Modeling phenological and physiological responses to climate warming in a hypothetical migratory songbird-mosquito system
Data files
Dec 12, 2025 version files 8.87 MB
-
README.md
13.88 KB
-
zipped_data_files.zip
8.86 MB
Abstract
Given the strong temperature-dependence of ectothermic vector physiology, climate warming is expected to profoundly impact many vector-borne diseases. Notably, endothermic hosts can also respond to warming by altering the timing of life history events like seasonal migration and reproduction, but relatively few predictive models of vector-borne disease have considered both phenological and physiological responses to climate warming. Here, we extend the Ross-MacDonald model for a vector-borne disease to incorporate temperature effects on host and vector phenology and physiology. We use this model to understand how projections of moderate and severe warming influence the emergence of a hypothetical vector-borne disease in a migratory bird. Modeled vector and host infection prevalence always increased under warming, and the increase was amplified when hosts failed to update their arrival phenology to keep pace with breeding site resources. While extreme warming scenarios yielded the highest infection prevalence, reduced vector survival in the hottest months caused late-season declines in transmission, altering seasonal patterns of infection. By considering host, vector, and parasite responses to temperature together, our modeling framework could be employed to help decipher otherwise non-intuitive wildlife infection outcomes under current and future climate conditions.
Code and data used in a model of vector-borne disease under climate warming scenarios. R and RStudio are required to run the model.
Description of the data and file structure
"VBD_phenology_physiology_2024.Rproj": R project that bundles the relevant scripts and data files. Save project file in the same folder as the unzipped data folders and the markdown (.rmd) files below. Open the project in R Studio and then access the markdown and data (.csv) files within the project.
RMD files:
"Netcdf_temps_Feb23_23.rmd" Converts netcdf files from Liess et al. 2022 to averaged, smoothed temperature time series in .csv format. Hosted in folder: 'Formatting_CNRM_temp_series' within zipped_data_files (see below);
Access necessary files from the Data Repository for the University of Minnesota: https://doi.org/10.13020/YV29-JY19
"Temperature_degday_scenarios.rmd" reads in smoothed temperature scenarios based on CNRM global climate model for Minnesota - historical temperatures (1980-1999), moderate warming (approx. +2.6C, relative concentration pathway 4.5 2080-2099) and severe warming (approx. +5.4C, relative concentration pathway 8.5 2080-2099). (Ref: https://doi.org/10.1029/2021EA001893). Calculates approximate cumulative degree days above 0C for each of the temperature scenarios; creates temperature and degree day data frames with annual patterns repeating for 20 years. Contains code for several figures related to temperature-dependent parameters.
"Model_pre-infection-dynamics_WNVintro.rmd" contains ordinary differential equations (ODEs) that yield 30-year pre-infection dynamics for hosts and vectors and daily temperature-dependent R0; contains ODEs for the 1 year of infection dynamics after a single infected vector is introduced. To run, user needs to actively switch between the 'only vectors advance phenology' (Vadv) and 'both hosts and vectors advance phenology' (BOTHadv) scenarios before running the for loops. inputs: "Tempramp_MN_Feb27_23.csv", "DD0_20yr_Mar1.csv" outputs: "OutHosts_Vadv_Nov2023.csv", "OutHosts_BOTHadv_Nov2023.csv", "WNVout_Vadv_Nov2023.csv", "WNVout_BOTHadv_Nov2023.csv"
"Main_VBDmodel_results.rmd" contains a review of the main methods and equations; it reads in data frames that contain pre and post-infection host and vector dynamics. Includes plots used in the main results section of the manuscript.
"sensitivity_fc_theta.rmd" focused sensitivity analysis assessing the impacts of the warming-induced cost to host reproduction and shift in vector emergence rate. To run, user needs to actively switch between the 'only vectors advance phenology' (Vadv) and 'both hosts and vectors advance phenology' (BOTHadv) scenarios, as well as fc and theta scaling parameters.
"physiology_sensitivity_10_27_25.rmd" (older version: "physiology_sensitivity.rmd") focused on sensitivity analysis exploring how holding bite rate, mosquito lifespan, or pathogen development rate constant with temperature impacts transmission potential. Updated version from 2025 holds parameters constant at a scenario-specific value: x at the mean breeding season temperature within each climate scenario. Older version held parameters constant at the same value across scenarios.
"01_Mainmod_earlier_end_breeding.rmd" script used to assess differences in model output given an earlier end to host breeding under warming. Appendix figures created.
"02_Mainmod_feccost_dismort.rmd" script exploring model outcomes with/without a host fecundity cost and disease-induced mortality. Used for Appendix figure S6.
