We have studied the active dispersal behaviour of Drosophila subobscum Coll. in a citrus orchard near Preveza (in N. W. Greece) following the experimental design proposed by Crumpacker & Williams and using fluorescent dusts to mark wild animals captured which were subsequently released. The parameters quantifying dispersal have been estimated and do not differ from the ones found in similar studies of Begon in English populations and of Burla & Greuter in Swiss populations. D. subobscura however does differ from its American relatives of the obscura group in having a lower dispersion activity. Density estimations of the autumn size indicate that natural populations in Greece are larger than the ones in England. From these densities we derived a tentative estimation of the effective size of the panmictic unit. This size is 20 to 340 times greater than the one estimated by a genetic method based on the temporal differences of allozyme frequencies. However an accurate ecological estimation of the effective size requires a density estimation of the minimal (winter) size and the figure calculated by us is no doubt an overestimation.
Drosophila species usually used in populationat genetic studies. The well known ecological technique of releasing marked individuals and recapturing them was used, at the beginning with visual mutants bred in the laboratory. Since then the technique has been improved in experimental design as well as in the marking technique by the use of ultra-violet fluores: cent dusts. The interest of these studies has been revived by the selectionist-neutralist controversy and especially by some models proposed by the Kimura school (Crow & Kimura, 1970) which predict that the immigration of even one individual per generation in every local population is enough to ~ansform them in to a single panmictic unit.
We have assembled some genetical information concerning the effective density in Greek natural populations of Drosophila subobscum (Loukas, Krimbas & Vergini, in press). With the present paper we start reporting ecological data which will eventually permit us to arrive at comparisons between ecological and genetical estimations. Thus the present paper reports an effort to quantify dispersal in this species and provides also some density estimations.