extreme_limits

authors: S.R. Supp, T.A. Cormier, and C.H. Graham, modified from P.A. Beck

R Code to run the analyses for the hummingbird extreme limits paper. The code creates the models that investigate how Molt stage (Molt_model.Rmd), Weight (Weight_model.Rmd) and arrival in Arizona (Arrival_model.Rmd) are affected by resource availability (NDVI) and physiological demand (operative temperature).

Files

• .Rmd (Rmarkdown) files describe each of the analyses and .html files generated from them.
• The data sets (.rdata files) are kept on a personal computer, for questions, contact C.H. Graham.
• Functions (.r files) called by the code described in the different .Rmd files.

Requirements

• aplpack
• bbmle
• ggplot2
• knitr
• lme4
• lattice
• latticeExtra
• maptools
• ncdf
• RNetCDF
• raster
• sirad
• timeDate
• zoo

Data

• Mexico_CFSR_level1.rdata - stored in shared Dropbox
• Mexico_CFSR_level2.rdata - stored in shared Dropbox
• Mexico_SOT_level1.rdata - stored in shared Dropbox
• Mexico_SOT_level2.rdata - stored in shared Dropbox
• Mexico_SOT_level3.rdata - stored in shared Dropbox
• NDVI_df.rdata - stored in shared Dropbox
• Torpor_df.rdata - stored in shared Dropbox
• annual.anom.df.rdata - Winter range (Nov-Mar) climate data from the hummingbird site points
• site.dat.rdata - Summarized site data
• yearly.catch.stat.rdata - Summarized hummingbird catch data from Arizona, arrival dates as quantiles
• yearly.climate.rdata - Summarized climate data, as quantiles, for precip, NDVI, and operative temperature

R Scripts (functions and data preparation)

• Calculate_SOT_from_Met.r - contains standard operative temperature calculations
• calculate_diffuse_fraction.r - contains diffuse fraction of radiation calculations
• Calculate_SOT_from_CFSR_Met.r - makes the dataframe allmet, makes Mexico_SOT_level3.rdata, uses functions from the above two files (Calculate_SOT_from_Met.r, and calculate_diffuse_fraction.r) to calculate Te and Tes, makes some plots of the met and temperature data.
• NCEP_CFSR_data_extract_v5.r - a function to summarize extracted data (netcdf files) per day, sum/mean/max/min
• compare_years_extracted_CFSR_data.r - uses Mexico CFSR_level2.rdata and makes annual.anom.df.rdata (which is later used to make yearly.climate.rdata)
• Make_yearly_enviro_dat.r - code to make yearly.climate.rdata file from the summarized raw data (annual.anom.df.rdata), to be used in the models for the manuscript results
• Plot_conditional_effects_interaction_models.r - Used for the models in the Rmd files: code for plotting conditional effects of the interaction models, to plot the effect of x1 on y, in a 2-level regeression model that includes an interaction between x1 and x2, and random varying intercept
• plot_yearly_NDVI_cumprecip_series.r - code to plot NDVI and cumulative precipitation over the time series. Figure 3: NDVI, cumulative precipitation, and ENSO
• Model_evaluation_v4.r - code for evaluating mixed models and quantile regresstion for the models in the .Rmd files.
• defunct code
• data_prep_for_transfer_PSAB_only.r - IGNORE

Rmd Scripts (generate the main results)

• Arrival_model.Rmd - model of Arrival/catch observed in Arizona and how it covaries with energetic demand and resource availability in Mexican wintering grounds
• Molt_model.Rmd - model of Molt observed in Arizona and how it covaries with energetic demand and resource availability in Mexican wintering grounds
• Weight_model.Rmd - model of Weight observed in Arizona and how it covaries with energetic demand and resource availability in Mexican wintering grounds

Replicating the results from the raw data

To run, check the pathnames in each of the files, make sure you have the required libraries installed, and a copy of the .Rdata files.

1. Convert temperature variable in the raw data (Mexico_CFSR_level1.rdata and Mexico_CFSR_level2.rdata) into Operative Temperature (Calculate_SOT_from_CFSR_Met.r). This refers to the scripts Calculate_SOT_from_Met.R and calculate_diffuse_fraction.R. Save as allmet within Mexico_SOT_level3.rdata.
2. Input Mexico_CFSR_level2.rdata into compare_years_extracted_CFSR_data.r to create annual.anom.df.rdata
3. Input annual.anom.df.rdata into Make_yearly_enviro_dat.r. This will create yearly.climate.rdata, which is summarized from Te.
4. Run Arrival_model.Rmd, Molt_model.Rmd, Weight_model.Rmd, and plot_yearly_NDVI_cumprecip_series.r on the main datafiles to regenerate the main figures, tables, and results for the manuscript.

Contact T.A.C., S.G., or P.S.A.B. for questions related to the code that summarizes the remote sensing data.

contact S.R. Supp

email: sarah@weecology.org