This repository contains the research compendium for ther project “Evaluating the cumulative effects of global changes on the ecological communities of the Scotian Shelf Bioregion”. It contains all the code required to import, format, and integrate the data needed for the assessment, as well as the code used to perform the analyses, figures, and the project report.

Since data should not be stored on GitHub, no data are found on this repository; the code structuring all steps of the assessment must therefore be executed to replicate the assessment. It should however be noted that certain datasets are not accessible online due to data sharing agreements, such as the fisheries data from the Department of Fisheries and Oceans Canada. It is therefore necessary to contact us or the data holders to obtain these data. Metadata and bibliographic files associated with every dataset used are however available.

Please cite this research compendium as follows:

Beauchesne D (2023) Research compendium for the assessment of cumulative effects of global changes on the ecological communities of the Scotian Shelf Bioregion. Consulted on [AAAA-MM-JJ]. https://github.com/Ecosystem-Assessments/nceadfo.

This research compendium is structured as follows:

• data/: contains the data for the assessment.

  • aoi/: spatial data on the area of interest
  • basemap/: spatial data used to generate the maps of the assessment
  • cea_modules/: formatted data used to perform the cumulative effects assessment
  • config/: configuration files used to access and format raw data and parameters used throughout the project (e.g. spatial projection, colors, bounding box, etc.)
  • data-abiotic/: formatted abiotic data used for species distribution modelling
  • data-biotic/: data on species distribution in the area of interest
  • data-integrated/: integrated datasets
  • data-metaweb/: metaweb of species interactions in the area of interest
  • data-raw/: raw data used for the cumulative effects assessment
  • drivers/: formatted data on environmental drivers in the area of interest
  • eDrivers/: formatted data used for the eDrivers platform
  • format_modules/: formatted data used to perform the cumulative effects assessment as .RData files for ease of execution
  • grid/: study grid for the cumulative effects assessment
  • metadata/: metadata and contacts for the raw data used for the cumulative effects assessment

• docs/: contains the html version of the report of the cumulative effects assessment

• figures/: contains all the figures generated for the cumulative effects assessment

• man/: contains the documentation for all R functions that are part of the research compendium

• output/: contains all the outputs from the cumulative effects assessment

  • cea/: community-aggregated cumulative effects assessment for species-scale and network-scale assessments
  • cea_difference/: difference in cumulative effects between the temporal periods considered for the assessment
  • cea_km2/: assessment of cumulative effects per km^2 for all taxa considered
  • cea_network/: results for the network-scale cumulative effects assessment for all taxa
  • cea_species/: results for the species-scale cumulative effects assessment for all taxa
  • exposure/: exposure of taxa to cumulative effects
  • footprint/: species richness and cumulative drivers

• R/: contains R functions developped for the assessment

  • fig_name.R: scripts to generate figures
  • fnc_name.R: generic functions used throughout the research compendium
  • format_modules.R: script to prepare .RData files for assessmemnt
  • gather_name.R: scripts to gather information on data used for the assessment
  • get_name.R: scripts to access base data for the project
  • make_name.R scripts to prepare the modules used for the assessment
  • out_name.R: scripts to perform the assessment and extract summaries
  • pipeline.R: script that executes the entirety of the assessment. ⚠️ if run, this script will take multiple days to run. Also take into consideration that the network-scale cumulative effects assessment should be run on externally on clusters like those offered by Compute Canada as each taxa should take at least 10-20 hours to execute locally depending on your hardware. Furthermore, not all data are available online due to data sharing agreements.
  • render_report.R: script to render the assessment report

• report/: contains the R Markdown version of the assessment report

• DESCRIPTION: research compendium metadata (authors, date, dependencies, etc.)

• README.Rmd: description of research compendium

The nceadfo assessment heavily relies on the pipedat package. As stated in the package description:

pipedat is a R package that provides analytical pipelines to access, load, and format a variety of data from multiple sources programatically. The goal of pipedat is to enhance the capacity scientists, planners and the wider public to prepare and perform complex and reproducible ecosystem-scale assessments requiring the integration of multiple spatial datasets such as cumulative effects assessments in the context of ecosystem-based management, and Marxan analyses for the establishment of individual and networks of MPAs. In its current format, pipedat is strictly experimental and in development. We are however hoping to further develop this initiative in the hopes of greatly enhancing the efficiency, transparency and reproducibility of large-scale environmental assessments.

The development of pipedat was first thought off through nceadfo, yet it is not a part of it. Now that I have seen the potential usability of pipedat, and how it could be formalized, there is a desire to continue building it and making it better and simpler with the lessons learned through the nceadfo assessment. As such, there is a pipedat release (v0.0.1-nceadfo) that provides the specific version of the package that was used for the nceadfo assessment. There is also a static version of the package available directly in the nceadfo research compendium in data/pipedat-package/, which means that a user could clone this research compendium and install the package directly, like this (for iOS):

cd ./data/pipedat-package
R CMD INSTALL .

Clone this repo and execute the following commands to execute the whole assessment.

⚠️ execution time is very long, and not all raw data are accessible online. This research compendium is thus not fully reproducible due to hardware limitations and data sharing agreements.

R -e 'library(devtools);document()'
R CMD INSTALL .
R
source("_pipeline.R")