Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
##################################
############ VERSION #############
##################################
v.00001.2023-04-04
##################################
###### GENERAL INFORMATION #######
##################################
1. Title of Dataset:
Data from: The fitness landscape of a community of Darwin’s finches
2. Author Information
Corresponding Investigator
Name: Marc-Olivier Beausoleil
Institution: McGill University
Email: marc-olivier.beausoleil@mail.mcgill.ca
Co-investigators
Name: ------------------- Institution: -------------------------
Paola L. Carrión McGill University
Jeffrey Podos University of Massachusetts Amherst
Carlos Camacho Estación Biológica de Doñana—CSIC
Julio Rabadán-González Observation.org
Roxanne Richard McGill University
Kristen Lalla McGill University
Joost A. M. Raeymaekers Nord University
Sarah A. Knutie University of Connecticut
Luis F. De León University of Massachusetts Boston
Jaime A. Chaves San Francisco State University and Universidad San Francisco de Quito
Dale H. Clayton University of Utah
Jennifer A. H. Koop Northern Illinois University
Diana M. T. Sharpe Harvard University
Kiyoko M. Gotanda Brock University, University of Cambridge,
Université de Sherbrooke, and McGill University
Sarah K. Huber Virginia Institute of Marine Science
Rowan D. H. Barrett McGill University
Andrew P. Hendry McGill University
3. Date of data collection:
2003-2020
4. Geographic location of data collection:
El Garrapatero, Santa Cruz Island, Galápagos, Ecuador (0°41'22.9" S, 90°13'19.7" W)
5. Funding sources that supported the collection of the data:
Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant and a Canada Research Chair (A.P.H.)
University of Cambridge (Clare Hall, Christ's College, and Newnham College),
British Ornithologists Union (K.M.G.)
European Society for Evolutionary Biology Godfrey Hewitt Mobility Award (K.M.G.)
NSERC Vanier and Banting Fellowships, and an FQRNT Postdoctoral Fellowship. (K.M.G.)
Field work was supported by Earthwatch in 2014 and 2015.
NSERC Discovery Grant and Canada Research Chair (R.D.H.B.)
NSERC Canada Graduate Scholarship, Biodiversity, Ecosystem Services and Sustainability NSERC CREATE (M.-O.B.)
Fonds de recherche du Québec, Nature et technologies Scholarship (208545) (M.-O.B.)
National Geographic Young Explorers Grants grant (#WW-170ER-17) (M.-O.B.)
Awards given by the Quebec Centre for Biodiversity Science (M.-O.B.)
6. Keywords
Darwin’s finches, Geospiza,
Adaptive landscapes,
Speciation,
Adaptive radiation,
Galápagos Santa Cruz,
Ecological theory,
fitness landscape
7. Recommended citation for this dataset:
Beausoleil, Marc-Olivier; Carrión, Paola L.; Podos, Jeffrey; Camacho, Carlos; Rabadán-González, Julio; Richard, Roxanne; Lalla, Kristen; Raeymaekers, Joost A. M.; Knutie, Sarah A.; De León, Luis F.; Chaves, Jaime A.; Clayton, Dale H.; Koop, Jennifer A. H.; Sharpe, Diana M. T.; Gotanda, Kiyoko M.; Huber, Sarah K.; Barrett, Rowan D. H.; Hendry, Andrew P., 2023, "The fitness landscape of a community of Darwin’s finches", https://doi.org/10.5683/SP3/0YIWSE, Borealis.
##################################
###### DATA & FILE OVERVIEW ######
##################################
1. Description of dataset and scripts
The data files are described using the folder structure below
- adaptive.landscapes.finches.zip contains ./data, ./output, ./script folders and LICENSE file.
The dataset and script were developed to
1. Estimate the fitness landscape for Darwin's ground finch species (Geospiza spp.)
at El Garrapatero over 2003 to 2020.
2. Consider theoretical expectations and previous empirical assertions regarding the topology of fitness
and adaptive landscapes.
We used the data from our long-term monitoring site El Garrapatero on Santa Cruz in the Galápagos, Ecuador.
We calculated lifespan as a fitness proxi from our recapture data to construct a fitness and adaptive landscape
using the beak length and depth.
The scripts and data and for the R language (R Core Team 2023; R version, 4.2.1 (Funny-Looking Kid)).
