
temp_t	Patch T attribute
habitat	Patch H attribute
trend	Fluctuation in T at a given timestep
mean_trend	Mean of T fluctuation,0 if no climate shift
precip_t	Defunct patch attribute from an early iteration of the program

Data variables for patch aggregated data (_patches):

Variable	Explanation
x,y	The x and y indices of a patch
Replicate	Replicate number
Timestep	The time step of data output
H_t	Hurst index for T attribute
H_h	Hurst index for H attribute
alpha	Niche breadth trade-off parameter
clim_scen	Climate scenario (DEFUNCT)
gradient	Strength of compositional heterogeneity (G)
pop	patch total population
richness	Patch lineage/species richness
shannon	Patch Shannon-Weiner diversity
simpson	Patch Simpson diversity
tdiff	Mean T_opt-patch T difference at current timestep
mt_tdiff	Mean T_opt-patch T difference on average over time
tft	Mean fitness for patch T at current timestep[^1]
mft	Mean fitness for patch average T over time[^1]
mfh	Mean fitness for patch H[^2]
time_fit	mean overall fitness at current timestep[^3]
mean_fit	mean overall fitness under patch average conditions over time[^3]
T_opt	patch mean T niche optimum
H_opt	patch mean H niche optimum
T_sd	patch mean T niche breadth
H_sd	patch mean H niche breadth
disp_l	Patch mean dispersal probability (Pdisp)
disp_g	Probability of dispersing via global dispersal (Pglobal)
fert	Patch mean intrinsic reproduction rate (R0), maximum expected offspring
temperature	Patch T attribute
habitat	Patch H attribute
trend	Fluctuation in T at a given timestep

[^1]: Fitness calculated for the T component of an organism's niche.

[^2]: Fitness calculated for the H component of an organism's niche.

[^3]: Combined product of fitness for T and H niche components.

Data variables for landscape time series (_trend):

Variable	Explanation
Replicate	Replicate number
Timestep	The time step of data output
clim_scen	Climate scenario (DEFUNCT)
Pop	Landscape total population
rich	Landscape total richness
simp	Landscape Simpson diversity index
shan	Landscape Shannon-Wiener diversity index
T_opt	Landscape mean T niche optimum
H_opt	Landscape mean H niche optimum
T_sd	Landscape mean T niche breadth (T_tol)
H_sd	Landscape mean H niche breadth (H_tol)
disp	Landscape mean dispersal probability (P_disp)
disp_g	Landscape mean global dispersal probability (P_global)
VT_opt	Landscape T_opt variance
VH_opt	Landscape H_opt variance
VT_sd	Landscape T_sd variance
VH_sd	Landscape H_sd variance
Vdisp	Landscape disp variance
Vdisp_g	Landscape disp_g variance
ft	Landscape mean fitness for patch T[^1]
fh	Landscape mean fitness for patch H[^2]
fit	Landscape mean fitness[^3]
tdiff	Landscape mean difference between T_opt and patch T at current timestep
avgtdiff	Landscape mean difference between T_opt and mean patch T
varft	Landscape variance in fitness for patch T[^1]
varfh	Landscape variance in fitness for patch H[^2]
varfit	Landscape variance in fitness[^3]
var_tdiff	Landscape variance in tdiff
var_avgtdiff	Landscape variance in avgtdiff
fert_max	Intrinsic reproduction rate (R0), maximum expected offspring
trend	Fluctuation in T at a given timestep
mean_trend	Mean of T fluctuation, 0 if no climate shift
grad	Strength of compositional heterogeneity (G)
H_t Hurst	Hurst index for T attribute (Spatial autocorrelation)
H_h Hurst	Hurst index for H attribute (Spatial autocorrelation)
alpha	Niche breadth trade-off parameter