This study examines how the spatial distribution of nests is related to the behavioral interactions of conspecific neighbors in a population of the seed-eating ant, Pogonomyrmex barbatus. Colonies live for 15-20 yr, reaching reproductive age and a stable size at =5 yr. Spatial distributions were measured for 6 yr (1988)(1989)(1990)(1991)(1992)(1993) in a population of =250 colonies of known age. The probability that a 1-yr-old colony occurs in a given location is related to the distance to, and ages of, its five nearest conspecific neighbors.One-year-old colonies are most likely to occur near small, 2-and 3-yr-old colonies.Neighboring colonies encounter each other when foragers of both colonies search the same area on the same day. The probability of an encounter between two colonies decreases with the distance between their nests. For colonies of all ages, encounters are most likely with their larger neighbors, ~5 yr old. Encounters are more likely if there was an encounter the previous day, and this effect can overwhelm the effect of distance. Two-yr-old colonies are more likely than colonies of other ages to lose ground in t;epeated encounters with a particular neighbor.Though encounters are frequent, their costs are low: few of the foragers on a trail that meets a neighbor's actually encounter an ant of the neighboring colony, and interaction with a neighboring colony does not increase the typical duration of a foraging trip. If foragers of two colonies do meet, most fights are brief, with few resulting in injury or death. These results suggest that the cost of conspecific neighbors searching the same ground for seeds may be greater than the cost of behavioral interaction itself. Exploitative competition may have more important effects than interference competition on founding colony survival and thus on the spatial distribution of nests.

Little is known of the e}ects of age structure on the

Background It is difficult to assess applicants' higher-order cognitive thinking skills during conventional resident interviews. Application metrics currently employed are useful indicators of academic and personal success in targeted areas, yet value of this information in predicting future clinical performance is limited.