Can captive rearing promote recovery of endangered butterflies? An assessment in the face of uncertainty

Captive rearing is increasingly used as an interim strategy to maintain at-risk butterfly populations while long-term recovery techniques are developed. However, it is seldom feasible to measure effects of captive rearing on small, fragile, and highly mobile organisms, such as butterflies, in wild populations. We use a series of general population viability models to assess the demographic effects of a number of captive rearing strategies, defined by the source of butterflies brought into captivity (the at-risk “small” population vs. a different, stable “large” population) and the number brought into captivity. In general, captive rearing increases population viability, although the benefits are small for rapidly declining or highly stochastic populations. Taking butterflies from the large population is consistently better than taking butterflies from the small population, as long as the large population is not also at imminent risk of extinction. We then modify these models to include two possible risks of captive rearing: captive-reared individuals might not perform as well as wild individuals, and captive-reared individuals might decrease the population growth rate of the wild population, e.g., by introducing deleterious alleles or diseases. Reductions in individual performance do not change optimal captive rearing strategies. Short-term extinction risk is also robust to small (≲5%) reductions in population growth rate. Although studying performance of captive-reared individuals in the wild is often not feasible, it is often possible to monitor performance while in captivity. Captive rearing is a robust way to maintain severely at-risk populations in the short-term, though it cannot replace long-term solutions.