1. Populations of bumble bees and other pollinators have declined over the past several decades due to numerous threats, including habitat loss and degradation. However, we can rarely investigate the role of resource loss due to a lack of detailed long-term records of forage plants and habitats.
2. We use 22-year repeated surveys of more than 262 sites located in grassland, forest, and wetland habitats across Illinois, USA to explore how the abundance and richness of bumble bee food plants have changed over the period of decline of the endangered rusty patched bumble bee (Bombus affinis).
3. We document a decline in abundance of bumble bee forage plants in forest understories, which our phenology analysis suggests provide the primary nectar and pollen sources for foundress queens in spring, a critical life stage in bumble bee demography. By contrast, the per-unit area abundance of food plants in primarily midsummer-flowering grassland and wetland habitats has not declined. However, the total area of grasslands has declined across the region resulting in a net loss of grassland resources.
4. The data for this project is provided alongside the scripts needed to reconduct analyses and generate figures.

# README

This directory contains all of the files needed to recreate the analysis and raw figure generation used in the manuscript "Long-term declines in early season forest plants critical to bumble bees."

By running all scripts in the order presented below, a user can first wrangle all raw data files needed (the "01" series within /data/data_wrangling), then conduct analysis and further number crunching ("02" series within /analysis), and  generate figures ("03" /figures/figure_wrangling). Lastly, two R Markdown files are provided that provide all of the necessary statistical output for the main text (and an HTML file) and the .docx supplemental file, respectively.  

For convenience, all outputs generated by scripts are provided in this directory so a user may skip directly to any analysis and figure generation scripts as they wish. In order to conduct a "clean run" of the project, simply remove files from the data/data_output, analyses/analyses_output, and figures/figure_partials directories. You will then only have the raw data (data/data_raw), the scripts described below, and the directory structure. You then need to run all scripts in order to yield a "clean" version of the two markdown files.

Detailed descriptions of scripts are provided at the top of each file.

01. data/data_wrangling

  01a_wrangling_plant_lists_binoms_only.R           
  01b_wrangling_usfws_plant_list.R             
  01c_wrangling_ctap_sites.R                    
  01d_wrangling_all_quadrats.R                       
  01e1_wrangling_landcover_extraction_ctap_sites.R    
  01e2_wrangling_landcover_extraction_ctap_sites.R    
  01f_NLCD_landcover_extraction.R
  01g_wrangling_plant_phenology.R                     
  01h_wrangling_rpbb_phenology.R                      
  01i_forest_and_shrub_data.R                       

02. analyses

  02a_analysis_cover_richness_change.R                       
  02b_analysis_cover_richness_by_habitat.R                   
  02c_analysis_phenology.R                                   
  02c01_analysis_phenology_all_species.R                     
  02d01_analysis_landscape_change.R                          
  02f_supplemental_repeated_analysis_different_plant_lists.R
  02g_proportion_by_species.R  
 
03. figures/figure_wrangling

  03a_figures_CTAP_site_maps.R            
  03b_figures_cover_richness_change.R     
  03c_figures_cover_richness_by_habitat.R
  03d_figures_phenology_v2.R              
  03f_figures_proportional_plants.R     
 
04. docs

  04a_maintext_results.Rmd
  supplements/04b_supplement_CTAP.Rmd
 
Additionally, there are a few files in `src` that cover functions used repeatedly or contain helpful things like figure color palettes, etc. These files are called as needed from the scripts above.

Two files, necessary to run analysis fully from raw-> complete were required to be removed for copyright issues. These files are bbna_midwestern_2019_10_31.csv and bbna_beespotter_to_find_caste.csv which are loaded within the initial data wrangling steps. The scripts should still be functional using the data_output intermediate files, but if these raw files are desired, please contact us and we can provide them through the proper channels.

Within data/data_raw, the following files can be found with brief descriptions.

  affinis_counties.csv: State and County that B. affinis is found in recent years
 
  CTAP_merged: a collection of files related to the CTAP program. The fields within these files are defined by the CTAP program, which can be found via references in main text.

  CTAP_nectar_plants: all plants within the CTAP database and their associated attributes. Species = plant species, PSID = plant species identifier, scientific name = scientific name, pollitiogent = what pollinators visit it, CHI.FLOWER.first = first date it flowers in the Chicago area, CHI.FLOWER.last = last date it flowers in the Chicago area, Origin = whether the plant is native to Illinois or note. Other fields not used in this study.

  nassTranslation.csv: a file of conversions between NASS crop categories as values and their categorical groupings. category = number, category2 = cover written value, new_cat = a higher grouping, new_cat2 = a yet even coarser grouping

  rpbb_historic_counties.csv: a list of all of the historic counties where B. affinis occurred. Coordinates are provided, but not used in this study.