Secondarily, if no alleles are lost in the founding of the new population, one can also perform simulations as above to determine the number of colonists at which there is a 95% chance of losing one or more alleles under each colonization scenario for each locus. This provides a second estimate of the minimum number of colonists that founded the sink population.

Three types of colonization events can be simulated. First, a founding event followed by an instantaneous increase to very large population size (FAST). Thus, all the allele frequency differences between the source and sink populations result from the initial founder event. Second, a founding event followed by an increase to a very large population size (>5000) with increases by a factor of X per generation (SLOW). This model may be more realistic, as genetic drift occurs in the early stages of the colonization after the initial founding event. Finally, the sink population is founded after a series of Y bottlenecks of the specified number of haploid colonists, thus simulating a stepping-stone process in the spread of the species (Y-STEP).

There are some assumptions inherent in these simulations. Our method uses a haploid model rather than a diploid model. Hence, some evolutionary dynamics may not be captured, such as skewed sex ratios or high variance in male mating success. Also, the model assumes that no genetic drift or selection has occurred in the source population subsequent to the migrants leaving to the sink population. We urge potential users of this model to keep these assumptions in mind when performing their calculations.