Food webs for three burn severities after wildfire in the Eldorado National Forest, California

We provide data on ecological community responses to wildfire, collected three years post-fire, across three burn conditions (unburned, moderate severity and high severity) in the Eldorado National Forest, California. The data were collected with 19 sampling methods deployed across 27 sites (nine in each burn condition) used to estimate richness, body size, abundance and biomass density for 849 species (including 107 primary producers, 634 invertebrates, 94 vertebrates). The sampling methods are detailed in a companion data paper. To maximize transparency and ease of use we have made our data available in four formats: Raw, tidy, temporary and summary. Raw data is as close as possible to the form in which it was collected. As such, raw data is not ready for analysis. We have provided tidy versions of each raw data set that are ready for analysis. Temporary data files are included for transparency but are used to create summary data files, and not intended to be informative as stand alone products. Summary data files provide estimated biomass densities and network structure for each of our three burn categories. All tidy and summary data files are accompanied by metadata files describing their contents. We provide the R code to transform raw data files into tidy, temporary and summary files. We have included the code necessary to reproduce our work. We have broken the code into three R scripts: Biomass density estimation, resource filters assignment and network assembly.

Methods
The methods used to collect the data are described in an companion data paper. The methods used to tidy and summarize the data are in the included R scripts.