"03_Earlier_breed_end_c_vh.rmd" script used to explore the impacts of ending breeding earlier under warming and turning fecundity cost and disease-induced mortality on/off. Appendix figure S7
S1_SupportingInformation_Appendix: contains- 1. 1. Model parameterization - hosts 2. 2. Model parameterization – vector and pathogen 3. 3. Model parameterization – hosts and vector phenology under climate warming 4. 4. Model analysis: calculating Temperature-dependent relative R0 5. 5. Model assumptions - limitations, implications, and potential model modifications 6. 6. Sensitivity analysis – warming impacts on host fecundity cost (c) and maximum vector emergence rate (εmax) 7. 7. Sensitivity analysis – impacts of physiological parameters on R0 8. Sensitivity analyses - eliminating the fecundity cost or disease-induced mortality 9. Sensitivity analyses - breeding season length 10. Sensitivity analyses - effects of non-competent hosts in the community~ ~
DATA
zipped_data_files.zip : zip file of data below
*For all CSV files generated by RStudio below, the first column (A) is created by the software and is redundant with time or DOY. It does not have a column name and will not be called by RMD files.
'temperature_data' folder
Column definitions:
'tempsmooth' = temperature (°C)
'DOY' = day of year (Julian date).
Files:
"Hist_1980_CNRM_1yrsmooth.csv": daily historical temperature values for 1 year; values from the MN model smoothed using the stats::loess function in RStudio.
"RCP45_2080_CNRM_1yrsmooth.csv": 1 year of daily temperature values for the moderate warming (RCP 4.5) scenario; values from the MN model smoothed using the stats::loess function in RStudio.
"RCP85_2080_CNRM_1yrsmooth.csv": 1 year of daily temperature values for the severe warming (RCP 8.5) scenario; values from the MN model smoothed using the stats::loess function in RStudio.
"Tempramp_MN_Feb27_23.csv": data frame with daily temperatures for all scenarios; columns: 'amb' = historical temperatures (1980-1999), 'rcp45' = moderate warming (relative concentration pathway 4.5), 'rcp85' severe warming (relative concentration pathway 8.5), 'DOY' = day of year (Julian date). This file contains 20 years of repeated temperature vectors; this is extended to 30 years in RMD files.
"DD0_20yr_Mar1.csv": The degree days above 0C accumulated for each day in the temperature scenarios. Each scenario has a column- 'degreeday_hist' = historical, 'degreeday_45' = RCP 4.5, and 'degreeday_85' = RCP 8.5. This file includes 20 years of degree days, and these vectors are extended to 30 years in RMD files.
'data' folder for Main model:
Column definitions:
Sh.0 = number of susceptible hosts in historical scenario
Sh.3 = number of susceptible hosts in moderate warming scenario
Sh.5 = number of susceptible hosts in severe warming scenario
Ih.0 = number of infected hosts in historical scenario
Ih.3 = number of infected hosts in moderate warming scenario
Ih.5 = number of infected hosts in severe warming scenario
Rh.0 = number of recovered hosts in historical scenario
Rh.3 = number of recovered hosts in moderate warming scenario
Rh.5 = number of recovered hosts in severe warming scenario
Sv.0 = number of susceptible vectors in historical scenario
Sv.3 = number of susceptible vectors in moderate warming scenario
Sv.5 = number of susceptible vectors in severe warming scenario
Ev.0 = number of exposed vectors in historical scenario
Ev.3 = number of exposed vectors in moderate warming scenario
Ev.5 = number of exposed vectors in severe warming scenario
Iv.0 = number of infected vectors in historical scenario
Iv.3 = number of infected vectors in moderate warming scenario
Iv.5 = number of infected vectors in severe warming scenario
prevalence.X = scenario-specific host infection prevalence; Ih.X / (Sh.X+Rh.X), where X is 0, 3, or 5 for historical, moderate warming, or severe warming, respectively.
prevalenceV.X = scenario-specific vector infection prevalence; Iv.X / (Sv.X+Ev.X), where X is 0, 3, or 5 for historical, moderate warming, or severe warming, respectively.
Bvh.X = scenario-specific transmission probability from vector to host, where X is 0, 3, or 5 for historical, moderate warming, or severe warming
Bhv.X = scenario-specific transmission probability from host to vector, where X is 0, 3, or 5 for historical, moderate warming, or severe warming
Mv.X = scenario-specific vector mortality rate, where X is 0, 3, or 5 for historical, moderate warming, or severe warming
q.X = scenario-specific pathogen development rate, where X is 0, 3, or 5 for historical, moderate warming, or severe warming
Nb.X = number of breeding adult hosts in scenario X, where X is 0, 3, or 5 for historical, moderate warming, or severe warming
Reff.X = daily, temperature-dependent R0, where X is 0, 3, or 5 for historical, moderate warming, or severe warming
DOY = day of year (Julian date)
time = daily timestep in 30-year sequence
Files:
"OutHosts_Vadv_Nov2023.csv": pre-pathogen introduction when ONLY vectors advance phenology under warming. Tracks SIR-SEI compartments, host and vector prevalence, temperature-dependent parameters, and daily R0 (Reff) for each of the three temperature scenarios (historic (.0), RCP4.5 (.3) and RCP 8.5 (.5)). Also includes day of year (DOY) and timestep (time) for the 30-year data frame.
"OutHosts_BOTHadv_Nov2023.csv": pre-pathogen introduction when BOTH hosts and vectors advance phenology under warming. Tracks SIR-SEI compartments, host and vector prevalence, temperature-dependent parameters, and daily R0 (Reff) for each of the three temperature scenarios (historic (.0), RCP4.5 (.3) and RCP 8.5 (.5)). Also includes day of year (DOY) and timestep (time) for the 30-year dataset.
