Spatial patterns of seed removal by harvester ants in a seed tray experiment.

Abstract

Using a selection of native grass and forb seeds commonly seeded in local restoration projects, we conducted a field experiment to evaluate the effects of seed species, distance of seed patches from nests, and distance between patches on patterns of seed removal by Owyhee harvester ants, Pogonomyrmex salinus (Olsen) (Hymenoptera: Formicidae). To provide context for ants' seed preferences, we evaluated differences in handling time among seed species. In addition, we assessed the influences of cheatgrass, Bromus tectorum (L.) (Poales: Poaceae), and Sandberg bluegrass, Poa secunda (J. Presl) (Poales: Poaceae), cover on seed removal. We found significant differences in removal rates among seed species. In general, seeds placed closer to nests were more vulnerable to predation than those placed farther away, and seeds in closely spaced patches were more vulnerable than seeds in widely spaced patches. However, the strength of these effects differed by seed species. Differences in handling time among seed species may help to explain these findings; the protective effect of from-nest distance was weaker for species that required less time to transport. For 2 of the seed species, there was an interaction between the distance of seed patches from nests and the distance between patches such that the protective effect of distance between patches decreased as the distance from nests increased. Cheatgrass and bluegrass cover both had small protective effects on seeds. Taken together, these results offer insight into the spatial ecology of harvester ant foraging and may provide context for the successful implementation of restoration efforts where harvester ants are present.