Bias in aging feral horses.

Abstract

Several investigators studying feral horses (Equus caballus) in the western U.S. have noted anomalies in the age distribution of captured horses that raised concern about the accuracy of aging technique. Possible biases in the aging technique were investigated by assembling records for 60,116 horses removed from public lands in Nevada, Oregon, and Wyoming between 1975 and 1987. Records were consolidated for each state and the resulting age distributions were compared to an expected distribution derived from a population model based on published demographic parameters. These comparisons revealed a tendency for 5-year-olds to be under represented while 6- and 7-year-olds were over represented. There were also higher than expected numbers of horses aged as 15 and 20 years. These apparent anomalies were consistent among states and also among years within each state. The pervasiveness of the trends demonstrates that several biases exist in the current aging technique. Given these uncertainties, it is recommended that development of age-specific demographic variables based on yearly increments beyond age 3 or 4 be avoided, instead lumping data into broad age classes whenever possible.