Behavioral components of dispersal in Drosophila mimica.

In the Hawaiian species, Drosophila mimica, it was necessary to determine the dispersal pattern in order to interpret genetic heterogeneity observed in the population. In addition, the pattern of colonization and speciation in the Drosophila community may have been affected by the dispersal behavior and response to infrequent conditions of environmental factors which would encourage sporadic movement. It was not possible in this experiment to measure such behavior directly, but predictions are possible.Dispersal patterns are formulated in terms of component behavioral responses to particular environmental cues. Under appropriate field measurements, the relationship of behavior patterns of D. mimica to air movement, light, humidity, and food substrate were sufficient to interpret a complex dispersal pattern in the field. Predictions were made for other movement patterns over longer distances and with infrequent occurrence.Present results are explainable by the response of D. mimica to low velocity air currents. Flies move into the current at velocities less than about 3.3 km/hr, and this response is consistent with field observations of insect and air movements. Visible light intensities from more than 10 to about 100 lux were most acceptable to this species, and movement was restricted to light periods.Based upon responses of D. mimica to wind, light, humidity, and attractive areas, long range dispersal would likely occur between two habitats when each had suitable food substrates, the weather was very humid, overcast, and when there was a slight air current between the habitats. Movement would be unidirectional only if air current patterns were stable. The most precise navigation would be at air current velocities below 3.3 km/hr, where olfactory orientation on the substrate in the new habitat would be possible and short flights would be upwind. However, longer range movements would be possible downwind at slightly higher velocities.Genetic data on D. mimica has been interpreted to reflect differences in selection between at least moderately isolated habitats. Our results indicate that, instead of isolated populations, these habitats contain populations that are part of a single deme with sporadic mixing. Genetic heterogeneity may temporarily result from intense localized selection pressure between periods of mixing.Furthermore, the presence of several sibling species (some undescribed) in Kipuka Puaulu and in several neighboring areas clearly illustrates the fragmentation of gene pools and the genetic differentiation which is possible. Since D. mimica might move from one kipuka to another, spatial isolation is unlikely to play a major role in speciation, whereas habitat selection and ethological isolation may be the principal mechanisms of speciation.