Fig. 2: Infection time distributions from our transmission network model for individuals involved in a putative transmission chain, along with the likely direction of transmission (red arrows) and the spatial context (see Fig. S5/S6 for information on other transmission events in the treatment and stable region). Grey circles encompassing puma silhouettes in the map insets represent known territorial overlap between individuals (based on observations from K. Logan) and is not representative of home range size. Orange: lag 1 period from the start of the no-hunting period when males are recruited to the study area and lag 2 period at the start of the hunting period until the population declines, light yellow: hunting pressure relieved, red: hunting pressure. Birth year is indicated by the cub silhouette, and death year of M73 is indicated by the black horizontal line. The orange horizontal line indicates when the FIVpco CO III lineage was introduced into this population based on node estimates from (Fountain-Jones et al. 2019). Red horizontal lines indicate transmission time distributions (overlap between infection time distributions) and ‘trans’ means ‘likely transmitted to’.
Fig. 3. Eliminating hunting mortality led to: (a) overall greater phylogenetic clustering (i.e., lower phylogenetic diversity) of FIVpco isolates standardized for sample number and (b) an increase in FIVpco population growth rate that was (c) strongly correlated with male population size rather than (d) female population size. (a) Standardized phylogenetic diversity (*: SES.PD, standardized effect size phylogenetic diversity calculated from 1000 posterior trees) estimates revealed strong patterns of phylogenetic clustering (smaller distances between isolates than expected by chance) when hunting pressure was relieved (negative values of SES.PD). Otherwise, FIVpco isolates were more dispersed across the tree (SES. PD ~ 0, indicated by the dashed line). Estimates of FIVpco prevalence (number of qPCR positives/total number sampled) are provided next to each box and whisker plot with number of individuals tested shown in parentheses (see Fig. S8 for estimates of prevalence across years). (b) Viral population growth rate was estimated using Bayesian phylodynamic reconstruction (Volz & Didelot 2018). See Fig. S7a for the corresponding skyline plot (effective population size through time estimated via thephylodyn model (Karcher et al. 2016)) for the treatment region and Fig. S7b/c for complementary plots for the FIVpco clade dominant in the stable region. Lg refers to the period between hunting being reinstated and the start of population decline.