Wright’s analyses led him to conclude that even small amounts of gene flow between isolated small populations could offset the adverse effects of genetic drift and inbreeding. That conclusion gave rise to a large body of work aimed at determining exactly how much gene flow, in the form of immigrants per generation, was necessary to offset the adverse effects of genetic deterioration. A rule of thumb of one immigrant per generation emerged (Kimura and Ohta, 1971; Lewontin, 1974; Spieth, 1974) and has been widely adopted in conservation practice. More recently, that rule of thumb has been challenged on the basis of the simplistic assumptions that were used in deriving it (e.g., Mills and Allendorf, 1996; Vucetich and Waite, 2000). At the time this report was prepared, it seemed likely that in real-world applications, one immigrant per generation would be an absolute minimum. Mills and Allendorf (1996) outlined a number of scenarios in which the number of immigrants per generations should probably exceed one, including scenarios in which at least one of the following is the case:

  • Inbreeding depression is believed to be occurring already.
  • Immigrants are closely related to each other or to the receiving population.
  • Effective population size is much lower than the number of animals present.
  • Social, behavioral, ecological, or logistical factors prevent single animals from immigrating successfully.
  • Immigrants are at a disadvantage in probability of survival and reproduction.
  • The receiving population has been isolated for many generations.
  • Extinction risk due to demographic or environmental variation is deemed to be very high unless there is aggressive supplementation.

The authors concluded that up to 10 immigrants per generation might be necessary to effect genetic restoration in those situations. Vucetich and Waite (2000) extended the analyses by modeling variation in population fluctuation and suggested that more than 20 immigrants per generation may be necessary if high population fluctuation leads to drastically reduced effective population size.

In addition to the number of animals to translocate, the interval for doing so must be determined. There are important practical and logistical considerations involved, but the translocation of animals for genetic restoration is usually thought of as being conducted on a per-generation basis. Therefore, one starting point is to determine the generation time of free-ranging horses. Eggert et al. (2010) constructed a pedigree for the Assateague Island horse population on the basis of molecular analyses and herd records and derived an estimate of 10 years. Goodloe et al. (1991) also derived an estimate of 10 years for horses on Cumberland Island, Georgia. Similarly, historical pedigree data on zoo populations of wild equids in North America all have generation time estimates of about 10 years (range 9.6 years in Somali wild ass to 10.4 years in Hartmann’s zebra6). Thus, it would be valid to consider 10 years as an appropriate interval for translocating animals between populations for genetic restoration. On the basis of the literature, it appears that translocation of 10 animals between populations every 10 years would be appropriate.

BLM is already experienced in the capture and transport of animals for population management, and the protocols for translocation would be similar to those currently used for gathers; only the destination of the removed animals would differ. Although the movement of animals among HMAs has the potential to facilitate the spread of pathogens (Champagnon et al., 2012), the probability of that could be minimized through observation


6 Association of Zoos and Aquariums website, www.aza.org. Accessed April 16, 2012.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement