READ ME FILE Two datasets, three R scripts, and one SYSTAT script accompany the manuscript, Population Regulation and Density-Dependent Demography in the Trinidadian Guppy Authors are: Joseph Travis (travis@bio.fsu.edu) Ronald Bassar (rdb4@williams.edu) Tim Coulson (tim.coulson@zoo.ox.ac.uk) Andres Lopez-Sepulcre (alopezsepulcre@wustl.edu) David Reznick (gupy@ucr.edu) Brief Summary: We used monthly mark-recapture data to examine population dynamics and demography in four experimental populations of guppies in natural streams. Three of the four populations displayed clear evidence of regulation. In all four populations, monthly adult survival rates were independent of biomass density or actually increased with increased biomass density. Juvenile recruitment, which is a combination of adult fecundity and juvenile survival, decreased as biomass density increased in all four populations. Demography showed marked seasonality, with greater survival and higher recruitment in the dry season than the wet season. All authors are responsible for data collection; Bassar has primary responsibility for data curation; Travis is responsible for data analyes and code. --------------------------------------------------------------------------------------------------------------------- File Name: DensityDependentGuppyDemographyDRYAD.csv Variable/Column: Description _________________________________________________________________________________________________________ STREAM: String variable identifying the stream (CA = Caigual, LL =Lower Lalaja, TY = Taylor, UL = Upper Lalaja) SAMPLING: Month of study M_PHI: male survival from the previous month as estimated from Program MARK SEMPHI: standard error of M_PHI LOGIT_M_PHI: logit transformed value of M_PHI LOGITSE_M_PHI: standard error of M_PHI on the logit scale F_PHI: female survival from the previous month as estimated from Program MARK SEF_PHI: standard error of F_PHI LOGIT_F_PHI: logit transformed value of F_PHI LOGITSE_F_PHI: standard error of F_PHI on the logit scale M_P: probability of capture for males as estimated from Program MARK MPSE: standard error of M_P M_P_UCL: upper 95% confidence interval limit for M_P F_P: probability of capture for females as estimated from Program MARK FPSE: standard error of F_P F_P_UCL: upper 95% confidence interval limit for F_P F_DENS_HAT: numerical density (number per square meter) of females, corrected for the probability of capture, as estimated from Program MARK SE_F_DENS: standard error of F_DENS_HAT M_DENS_HAT: numerical density (number per square meter) of males, corrected for the probability of capture, as estimated from Program MARK SE_M_DENS: standard error of M_DENS_HAT F_REC_DENS_HAT: recruitment density of females (number of females recruited per adult female) from the previous month to the current month, corrected for the probability of capture, as estimated from Program MARK SE_FREC_DENS: standard error of F_REC_DENS_HAT M_REC_DENS_HAT: recruitment density of males (number of females recruited per adult female) from the previous month to the current month, corrected for the probability of capture, as estimated from Program MARK SE_MREC_DENS: standard error of M_REC_DENS_HAT TOTAL_DENS_HAT: numerical density of all individuals (number per square meter, males plus females), corrected for the probability of capture, calculated from the sum of male and female densities SE_TOTAL_DENS: standard error of TOTAL_DENS_HAT, calculated as described in Appendix 1 of the Supplemental Materials TOT_REC_DENS: density of all recruits (females plus males) per adult female, calculated as the sum of recruitment densities of males and females SE_TOT_REC_DENS: standard error of TOT_REC_DENS, calculated as described in Appendix 1 of the Supplemental Materials FEMMEANMASS: average body mass in grams of all adult females captured in that month MALEMEANMASS: average body mass in grams of all adult males captured in that month FEMSEMASS: standard error of FEMMEANMASS MALESEMASS: standard error of MALEMEANMASS TOTBIOMDENS: Biomass density (grams per square meter) of all individuals, calculated as described in Appendix 1 of the Supplemental Materials FEMVAR: Variance in female biomass density, calculated as described in Appendix 1 of the Supplemental Materials MALEVAR: Variance in male biomass density, calculated as described in Appendix 1 of the Supplemental Materials VARTOTBIOMDENS: Variance in TOTBIOMDENS, calculated as described in Appendix 1 of the Supplemental Materials SETOTBIOMDENS: Square root of VARTOTBIOMDENS CVTOTBIOMDENS: Coefficient of variation in total biomass density, calculated by dividing SETOTBIOMDENS by TOTBIOMDENS ------------------------------------------------------------------------------------------------------------------- Survival Analyses File Name: survival.csv Each row represents an individual in the data set. IMPORTANT: Opening this file in Excel will change irreversibly the values of "ch" Variable/Column: Description _________________________________________________________________________________________________________ ID: individual identification number ch: Capture history over the course of the study. 0 = not caught, 1 = caught. Stream: String variable identifying the stream (CA = Caigual, LL =Lower Lalaja, TY = Taylor, UL = Upper Lalaja) sex: String variable identifying the sex of the individual. M = male; F = female. File Name: cap.dates.csv Each row represents a capture event in a stream. Variable/Column: Description _________________________________________________________________________________________________________ stream: String variable identifying the stream (CA = Caigual, LL =Lower Lalaja, TY = Taylor, UL = Upper Lalaja) sampling: Numeric variable describing the month of the experiment. cap.date: Date of capture. int: interval in days between capture events. ------------------------------------------------------------------------------------------------------------------- All analyses conducted in standard Program R (R Core Team 2020) Package dlm used to analyze state-space models: R (v3.6.3). Survival Analyses:R (v4.2.1) and packages RMark (v3.0.0) and reshape2 (v1.4.4). Program Mark required to run RMark (see: https://urldefense.com/v3/__http://www.phidot.org/software/mark/__;!!PhOWcWs!xiVH7REeIdg4ss8E_zPMMs_a-GP3SmCM4tJZ4uJh7WYx83vEiKFME4qD5-YVpGd56D24ZlNqOs7-5-sBhM--$ ) to install Program Mark. -------------------------------------------------------------------------------------------------------------------- POPAN Models Density Paper.R contains script for analyzing mark-recapture data in Program MARK to obtain monthly survival and recruitment data for each of two sexes in each of four streams R code for simulated relationships of demographic variables on density.R contains R script for simulating 1000 regressions of each vital rate on density, drawing from a normal distribution of vital rates (logit transformed survival rates or recruitment rates) and a normal distribution of log density, distributions defined by means and variances obtained from Program MARK R code for state space models using package dlm.R contains R script for building and comparing state-space models of the monthly changes in log biomass density using the package dlm ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Using SYSTAT to use linear models for analyzing vital rates as functions of density and season SYSTAT script based on SYSTAT Version 12.0 (proprietary software), used within the Windows operating system. First step is to load the data. This can be done by clicking on the "File" name on the toolbar, clicking on "Open," then clicking on "Data." Systat will then offer a list of files as is done in any Windows application. Clicking on the dataset will load it for use in SYSTAT. The script provided can be implemented within the General Linear Model function. The function can be accessed in two ways: 1. One can click on the SYSTAT toolbar under "Analyze"and find the General Linear Model tool. One then clicks on "Estimate Model" and fills in dependent and independent variables from the list of variables. The output will appear immediately. Additional options allow testing of specific factors and a priori contrasts if desired. 2. One can copy and paste the script directly into the "Interactive" tab at the bottom of the screen. This by-passes the "point and click" system and functions in the same manner as using R scripts or SAS scripts.