Bayesian Trophic Position Calculation with Stable Isotopes

Estimates the trophic position of a consumer relative to a baseline species. It implements a Bayesian approach which combines an interface to the 'JAGS' MCMC library of 'rjags' and stable isotopes. Users are encouraged to test the package and send bugs and/or errors to [email protected]

tRophicPosition is a colaborative project of Chris Harrod, Andrew L. Jackson, Claudio Quezada-Romegialli and others. It consists of an R package incorporating a Bayesian model for the calculation of consumer trophic position using stable isotopes with one or two baselines. It uses the powerful approach of Markov Chain Monte Carlo simulations provided by JAGS and the statistical language R.

As of 29th of January 2018, we are releasing the version 0.7.5 of the package, the fourth release version. If you find any error, please send it to [email protected] and/or raise an issue in the GitHub page. Also you can also send a direct email to clquezada at harrodlab.net.

If you are interested in joining the tRophicPosition support group, do it at https://groups.google.com/d/forum/trophicposition-support

NEWS (29 January 2018)

• Added the function extractPredictiveData() to perform a posterior predictive model-checking procedure.
• Implemented code coverage and unit test.
• Improved the code (to make it clearer, considering width and the assignment operator).
• Changed some arguments (consumer instead of species, group instead of community and others) in loadIsotopeData(), extractIsotopeData() and other functions (while maintaining old arguments as compatibility)
• Added a a procedure for checking errors on priors arguments in jagsBayesianModel and related functions
• Removed MCMCvis from the Short guide to tRophicPosition vignette, as that package is no longer in CRAN.

NEWS (12 October 2017)

• Added the function fromParallelTP() to extract the data from parallel calculations (see the examples on GitHub).
• Implemented the Bhattacharrya Coefficient (from the package dispRity) when comparing two or more posterior distributions. As dispRity is only available through GitHub, when users have not installed the package we use the internal function bhat.coeff based on dispRity package.
• Improved credibilityIntervals(). Now it can group samples by colour, accept manual colours and legend position.
• Added two example datasets: Inari and Kilpis (Finnish_Lakes), and Roach, and including some examples in the GitHub page.

Installation

The stable version of tRophicPosition is hosted on CRAN, and is installed like any other package:

install.packages("tRophicPosition")

Remember to install JAGS for your platform as well.

The development version of tRophicPosition must be installed from GitHub. For this, install the package devtools, and then use the function install_github():

install.packages("devtools")
devtools::install_github("clquezada/tRophicPosition", build_vignettes = TRUE)

And then see the vignettes we have prepared to describe how to use the package:

browseVignettes("tRophicPosition")

Online vignettes

Can be found in GitHub here.

Credits

tRophicPosition relies on some code from coda and dispRity. At the moment dispRity is only available through GitHub, so it is not included in Suggests. tRophicPosition included the Bhattacharrya Coefficient (bhatt.coeff function from dispRity) and plotMCMC (from coda) for convenience.

Citation

While the paper describing tRophicPosition is not published (currently under revision at MEE), to cite the package in publications please use:

Claudio Quezada-Romegialli, Andrew L Jackson and Chris Harrod (2018). tRophicPosition: Bayesian Trophic Position Calculation with Stable Isotopes. R package version 0.7.5. https://cran.r-project.org/package=tRophicPosition doi:10.5281/zenodo.1161826

For previous releases please use the corresponding version and doi when citing:

• v0.7.5 (29 January 2018) 10.5281/zenodo.1161826

• v0.7.3 (12 October 2017) 10.5281/zenodo.1009571

• v0.7.0 (12 June 2017) 10.5281/zenodo.806048

Future releases and how to get support

You are encouraged to use tRophicPosition with your own data, test the package and see if there are any issues or problems. You can send your questions or commentaries to the google group tRophicPosition-support or directly to the email trophicposition-support at googlegroups.com. You can send your questions to http://stackexchange.com/ http://stackoverflow.com/ or even Facebook (stable isotope ecology group).

We are constantly working on future releases of tRophicPosition, so feedback is very much appreciated.

To DO

Things to do in tRophicPosition