Meffert chose to conduct her experiments on common houseflies because they breed every few weeks. This model system allowed her to simulate several years of breeding for endangered species that breed only once or twice per year.
In order to simulate small populations of rare animals in zoos, breeding lines were founded by just five pairs of males and female flies. Moreover, because flies lay many eggs, and most higher-order animals have small broods, each fly breeding line was allowed to grow by no more than 50 percent per generation.
Each time a pair of flies mated, researchers measured the fitness of the pair by counting the number of eggs the female laid, and by measuring the number of eggs that hatched.
In the MAI breeding scheme, brothers and sisters were never allowed to mate and were instead paired only with offspring from other breeding pairs. In the alternate breeding scheme, the females were allowed to contribute a disproportionate share of offspring to subsequent generations, based on their fitness scores.
The release into the wild was simulated by adding environmental stress. In the "post-release" breeding scenario, temperatures in incubators were altered on a 12-hour cycle based upon the high and low temperatures recorded in Houston that day.
In both strategies, all breeding lines survived in captivity. In the "wild" setting, four of the inbred populations went extinct, compared to just one extinction for the MAI lines.
"Apparently, the stressful environment served to select only the most fit lines while the more benign environment allowed low-fitness lines to persist in captivity," said Meffert.
Though the findings clearly support the case for expensive MAI breeding schemes for some endangered animals, Meffert suspects that might not be the