Single large or several small? Population fragmentation in the captive management of endangered species

Captive populations of endangered species are typically maintained effectiv ely as single random-mating populations by translocating individuals betwee n institutions. Genetic, disease, and cost considerations, however, suggest that this may not be the optimal management strategy. Genetic theory predi cts that a pooled population derived from several small isolated population s will have greater genetic diversity, less inbreeding, and less genetic ad aptation to captivity than a single large population of equivalent total si ze, provided there are no population extinctions. These predictions were te sted using populations of Drosophila with effective size comparisons of 50 vs. 2 x 25; 100 vs. 2 x 50 vs. 4 x 25, and 500 vs. 2 x 250 vs. 4 x 100 + 2 x 50 vs. 8 x 25 + 6 x 50. Populations were maintained at the indicated size s as separate pedigreed populations for 50 generations. The several small t reatments were subsequently pooled and maintained for eight to 10 generatio ns prior to determination of fitness and evolutionary potential. Several sm all populations (pooled), when compared to single large populations of equi valent total size, were found to have lower average inbreeding coefficients , significantly higher reproductive fitness under competitive conditions, s imilar fitness under benign captive conditions, higher genetic diversity, a nd equivalent evolutionary potential. Trends favored the several small (poo led) populations in all comparisons at population sizes of 50 and 100. We r ecommend that endangered species in captivity be maintained as several smal l populations, with occasional exchange of genetic material. This has genet ic benefits over current management both in captivity and especially for re introductions, as well as reducing translocation costs and risks of disease transfer. (C) 1998 Wiley-Liss, Inc.