Dataset 1. Data behind Figure 1a. Columns indicate the population size (N), density (Density), detection probability for a single animal over a single trap-day (N1dailydetectionprob), the probability of detecting one of N wolverines in a single trap-day (RNdailydetectionprob) calculated from the equation taken from Royle and Nichols [7]: RNdailydetectionprob =1-(1- N1dailydetectionprob) N, and the approximated single trap-day probability of detecting a population of wolverines at a given density (approxdailydetectionprob) based on the equation: approxdailydetectionprob =E(d/d) where E is the known daily detection probability of a reference population at density d and d is the density of the population being surveyed. The reference population in this example had a density of 0.00122 wolverines/km2 and E=0.03, corresponding to the single trap-day detection probability for 10 individuals. When E(d/d*)>1, the approximate daily detection probability was set to 1.0 since probabilities are restricted to the range 0-1. Dataset 2. Data behind Figure 1b. Data table indicates for initial population sizes (N0) from 2-25 wolverines, the extinction (probextinct) and corresponding persistence (probpersist) probabilities of 10,000 simulations in program VORTEX assuming the demographic parameters in Appendix A. Dataset 3. Data behind Figure 2. The columns show the probability of detecting at least one wolverine in a population inhabiting Sequoia-Kings Canyon National Parks at a given density (density) assuming 982 trap-days (detection982) or 1418 (detection1418), and the corresponding probability of failing to detect a viable population assuming 982 (vpnondetection982) or 1418 (vpnondetection1418) trap-days. A viable population here was defined as a population that persists at least 25 years. Dataset 4. List of photographs from baited camera stations showing animals. This dataset includes a picture ID, the species observed in the picture, camera ID, Site ID, date and time recorded for each picture and any associated notes. The picture ID is a unique alphanumeric string assigned to each photo file by the camera when the picture was taken. The camera ID is a 4-digit number assigned to each camera and corresponding to the year on dates recorded on pictures taken by the camera. The date programmed into each camera was set to the correct day and month but assigned a unique year to ensure that photographs from the camera could later be tied to the correct baited camera station. Each camera ID corresponds to a single site ID. The site ID is a two letter identification assigned to each baited camera station and corresponds to the site ID listed in Table 1. The date and time for each picture taken are the day, month, and time of day recorded by the camera on the picture file.

Dataset 1. Data behind Figure 1a. Columns indicate the population size (N), density (Density), detection probability for a single animal over a single trap-day (N1dailydetectionprob), the probability of detecting one of N wolverines in a single trap-day (RNdailydetectionprob) calculated from the equation taken from Royle and Nichols [7]: RNdailydetectionprob =1-(1- N1dailydetectionprob) N, and the approximated single trap-day probability of detecting a population of wolverines at a given density (approxdailydetectionprob) based on the equation: approxdailydetectionprob =E(d/d) where E is the known daily detection probability of a reference population at density d and d is the density of the population being surveyed. The reference population in this example had a density of 0.00122 wolverines/km2 and E=0.03, corresponding to the single trap-day detection probability for 10 individuals. When E(d/d*)>1, the approximate daily detection probability was set to 1.0 since probabilities are restricted to the range 0-1. Dataset 2. Data behind Figure 1b. Data table indicates for initial population sizes (N0) from 2-25 wolverines, the extinction (probextinct) and corresponding persistence (probpersist) probabilities of 10,000 simulations in program VORTEX assuming the demographic parameters in Appendix A. Dataset 3. Data behind Figure 2. The columns show the probability of detecting at least one wolverine in a population inhabiting Sequoia-Kings Canyon National Parks at a given density (density) assuming 982 trap-days (detection982) or 1418 (detection1418), and the corresponding probability of failing to detect a viable population assuming 982 (vpnondetection982) or 1418 (vpnondetection1418) trap-days. A viable population here was defined as a population that persists at least 25 years. Dataset 4. List of photographs from baited camera stations showing animals. This dataset includes a picture ID, the species observed in the picture, camera ID, Site ID, date and time recorded for each picture and any associated notes. The picture ID is a unique alphanumeric string assigned to each photo file by the camera when the picture was taken. The camera ID is a 4-digit number assigned to each camera and corresponding to the year on dates recorded on pictures taken by the camera. The date programmed into each camera was set to the correct day and month but assigned a unique year to ensure that photographs from the camera could later be tied to the correct baited camera station. Each camera ID corresponds to a single site ID. The site ID is a two letter identification assigned to each baited camera station and corresponds to the site ID listed in Table 1. The date and time for each picture taken are the day, month, and time of day recorded by the camera on the picture file.