GROWTH-RATES OF FERAL HORSE POPULATIONS

We studied growth rates of feral horse (Equus caballus) populations in the western United States by consolidating aerial count records for 103 management areas administered by the Bureau of Land Management (BLM). Data from most areas could not be used because of free exchange of horses between neighboring areas and/or inadequate or incomplete data. Twenty-two areas provided data that were used to assess the accuracy of aerial counts with the index-removal method. Most (24/31) aerial counts consistently detected a large proportion (85-105%) of the horses. Replicate index-removal estimates provided little evidence of progressively higher sighting probabilities that would inflate growth rate estimates, and we concluded that aerial counts could be used to obtain reliable estimates of population growth rates. We used log-linear regression to estimate finite annual growth rates (lambda). Estimates of lambda for 12 areas with adequate data ranged from 1.15 to 1.27 with a. mean of 1.21. Population simulations from age-specific reproduction and survival data collected independent of our study produced a similar range of growth rates. We concluded that many feral horse populations are currently being maintained at levels below which density-dependent responses operate and are, therefore, increasing at or near their biological maximum.