2. Folder content structure and file List:
- Use the script `00_a.runall.R` to run the scripts in the intended order using the folder structure below.
- Note:
- Folders are noted with "./folder_name"
- File are shown with an extension such as:
- .csv: Comma separated values
- .gpkg: GeoPackage (read with `sf::st_read()` function)
- .gif: Graphics Interchange Format
- .json: JavaScript Object Notation
- .pdf: Portable Document Format
- .png: Portable Network Graphics
- .R: R script
- .RDS: Stores a singe R object (read with `base::readRDS()` function)
- .RData: Stores multiple R object (read with `base::load()` function)
- .txt: text file
- .Rproj: project file
```
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
├── README # This file!
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
├── ./data
│ ├── bird.data.RData and bird.data.csv # Main dataset for capture-recapture of birds at
│ │ # El Garrapatero, Santa Cruz Island.
│ ├── ./climate
│ │ ├── climate_puerto-ayora_V.XXXX-XX-XX.csv # Meteorological station data (precipitation)
│ │ ├── ./ENSO
│ │ │ └── oni.ascii_V.XXXX-XX-XX.txt # Oceanic Niño Index
│ │ └── ./Kleindorfer.et.al._2006
│ │ ├── H.L.lands.fig3.Kleindorfer.2006.png # figure 3 from Kleindorfer et. al. 2006
│ │ ├── H.L.lands.fig3.Kleindorfer.2006.csv # Extracted data from figure 3 from
│ │ # Kleindorfer et. al. 2006 using WebPlotDigitizer
│ │ └── Kleindorfer.et.al.2006.WebPlotDigitizer.json # WebPlotDigitizer data file
│ ├── ./maps
│ ├── Galapagos.EG.nets.csv # Approximate mistnet GPS at El Garrapatero
│ ├── galapagos.gis.gpkg # Geopackage with 2 layers:
│ │ # 1. polygon of all islands,
│ │ # 2. line string of roads on all islands
│ └── map.back.gpkg # Geopackage of polygons for background of maps
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
├── ./scripts
│ ├── 00_a.runall.R # Script to source all the other scripts in order
│ ├── 0.0_initialize.R # Contains packages and functions, (session info)
│ ├── ./adaptive_landscapes
│ │ ├── adaptive.landscape.calculation.R # Calculates the 2d-adaptive landscape for Darwin's
│ │ │ # finches based on the fitness landscape
│ │ └── gif.walking.adaptive.landscape.R # Generate a GIF on how to compute adaptive landscapes
│ ├── ./climate
│ │ └── climate.analysis.R # Compute sum precipitation per month for all years in
│ │ # study and climate high- lowland G. fuliginosa image
│ ├── ./CMR
│ │ └── cmr.test.R # Capture-Mark-Recapture calculation on the data
│ ├── ./fitness_surfaces
│ │ ├── fit.land.analysis.R # Prepare the data for GAM and prints tables
│ │ ├── fit.land.analysis_2.R # GAM fitting
│ │ ├── fit.land.analysis_3.R # Plot fitness landscape, euclidean distance on
│ │ │ # landscape, find fitness maxima, prospective
│ │ │ # selection, tables of walk on the fitness landscape
│ │ ├── fit.land.analysis_4.R # Plot 3D fitness landscape for peak visualization.
│ │ ├── fit.land.walk.R # Figures of the walk on the fitness landscape
│ │ │ # (showing valleys and peaks)
│ │ ├── plot_fitness_no_model.R # Plot fitness landscape using a smoothed average of
│ │ │ # raw fitness values across all phenotypic
│ │ │ # combinations for raw values.
