1. Large carnivore restoration programs are often promoted as capable of providing ecosystem services. However, these programs rarely measure effects of successful restoration on other economically and ecologically important species. In South Florida, while the endangered Florida panther (Puma concolor coryi) population has increased in recent years due to conservation efforts, the population of its main prey, the white-tailed deer (Odocoileus virginianus), has declined in some regions. The extent to which panther predation has affected deer populations has been difficult to assess because several other factors have changed during this period, including hydrology and hunting regulations.

2. We collected known-fate survival data on 241 GPS-collared adult deer (156 females and 85 males) from 2015 to 2018 in the Florida Panther National Wildlife Refuge and the Big Cypress National Preserve in Florida, USA, to assess effects of panther predation on the deer population, while also evaluating the impacts of hunting and hydrology.

3. Predation was the primary cause of death (110 of 134 mortalities), and 87% of predation events were attributed to panthers, a much greater rate than reported by studies conducted before the panther genetic restoration effort initiated in 1995. One deer was legally harvested, and two were likely killed by poachers. Increasing water depth decreased female survival but had little impact on male survival, and drowning was never a cause of mortality.

4. Females had greater survival probability than males, except during fawning season. From 2015 to 2018, annual survival rates increased from 0.61 (0.52-0.70) to 0.86 (0.79-0.91) for females, and from 0.45 (95% CI: 0.33-0.58) to 0.79 (0.69-0.86) for males.

5. Synthesis and applications – High predation rates, coupled with previous evidence of low recruitment of deer in South Florida, suggest that it will be challenging to meet society’s competing demands for large predator restoration and sustainable deer harvest. Deer hunting in the area must remain tightly controlled, for now, if it is to be sustainable, and managers should seek to mitigate effects of high waters and improve deer habitat quality to increase deer population viability. Future work should closely monitor the deer population to assess if management actions can increase vital rates and abundance in the context of high predation rates.

We collected known-fate survival data on 241 GPS-collared adult deer (156 females and 85 males) from 2015 to 2018 in the Florida Panther National Wildlife Refuge and the Big Cypress National Preserve in Florida, USA

We obtained hydrological data from the Everglades Depth Estimation Network database (Telis, 2006). We downloaded daily median water levels from water gages located in our study area (BCA1, BCA12, BCA17 and BCA18) for the study period, and corrected these values to obtain actual water depth (i.e., water height in relation to the ground).

Covariates have been scaled, scaling factors are indicated in the metadata README.txt file.

coda.samples.dic.R
  R script, alternative to coda.sample, while allowing for computation of DIC. Based on http://ihrke.github.io/dic_jags.html.

code.R
  R script, contains sample code to load data, run best model, and extract some information

covars.RData
  R Data, data frame containing covariates for the analysis:
     $ deerID        : Deer unique ID
     $ day           : Date (format: "YYYY-MM-DD")
     $ sex           : Sex ; "f"=female,"m"=male
     $ season        : Biological season ; "fawning", "rearing", "rut", "postrut"
     $ location      : Site ; "BCNP": Big Cypress National Park, "FPNWR": Florida Panther National Wildlife Refuge
     $ waterDepthAvgS: water depth (scaled) ;                                Backtransformation: Water depth (in m) = (X'*s + c) * k     with s= 1.067989 , c= 1.116642 , and k= 0.3048
     $ time          : time since the begining of the study (scaled)         Backtransformation: Time (in days) = X'*s + c                    with s= 412.3724  and c= 34
     $ days.wetS     : time since the last dry day (scaled)                  Backtransformation: Dry time (in days) = X'*s                    with s= 228.797
     $ days.fawnS    : number of days since/to peak fawning (scaled)         Backtransformation: Fawning time (in days) = X'*s + c            with s= 52.89494  and c= 92.18487
     $ days.rutS     : number of days since/to peak rut (scaled)             Backtransformation: Rut time (in days) =  X'*s + c               with s= 52.86728  and c= 90.28011
     $ year          : study year
     $ age           : deer age
     $ age.fac       : deer age as factor ( <2 or >= 2)
     $ age2          : deer age (squared)

data.jags.all.RData
  R Data, list containing deer survival data for all sources of mortality:
     $ fateTime: Individual duration in the study (computed as, if Death: last-first ; if Right-censored: last-first+1)
     $ nDeer   : Number of individuals
     $ nDays   : Total number of days in the study
     $ first   : Time of entry in the study for each individual
     $ last    : Time of exit from the study for each individual

data.jags.panther.RData
  R Data, list containing deer survival data censoring non-panther related sources of mortality:
     $ fateTime: Individual duration in the study (computed as, if Death by panther: last-first ; if Right-censored: last-first+1)
     $ nDeer   : Number of individuals                                                                                
     $ nDays   : Total number of days in the study                                                                    
     $ first   : Time of entry in the study for each individual                                                       
     $ last    : Time of exit from the study for each individual

deer.jags.R
  R script, internal function to prep data based on specified model, call JAGS and run analysis

JPEbledSA1. Model code for JAGS.txt
  JAGS code for the survival model used in the paper

README.txt
  This file, contains general information on what each element contains and does.