Metrics of predation: perils of predator-prey ratios

We developed an original modeling approach using program StellaA (R) to investigate the usefulness of predator-prey ratios (PPRs) for interpreting top-down and bottom-up forcing on moose Alces alces. We included density-dependent feedbacks for the moose population, allowed K to vary based on amount and quality of available forage for moose, integrated effects of compensatory mortality, and added time lags in wolves Canis lupus tracking the moose population. Modeling scenarios we developed included bottom-up and top-down regulation as predetermined outcomes. We then evaluated whether PPRs would reflect the various combinations of trajectories of predator and prey populations under top-down versus bottom-up regulation. The resulting patterns of PPRs were impossible to disentangle from one another, and did not provide reliable insights into whether top-down or bottom-forcing occurred, especially over short time spans where critical decisions related to management of moose and wolves might be necessary. Only under top-down regulation did PPRs reflect the degree of predation experienced by moose, but in that instance, knowledge of top-down regulation must be known a priori to correctly interpret PPRs. Potential problems with interpreting PPRs include their double-variable nature, which resulted in the failure to reflect patterns of increase and decrease for predators and prey. We suggest that confidence intervals for PPRs be calculated from a binomial, similar to that proposed for sex and age ratios, which should help discourage the inappropriate use of this metric. We caution that the temptation to use PPRs often is irresistible, but their reliability is highly questionable. We provide an alternative method to using PPRs or other predation metrics for determining whether top-down or bottom-up forcing is occurring by adopting an approach based on the physical condition and life-history characteristics of prey.