"WNVout_Vadv_Nov2023.csv": post-pathogen the introduction when ONLY vectors advance phenology under warming. tracks same compartments and variables as pre-WNV data frame for 1 year of WNV infection
"WNVout_BOTHadv_Nov2023.csv": post-pathogen introduction when BOTH hosts and vectors advance phenology under warming. tracks same compartments and variables as pre-WNV data frame for 1 year of WNV infection
SENSITIVITY ANALYSES:
Physiology sensitivity analysis - 'sensitivity_data' folder:
The following csv files contain host and vector dynamics and daily R0 calculated when holding one physiological parameter constant; there is a csv file for each phenology scenario (Vadv = vectors advance, BOTHadv = both hosts and vectors advance). Column definitions are the same as 'data' folder files above.
- vector bite rate constant:
"OutHosts_Vadv_constant_a.csv"
"OutHosts_BOTHadv_constant_a.csv" - vector lifespan constant:
"OutHosts_Vadv_constant_ls.csv"
"OutHosts_BOTHadv_constant_ls.csv" - pathogen development rate constant:
"OutHosts_Vadv_constant_q.csv"
"OutHosts_BOTHadv_constant_q.csv"
Reproductive cost (f_c) and emergence rate (theta) scaling:
pre-infection dynamics - 'sensitivity_data' folder:
The following csv files contain pre-infection host and vector dynamics and daily R0 calculated when varying either the reproductive cost of late arrival (with parameter f_c) or maximum vector emergence (with parameter theta). There are csv files for each phenology scenario (Vadv = vectors advance, BOTHadv = both hosts and vectors advance phenology). Column definitions are the same as 'data' folder files above.
"data_fc_sens_Vadv_Nov2023.csv" #only vectors advance phenology, variable cost (f_c); constant emergence
"data_fc_sens_BOTHadv_Nov2023.csv" #both hosts and vectors advance phenology, variable cost (f_c); constant emergence
"data_theta_sens_Vadv_Nov2023.csv" #only vectors advance phenology, variable emergence (theta); constant cost (f_c)
"data_theta_sens_BOTHadv_Nov2023.csv" #both hosts and vectors advance phenology, variable emergence (theta); constant cost (f_c)
post-infection dynamics - 'sensitivity_output' folder:
The following csv files contain post-infection host and vector dynamics when varying either the the reproductive cost of late arrival (with parameter f_c) or maximum vector emergence (with parameter theta). There are csv files for each phenology scenario (Vadv = vectors advance, BOTHadv = both hosts and vectors advance phenology). Column definitions are the same as 'data' folder files above.
"output_fc_sens_Vadv_Nov2023.csv" #only vectors advance phenology, variable cost (fc); constant emergence (theta)
"output_fc_sens_BOTHadv_Nov2023.csv" #both hosts and vectors advance phenology, variable cost (fc); constant emergence (theta)
"output_theta_sens_Vadv_Nov2023.csv" #only vectors advance phenology, variable emergence (theta); constant cost (fc)
"output_theta_sens_BOTHadv_Nov2023.csv" #both hosts and vectors advance phenology, variable emergence (theta); constant cost (fc)
Non-focal hosts / bite dilution - 'sensitivity_data' folder:
These two CSV files contain summary data on transmission potential and pre-infection, end-of-season host abundance when non-competent hosts are incorporated in the community (N_c). There are files for each phenology scenario (Vadv = vectors advance, BOTHadv = both hosts and vectors advance phenology).
"data_Nc_sens_Vadv_Nov2023.csv" #only vectors advance phenology
"data_Nc_sens_BOTHadv_Nov2023.csv" # both hosts and vectors advance phenology
*Column definitions: *
Nc = number of non-competent hosts (0, 200, or 800)
R0day_X = first day of the year (Julian date) sustained transmission is possible (i.e., R0>1); X corresponds to temperature scenario, "hist," "rcp45," or "rcp85" for historical, moderate warming, or severe warming, respectively
peakR0_X = maximum value of R0, where X represents temperature scenario, "hist," "rcp45," or "rcp85" for historical, moderate warming, or severe warming, respectively
Shlast_X = final susceptible host abundance value, where X represents temperature scenario, "hist," "rcp45," or "rcp85" for historical, moderate warming, or severe warming, respectively
We use differential equations to track population and infection dynamics of a migratory avian host and mosquito vector over a single host breeding season, following the first introduction of a novel pathogen (i.e., assuming no pre-existing immunity in the host). We incorporate climate warming impacts on host and vector phenology and vital rates, assessing outcomes under moderate and severe warming scenarios. Warming affects mosquito emergence timing and, in some scenarios, the arrival date of birds (which in turn influences reproductive success). At the same time, temperature dependence in transmission traits is incorporated by linking mosquito mortality rate, bite rate, and pathogen development rate to daily temperature data. For tractability, we assume that only the focal host species can acquire and transmit the pathogen, and that mosquitoes preferentially bite this host.