A flexible point and variance estimator to assess bird/bat fatality from carcass searches

Abstract

Estimation of bird and bat fatalities due to collision with anthropogenic structures (such as power lines or wind turbines) is an important ecological issue. However, searching for collision victims usually only detects a proportion of the true number of collided individuals. Various mortality estimators have previously been proposed to correct for this incomplete detection, based on regular carcass searches and additional field experiments. However, each estimator implies specific assumptions/restrictions, which may easily be violated in practice. In this study, we extended previous approaches and developed a versatile algorithm to compute point and variance estimates for true carcass numbers. The presented method allows for maximal flexibility in the data structure. Using simulated data, we showed that our point and variance estimators ensured unbiased estimates under various challenging data conditions. The presented method may improve the estimation of true collision numbers, as an important pre-condition for calculating collision rates and evaluating measures to reduce collision risks, and may thus provide a basis for management decisions and/or compensation actions with regard to planned or existing wind turbines and power lines.