Endangered Florida panther population size determined from public reports of motor vehicle collision mortalities

Reliably estimating the abundance of rare or elusive animals is notoriously difficult. An archetypical example is the endangered Florida panther, whose conservation status is intrinsically linked to population size, but for which reliable abundance information is lacking across its range. This is due not only to the inherent difficulty of sampling a rare and elusive species over a large geographic area, but also because of restricted scientific access to private land. Human interactions with wildlife are a regular occurrence, and interactions with non-scientists constitute an important and underutilized source of information about species distribution and abundance. For example, motor vehicle collisions with Florida panthers are recurrent on the vast network of roads within the public and private lands comprising its range in southern Florida, USA. Capitalizing on a tendency for the public to report collisions with species of concern to wildlife officials, we describe a novel methodology using public reports along with routine telemetry monitoring data to produce the first statistically defensible population estimates for the Florida panther across its entire breeding range. In essence, our approach uses traffic volume and road density to estimate the probability of motor vehicle collision mortality from telemetered animals and models counts reported by the public accordingly. Despite low motor vehicle collision mortality probabilities, our methodology achieved abundance estimates of reasonable precision (29% CV) that was similar to that of previous panther studies using conventional approaches on much smaller study areas. While recovery criteria require establishment of three distinct populations of 240 Florida panthers, we found this single population may never have exceeded 150 individuals from 2000 to 2012.Synthesis and applications. By extracting critical demographic information from underutilized aspects of human-wildlife ecology, our citizen-based approach can cost less than conventional alternatives and could conceivably be used for long-term population monitoring of other species over broad geographic areas, for example from reports of avian wind farm collisions, beached whales or marine mammal boat strikes. An additional benefit is that it can be applied to historical data sets of carcass recovery programmes, in our case permitting abundance estimation over a 13-year period. By extracting critical demographic information from underutilized aspects of human-wildlife ecology, our citizen-based approach can cost less than conventional alternatives and could conceivably be used for long-term population monitoring of other species over broad geographic areas, for example from reports of avian wind farm collisions, beached whales or marine mammal boat strikes. An additional benefit is that it can be applied to historical data sets of carcass recovery programmes, in our case permitting abundance estimation over a 13-year period.