This repository contains the data and code that supports the manuscript: Katelyn B.S. King, Henrique C. Giacomini, Kevin Wehrly, Hernán López-Fernández, Andrea K. Thomer, Karen M. Alofs. Using historical catch data to evaluate predicted changes in fish relative abundance in response to a warming climate.
The manuscript investigates if models built using spatial environmental gradients can reliably predict population changes through time in response to climate warming. We use largemouth bass abundance across Michigan lakes to validate hindcast.
The 'Code' folder includes the following scripts:
01_data_explore includes exploring driver variables
02_map_code includes code for mapping sample sites
03_model1 includes the methods for running model 1 and calculating deviance, residual deviance, and Bayes R2
03_model2 includes the methods for running model 2 and calculating deviance, residual deviance, and Bayes R2
04_deviance_resid includes the methods for comparing the residuals from the two time periods
05_effects_plots_model1 includes the methods for creating effects plots for model1
05_effects_plots_model2 includes the methods for creating effects plots for model2
06_obs_vs_pred_Fig6 includes the methods for plotting the observed vs predicted values from the two time periods
07_hindcast_abundance includes the methods for hindcasting abundace from model 2
The 'Data' folder includes the following datasets:
contemp_lmb_dat.csv which are the contemporary lake attributes and catch data used in the models. See the Metadata.xlsx file which describes the variables.
historical_lmb2_with_secchi.csv which are the historical lake attributes and catch data used to validate the models. See the Metadata.xlsx file which describes the variables.
Humphries_table.csv which is the authority file for lake ids and latitude/longitude for Michigan lakes. See the Metadata.xlsx file which describes the variables.
lake_surface_temp.csv which are the modeled yearly lake temperatures from James Breck Michigan Department of Natural Resources. See the Metadata.xlsx file which describes the variables.
MDNR_FMUs folder contains the shapefiles for the Fisheries Management Units
output folder contains the density estimates, temperature changes, and the residual deviance for all of the folds from the model for the contemporary and historical catch data.
The 'figures' folder includes the figures made in R for the manuscript