The analysis of GOM Fish Trait and Phenology
If you use this code, please cite our paper: Fujiwara, M., A. Simpson, M. Torres‐Ceron, and F. Martinez‐Andrade. 2022. Life‐history traits and temporal patterns in the incidence of coastal fishes experiencing tropicalization. Ecosphere 13, <10.1002/ecs2.4188>.
Data S1
GAM analysis and subsequent model selection and analysis to compare patterns with fish traits
Author of the material provided in DataS1.zip Author 1: Masami Fujiwara Affiliation: Department of Ecology and Conservation Biology, Texas A&M University Address: 534 John Kimbrough Blvd, College Station, TX 77843-2258, USA Email: fujiwara@tamu.edu
File list (files found within DataS1.zip)
a_GAM_ANALYSIS.R
b_extractExplanatoryVars.R
b_scaling_factor.R
c_ModelSelection.R
d_comparisons.R
e_plotpatternfigures.R
f_plotFigurebySpeciesv.R
The following data will be available upon request.
DATA.Rdata
Demography.Rdata
ExpVars.rdata
SCALE_FACTORS_S.Rdata
STATION.Rdata
TPWD.Rdata
Description
a_GAM_ANALYSIS.R
• GAM is fit to presence-absence data using binomial model. • Environmental variables are the residuals after fitting a model with smooth trends of year and season. • Four covariates (temperature, salinity, DO, SL) are fitted one at a time. • Including the no-covariate model, there are ten models for each species. • The best model for each species is selected with "c_ModelSelection.R". • Input data are o DATA.Rdata o STATION.Rdata o TPWD.Rdata • Outputs are saved in o ResultsGAM.Rdata
b_extractExplanatoryVars.R
• Calculate the scaled explanatory variables. • These were calculated in a_GAM_ANALYSIS.R, but I forgot to save them in the script. • I just copied a part of the script and pasted here and saved the explanatory variables.
b_scaling_factor.R
• Calculate the scaling factors used for scaling explanatory variables. • These were calculated in a_GAM_ANALYSIS.R, but I forgot to save them in the script. • I just copied a part of the script and pasted here and saved the scaling factors.
c_ModelSelection.R
• The best model for each species is selected based on AIC. • Take the results from GAM_ANALYSIS.R (ResultsGAM.Rdata) and saved in "Seine_Best.Rdata".
d_comparisons.R
• Conduct a cluster analysis on the yearly and seasonal trends. • Compare the yearly and seasonal trend patterns and demographic characteristics of fishes • Take the information in "Demography.Rdata" (fish trait data). • Take the information in "Seine_Best.Rdata" from "c_ModelSelection.R". • Results are saved in 'clusters.Rdata' • Figures produced are Figures 4 and 5.
e_plotpatternfigures.R
• Take the information in “cluster.Rdata” and “SCALE_FACTORS_S.Rdata”. • Plot figures of patterns Figures 2 and 3.
f_plotFigurebySpeciesv.R
• Create Appendix S4.
DATA.Rdata
[1] "TPWD_ID": Texas Parks and Wildlife Department fish species ID number
[2] "station_id": Texas Parks and Wildlife Department station ID number (unique for each sampling occasion)
[3] "major_area": ID for bays 1: Sabine Lake, 2: Galveston Bay, 3: Matagorda Bay, 4: San Antonio Bay, 5: Aransas Bay, 6: Corpus Christi Bay, 7: Upper Laguna Madre, and 8 Lower Laguna Madre
[4] "year": Year of sampling
[5] "month": Month of sampling
[6] "cpue": Catch per unit effort
[7] "catch": Number of individuals caught
[8] "mean_length": Mean length of selected individuals (not all individuals)
[9] "day": Day of sampling
[10] "season": Season (1-24) of sampling (see the main text)
[11] "sci_name": Scientific name of the species
Demography.Rdata
• Fish trait date for 275 species of fishes found in the TPWD bag seine data. • Each list item is for each species as described below:
Information from FishLife R Package.
