Postprint version. Published in Biological Invasions, Volume 3, Issue 2, June 1, 2001, pages 103-111.
The definitive version is available at http://dx.doi.org/10.1023/A:1014559617282.
The Mediterranean fruit fly (Ceratitis capitata) is an invasive agricultural pest with a wide host range and a nearly global distribution. Efforts to forgo the medfly''s spread into the United States are dependent on an understanding of population dynamics in newly established populations elsewhere. To explore the potential influence of demographic and historical parameters in six medfly populations distributed from Mexico to Peru, we created population genetic null models using Monte Carlo simulations. Null expectations for genetic differentiation (F ST) were compared with actual sequence variation from four highly polymorphic nuclear loci. Four colonization scenarios that were modeled led to unique genetic signatures that could be used to interpret empirical data. Unless current gene flow across Latin America was assumed to be very high, we could reject colonizations consisting of multiple introductions, each of low genetic diversity. Further, if simulated populations were small (N e = 5 × 102 individuals per population), small invasions from a single source consistently produced F ST values comparable to those currently observed in Latin America. In contrast, only large invasions from diverse sources were compatible with the observed data for large populations (N e 5 × 103). This study demonstrates that alternative population genetic hypotheses can be tested empirically even when departures from equilibrium are extreme, and that population genetic theory can be used to explore the processes that underlie biological invasions.