│ │ └── projection_pursuit_regression.R # Projection pursuit regression (PPR)
│ ├── ./functions
│ │ ├── 00_info_session.R # Prints packages used and operating system in file
│ │ └── 0.1_misc.R # More helper functions
│ │ ├── cor_fun.R # Used for ggpairs in fit.land.analysis.R
│ │ ├── euclidean_dist.R # Used to calculate the euclidean distance and get
│ │ │ # the angle between vectors in fit.land.analysis_3.R
│ │ ├── intersect.lines.R # Draw perpendicular lines in a plot for prospective
│ │ │ # selection in fit.land.analysis_3.R
│ │ ├── isolate_surface.R # Helper function to find fitness maxima on a surface
│ │ │ # in R in fit.land.analysis_3.R
│ │ └── plot_gams.R # Plots gam output as surface and collects data of the
│ │ # surface and its standard error
│ ├── ./maps
│ │ └── Maps sites Galapagos drawing net and sites.R # Generates maps of islands, and mistnets locations
│ └── Repeat.R # Repeatability measures
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
├── ./output
│ ├── ./adaptive_land.animation
│ │ └── ./_fitplot.gif # Contains images to construct a GIF on how to
│ │ │ # calcualte an adaptive landscape
| │ └── [...] Miscellaneous files # Files for .gif construction
│ ├── ./data.out
│ │ ├── ./adpt_land.sp
│ │ │ ├── all.sp_EG_fortis_small.RDS # Adaptive landscape empirical data
│ │ │ └── all.sp_EG.RDS # Adaptive landscape simulated data
│ │ ├── ./fitness.landscape.data
│ │ │ ├── all.local.max_pca.RDS # Phenotypic position and value of fitness local
│ │ │ │ # maxima for all species (PCA traits)
│ │ │ ├── all.local.max.RDS # Phenotypic position and value of fitness local
│ │ │ │ # maxima for all species (beak traits)
│ │ │ ├── fitness.summary.RDS # Combination of all.local.max and w.data
│ │ │ ├── min.fit.sca.comp.RDS # Walking the fitness surface fitness-peak between
│ │ │ │ # species: Peak to peak measurements
│ │ │ ├── summary.traits.finches.RDS # Summary of the beak and body traits YEARLY (in
│ │ │ │ # fit.land.analysis.R)
│ │ │ └── w.data.RDS # Average and predicted fitness at trait average (and
│ │ │ # average + sd of beak traits)
│ │ ├── ./ggplot_data
│ │ │ └── ggp.fit.no.model.RDS # GGplot data for fitness landscape without model
│ │ ├── ./landscape_data
│ │ ├── my.persp.data_pca.RDS # Data from plot.gam.cust() function using GAM with
│ │ │ # PCA traits
│ │ └── my.persp.data.RDS # Data from plot.gam.cust() function using GAM with
│ │ # beak traits
│ │ ├── ./tables
│ │ ├── between.species.pheno.eucli.csv # Euclidean distance of each mode for pairwise
│ │ │ # mean phenotypes (lower portion) and for pairwise
│ │ │ # G-transformed mean phenotypes (upper portion).
│ │ ├── euclidean.csv # Table of Euclidean distances of population mean
│ │ │ # to the position of the peaks on the fitness
│ │ ├── fitness.diff.within.species.adaptive.walk.csv # % fitness loss from fitness peak and fitness at
│ │ │ # phenotypic mean
│ │ ├── fitness.landscape.features.csv # % fitness loss from fitness peak and fitness at
│ │ │ # phenotypic mean
│ │ ├── pca.loadings_EG.csv # Principal component analysis loadings
│ │ ├── recapnumber_EG.csv # Recapture data at El Garrapatero
│ │ └── sampling.effort.sampling.range_EG.csv # Sampling effort for the ground finches studied
│ │ # with minimum and maximum date of visit and
│ │ # number of individual of each species captured.
│ ├── ./fitland_interactive
│ │ └── plotly data, interactive graph and 3D plots # Various files to plot landscape in 3D
│ ├── ./images
│ │ ├── ./climate
│ │ │ ├── fuliginosa.beak.length.precipitation.png # Precipitation pattern associated with High-
│ │ │ │ # lowlands for G. fuliginosa
│ │ │ └── precipitation.Puerto.ayora.el.nino.la.nina.png # Precipitation pattern for the length of the
│ │ │ # study with ONI El Nino and La Nina
│ │ ├── ./data_exploration
│ │ │ ├── density.traits_and_pca_EG.png # Density of birds for beak and PC traits
│ │ │ ├── effort.plot.per.species_EG.png # Number of captures of Darwin's finches for
│ │ │ │ # the study system
│ │ │ └── pairs.traits_EG.png # Pairs plot for all beak, body and PCA traits
│ │ ├── ./landscape_plots
│ │ │ ├── adaptiveland_empirical_EG.png # Compared simulated-vs-empirical adaptive
# landscapes. Phenotypic dist. G. fortis small
│ │ │ ├── fit.surf.adapt.land.simulations_complete_single_ggpt_EG.[pdf, png] # Fitness landscape
│ │ │ │ # without model, with GAM
│ │ │ │ # and adaptive landscape
│ │ │ │ # for all finches.
