Dispersal rates for freshwater invertebrates are often inferred from population genetic data. Although genetic approaches can indicate the amount of isolation in natural populations, departures from an equilibrium between drift and gene flow often lead to biased gene flow estimates. I investigated the genetic population structure of the pond-dwelling fairy shrimp Branchinecta coloradensis in the Rocky Mountains of Colorado, U.S.A., using allozymes. Glaciation in this area and the availability of direct dispersal estimates from previous work permit inferences regarding the relative impacts of history and contemporary gene flow on population structure. Hierarchical F statistics were used, to quantify differentiation within and between valleys (theta(SV) and theta(VT) respectively). Between valleys separated by 5-10 km, a high degree of differentiation (theta(VT) = 0.77) corresponds to biologically reasonable gene flow estimates of 0.07 individuals per generation, although it is possible that this value represents founder effects and nonequilibrium conditions. On a local scale (less than or equal to 110 m), populations are genetically similar (theta(SV) = 0.13) and gene flow is estimated to be 1.7 individuals exchanged between ponds each generation. This is very close to an ecological estimate of dispersal for B. coloradensis via salamanders. Gene flow estimates from previous studies on other Anostraca are also similar on comparable geographic scales. Thus, population structure in B. coloradensis appears to be at or near equilibrium on a local scale, and possibly on a regional scale as well.
