population and community properties

Metacommunity dynamics simulations for diatoms in Antarctic ponds

Abstract: 

We use metacommunity simulations to understand how local and regional community assembly dynamics influence the regional biodiversity patterns that we observe in the McMurdo Dry Valleys ecosystem. A metacommunity refers to a network of communities in an ecosystem that are connected to one another by the dispersal of biota among sites. For example, ponds in the McMurdo Dry Valleys share common diatom species that are likely dispersed among neighboring ponds by wind.

We have written a metacommunity model for the R programming language (MCSim, Sokol et al. 2015) to explore what biodiversity patterns look like for different metacommunity connectivity scenarios. This gives us the opportunity to model alternative hypotheses and explore how biodiversity metrics, such as beta-diversity, can be expected to respond to changing ecosystem connectivity for different sets of assumptions about the underlying dynamics that could potentially control biodiversity in the Dry Valleys. 

Dataset ID: 

8006

Associated Personnel: 

383

Short name: 

metacommunity dynamics

Data sources: 

Metacommunities Data INPUTs
Metacommunities Data OUTPUTS and code example

Methods: 

Field Sample Analysis and Model Approach

We are parameterizing metacommunity simulations by using observational data collected across various MCM core experiments to characterize local habitat characteristics and the composition of biotic communities in aquatic and terrestrial habitats. For example, we have modeled different metacommunity scenarios based on the diatom communities observed in microbial mats sampled in ponds on Ross Island and in the McMurdo Dry Valleys (observational data described in Sakaeva et al. 2016). For each pond, we have measured a suite of environmental variables (e.g., pH, conductivity, nutrient concentrations) and recorded GPS locations. Similar to the modeling approach described by Sokol et al. (2015), we use the pond GPS coordinates to create a network of ponds in 2-dimensional space for the metacommunity simulations. We use the first axis of a principal component analysis (PCA) of pond environmental variables to create a composite environmental variable, used to characterize the local habitat at each pond in a simulation scenario. Diatom counts from cores collected from microbial mats from each pond were used to set the initial metacommunity conditions for all simulation scenarios. Similarly, we are also using this approach to understand the importance of connectivity for biodiversity data collected in nematode and soil bacteria studies at the MCM LTER. 
 
Modeling code used:

MCSim for the R statistical language. Latest version available on Github at https://github.com/sokole/MCSim (version used for this work is https://github.com/sokole/MCSim/releases/tag/v0.4.1.9001)

Data Input Files

The input files used for the Metacommunities simulations, all the empirical data used are included in a Zip file (see above). Four csv files contain information on the communities densities. These include the following taxa (coded): Ataylor, Ccymat, Claevis, Cmolest, Fpellic, Holigot, Hantzsc, Habundan, Hamphiox, Hmuelle, Helongat, Humidoph, Hhyperaus, Haustral, Hparall, Lutico, Laustro, Ldolia, Lgauss, Llaeta, Levolut Lmuticop LmLevk, LmWest, Lpermutic, Lvermeu Matomu Mpermit Mvar1, Mueller Mmerid Msupra, Mperaus, Nseibig, Nshack Ngrega Nlineo, Ncommu, Nwesto, Ppapil, Slatis, Lunknown, Cheam1Input locations are in two files, including UTM coordinates as well as Lat/Lon coordinates. The lake index serves as linkage for chemical, physical and taxa data.

The pond or lakes chemistry empirical inputs include concentrations for Calcium, Iron, Chlorine, Bromine, Phosphates, Nitrites, Nitrates, Sulfates, Dissolved Organic Carbon and Amonia. Physical paramenters like the pH, the Conductivity and Temperature.

Data Output Files

Output files for example metacommunity simulations. Simulation results include species counts for each timestep from two different simulation scenarios. Example R code necessary to run these simulation scenarios is included

R-code.  Best refer to github link, but here  an example for Diatoms is included in the Output ZIP file.  Eric provided you with a nice tutorial on how to use the MCSim simulation, follow it at: http://rpubs.com/sokole/159425

Maintenance: 

Modeling code is managed in github, and the data output snapshots are stored here.  

Subscribe to RSS - population and community properties