│ │ │ ├── fit.surf.PCA_ggpt_EG.png # Fitness landscape with PCA traits
│ │ │ ├── fit.surf.rev1_ggpt_EG.png # Fitness landscape with more contour levels
│ │ │ ├── fit.surf.walk_ggpt_EG.png # Walking on the fitness landscape and valleys
│ │ │ │ # between peaks of different species and between
│ │ │ │ # peak and phenotypic average within species
│ │ │ ├── fit.surf.yr.chg_ggpt_EG.png # Fitness landscape with phenotypic average in
│ │ │ │ # every year
│ │ │ ├── fitplot.3D_select_EG.png # GAM select = TRUE or FALSE.
│ │ │ ├── fitplot.average.change_se_ggplt_EG.png # Standard error of fitness landscape
│ │ │ ├── fitplot.no.model_ggplt_EG.png # Fitness landscapes without model.
│ │ │ ├── plot.3D.adaptive.land_EG.png # 3D adaptive landscape plot
│ │ │ ├── plot.3D.fit_land_EG.png # 3D fitness landscape plot
│ │ │ ├── plot.3D.fit_land_oneleg_EG.png # 3D fitness landscape plot (1 legend)
│ │ │ └── ppr_ggpt_EG.png # Projection pursuit regression
│ │ ├── ./maps
│ │ │ └── santa_Cruz_and_sampling_area_EG.png # Figure of Santa Cruz island and sampling sites
│ │ └── prospective_selection_EG.png # Figure of prospective selection
│ ├── ./model.out
│ │ ├── GAM_model.selection.txt # GAM model comparisons
│ │ ├── poisson.spline.model_EG.RDS # GAM model for beak traits
│ │ ├── poisson.spline.model_pca.RDS # GAM model for PC traits
│ │ └── spline.model.parameters.csv # Parameters of spline beak traits
│ ├── ./session_info
│ └── session_information.txt # R Session information (packagesand operating system)
│
└── [...] Miscellaneous files # LICENSE, .Rproj, .gitignore...
```
##################################
### METHODOLOGICAL INFORMATION ###
##################################
Please refer to the article.
The main data file is `bird.data.RData` which contains 3428 rows and 64 columns.
- Each row correspond to a bird tagged with a unique Monel band (BANDFINAL).
- Species information is given in sp2, and sp2.fct.
- Beak measurements were taken multiple times (Beak.{Length, Depth, Width}.{A, B, C})
- The median beak traits was calculated and stored in 'Median{beak trait}'
- If a bird was recaptured and measure more than once, an average of all median beak measurements
is recorded in avg.{mbl, mbw, or mbd}.
- Body measurements are in Tarsus, Wing.Chord, and Mass.
- A capture history (ch) is provided along a 'corrected' version of the capture history:
- when a bird was capture at time T and seen at time T+2 but not seen in T+1, the corrected ch 'fills'
the T+1 with a infered 'observation'.
- The fitness measure is in mxcpois which is the last year a bird was captured minus the first year
(or maximum count - 1).
- GPS location of mistnet locations and sites in `Galapagos.EG.nets.csv`
- For maps:
- `galapagos.gis.gpkg` contains the Galapagos island contour and known roads
- `map.back.gpkg` contains polygons of land features at El Garrapatero.
Other datasets:
The precipitation data `climate_puerto-ayora_V.XXXX-XX-XX.csv`:
- Source: https://www.darwinfoundation.org/images/climate/climate_puerto-ayora.csv
The ENSO data `oni.ascii_V.XXXX-XX-XX.txt`
- Source: https://www.cpc.ncep.noaa.gov/data/indices/oni.ascii.txt
The highland and lowland G. fuliginosa data `H.L.lands.fig3.Kleindorfer.2006.csv`
- Source: Figure 3 from Kleindorfer et. al. 2006
##################################
########### REFERENCES ###########
##################################
Reference of functions, data and code are provided inside scripts
R Core Team. 2023. R: A language and environment for statistical computing. R Foundation for Statistical Computing.
Vienna, Austria.