Papers in the Biological Sciences


Date of this Version



Tenhumberg in Reintroduction Specialist Group Australasian Newsletter (April 2002).


Copyright 2002, RSG. Used by Permission.


There are many uncertainties that must be evaluated when captive breeding is considered. We have begun examining methods for supporting captive breeding decision-making by combining stochastic models with optimization methods. Captive breeding decisions should be state dependent- i.e., the best decision when there are 20 animals left might be different from one when there are 100 animals left. The method we use to search for state dependent decision that minimize the risk of extinction for a species is called Stochastic Dynamic Programming. We construct stochastic dynamic programming to identify the optimal size of translocations between captivity and the wild. For an initial test we parameterized the model with data on Arabian oryx (Oryx leucoryx). A key result is the importance of captive breeding in minimizing the extinction risk of a species in the wild if we can be sure that the captive population will fare better than the wild population. if the wild population is small the entire wild population is best transferred to a captive breeding facility even if the population in the wild is growing.