Due to continued habitat destruction and species extirpations, the need to use captive breeding for conservation purposes has been steadily increasing. However, the long-term demographic and genetic effects associated with releasing captive-born individuals into the wild remain largely unknown. We developed forward-time, agent-based models to understand these effects.
Schematic representation of the supplementation population models. All models were age and stage-structured and all populations in the wild had the following steps (a): reproduction, density dependent mortality, density independent mortality, growth, and an additional chance of mortality for individuals that had reduced fitness in the wild due to captive ancestry. Every generation, one effective migrant from a neighboring population was introduced. For captive breeding (b), all captive parents were wild-born individuals collected as adults, were only used in a single pair, and were never used for more than a single year. For supplementation scenarios, individuals born in captivity were added to the wild population in order to bolster population size (c). All scenarios were compared to control runs where a captive breeding program was never initiated.
Long-term effects of captive-breeding supplementation
Using the model described above, we found that releasing even slightly less fit captive-born individuals to supplement wild populations can result in substantial reductions in population sizes and genetic diversity over the long term, provided that the fitness reductions are heritable. For example, releasing captive-born individuals with only a 3% reduction in wild fitness resulted in a 20% reduction in population size and a 6% reduction in neutral genetic diversity 100 years after the release of captive-born individuals ceased. Our results suggest that conservation-focused captive breeding programs should implement measures to prevent even small heritable reductions in fitness in the wild environment.
Comparison of population sizes (a, c) and genetic diversity (b, d) for supplemented populations when habitat quality was reduced at year 50 and was never repaired (a, b) and when habitat quality was reduced at year 50 and was repaired at year 100 (c, d). Example is shown using western toad as model species, and genetic diversity is displayed relative to the mean number of alleles present each year in the control simulations. Under both scenarios, supplemented populations are smaller and contain less genetic diversity than populations that were not supplemented.
Bias in our understanding of supplementation effects
In order to evaluate the effectiveness of supplementation programs, the reproductive success of captive-born individuals released into the wild is often compared to the reproductive success of wild-born individuals in the recipient population (relative reproductive success, RRS). Using the model described above, we showed that when captive ancestry in the wild population reduces mean population fitness, estimates of RRS are upwardly biased, meaning that the relative fitness of captive-born individuals is over-estimated. This phenomenon has long-term conservation impacts since management decisions regarding the design of a supplementation program and the number of individuals to release can be based, at least in part, on RRS estimates.
Population size (a) and mean population fitness (b) illustrated before, during, and after the release of captive-born individuals. In both panels, supplemented populations are indicated by blue (dark) lines with the lighter portion indicating the years during which supplementation occurred. Control populations, which did not have a captive breeding program, are plotted behind the supplemented populations (gray lines). For illustrative purposes, genetic adaptation to captivity resulted in a total fitness reduction of 30%. The insets in panels b and d show the mean population tness over the entire simulated time period, with the shaded boxes highlighting the years displayed in the main panels.