Abstract
Hundreds of millions to greater than one billion North American birds are directly killed each year by human stressors, including collisions with man-made structures, predation by feral and pet cats, intentional and accidental poisoning, and pollution. Because these causes of mortality are increasingly abundant and because some result in large bird die-offs, they have received both increased scientific attention and general media coverage. However, quantifying bird mortality remains imprecise and methods to assess whether these losses cause important biological impacts remain underdeveloped. If local mortality studies followed rigorous design and sampling schemes, allowing comparison of data and scaling up of mortality estimates to broad regions, this could lead to improved analyses. Several analytical techniques – including hierarchical and population models – show potential for improving quantification of anthropogenic mortality of birds and inference of population-level effects. Results arising from improved study designs and analytical techniques will more effectively inform decision making about policies and regulations aimed at reducing avian mortality and minimizing population impacts.