[1] "Amax": Maximum age from FishLife R package
[2] "Amat": Age of maturity from FishLife R package
[3] "Linf": Asymptotic length from FishLife R package
[4] "Kapp": von Bertalanffy growth parameter from FishLife R package
[5] "Lmat": Length of maturity from FishLife R package
[6] "Lamb": The maximum asymptotic population growth rate from FishLife R package
Information from Texas Parks and Wildlife Department data
[7] "common_name": Common name of fish species
[8] "TPWD_ID": Texas Parks and Wildlife Department fish species ID number
[9] "seine": How frequently the species was caught in bag seine.
[10] "trawl": How frequently the species was caught in bay trawl.
[11] "gillnet": How frequently the species was caught in bay gill net.
Information from FishBase
[12] "FB_SPECI": Species information downloaded from FishBase.
[13] "FB_ESTIM": Estimated trait information downloaded from FishBase.
[14] "FB_FECUN": Fecundity data downloaded from FishBase.
[15] "FB_ECOLO": Ecological information downloaded from FishBase.
[16] "FB_MATUR": Maturity information downloaded from FishBase.
[17] "FB_FOODI": Food item information downloaded from FishBase.
[18] "FB_REPRO": Information on reproduction downloaded from FishBase.
[19] "FB_LENGT": Length information downloaded from FishBase.
[20] "sci_name": Scientific name.
Derived parameters
[21] "a_lw": Allometric coefficient for length weight relationship.
[22] "b_lw": Allometric coefficient (exponent) for length weight relationship.
[23] "M_u": Mortality at 1 g under Lorenzen (1996) model.
[24] "B_L": Lorenzen (1996) exponent for the weight-based mortality model.
[25] "A0": estimated von Bertalanffy growth model parameter (hypothetical age at size 0).
[26] "Age": Age from 1 to the maximum age of the species.
[27] "L": Length at given age.
[28] "W": Weight at given age.
[29] "M": Instantaneous mortality age given age.
[30] "S": Finite per-capita survival rate.
[31] "A1": Population matrix when the asymptotic population growth rate was 1.
[32] "A2": Population matrix when the asymptotic population growth rate was “Lamb”
[33] "GT1": Generation time estimated using population matrix A1.
[34] "NR1" The expected number of years in mature stage calculated with A1.
[35] "DR1" Damping ratio estimated from population matrix A1.
[36] "GT2": Generation time estimated using population matrix A2.
[37] "NR2": The expected number of years in mature stage calculated with A2.
[38] "DR2": Damping ratio estimated from population matrix A2.
ExpVars.rdata
The environmental covariates scaled so that the mean is 0 and standard deviation is 1.
SCALE_FACTORS_S.Rdata
The mean and standard deviation of environmental variables.
STATION.Rdata
• Station specific information.
[1] "major_area": ID for bays 1: Sabine Lake, 2: Galveston Bay, 3: Matagorda Bay, 4: San Antonio Bay, 5: Aransas Bay, 6: Corpus Christi Bay, 7: Upper Laguna Madre, and 8 Lower Laguna Madre
[2] "year": year of sampling
[3] "month": month of sampling
[4] "station_id": Texas Parks and Wildlife Department station ID number (unique for each sampling occasion)
[5] "temperature": temperature
[6] "salinity": salinity
[7] "diss_oxygen": dissolved oxygen
[8] "turbidity": turbidity
[9] "latitude": latitude
[10] "longitude": longitude
[11] "depth": depth
[12] "msl": mean sea level
[13] "day": day of sampling
[14] "season": season (1-24) of sampling (see the main text)
[15] "date": date of sampling
[16] "DoY": day of the year of sampling
TPWD.Rdata
• Species specific information from the original data.
[1] "common_name": Common name
[2] "sci_name": Scientific name
[3] "TPWD_ID": Texas Parks and Wildlife Department fish species ID number
[4] "seine": How frequently the species was caught in bag seine.
[5] "trawl": How frequently the species was caught in bay trawl.
[6] "gillnet": How frequently the species was caught in gill net.