干旱区生态环境知识资源中心

The responses of animals to patch structure at different spatial scales influence how processes such as foraging and species interactions vary with spatial scale or translate across levels of organization. We investigated how seed harvesting by the western harvester ant (Pogonomyrmex occidentalis) was affected by vegetation structure in a shortgrass steppe ecosystem at levels of organization ranging from individual ants to populations and at spatial scales from 0.1 m(2) to 15 ha. In the first of two experiments, we used a hierarchical experimental design to assess the magnitude and Variation in seed harvesting (1) among ant colonies in three pastures differing in livestock grazing intensity, (2) among colonies within pastures, and (3) among plant cover types within colonies. We found that seed harvesting varied significantly across pastures and showed a weak interaction with fine-scale plant cover. Seed removal was greatest from patches of bare ground in a lightly-grazed pasture with a heterogeneous structure. Ant colonies that removed fewest seeds were located in a prairie dog (Cynomys ludovicianus) colony where we commonly observed short-horned lizards (Phrynosoma douglassi), which are specialist predators on harvester ants. The second experiment served to partition variation in seed removal according to fine-scale effects of seed species, plant cover type, and distance of seed patches from ant nests. Two of five species of native seeds comprised 91% of all seeds harvested, with most seeds taken from patches of bare ground near ant nests.

To relate seed harvesting to ant foraging behavior, we mapped foraging movements of individual ants from colonies located in the experimental areas. The distance travelled from the nest and the foraging speed of ants differed among colonies in the three pastures, and were related to bare ground coverage near colonies. Individual ant movements partly explained differences in seed removal among fine-scale cover types, but processes operating at the population level, such as predation and grazing intensity, largely contributed to the variation in seed removal. This finding was supported by Monte Carlo simulations, which showed that variation in seed harvest changed nonlinearly across scales and was usually greatest above the level of the colony. Processes operating at the scales of individual foraging and patch use were different than those at broad scales, where Vegetation structure and predator abundance affected variation in seed-harvesting activity among ant colonies.

Our results suggest that processes, such as seed harvesting by ants, may change across spatial scales and hierarchical levels in response to factors that operate at different scales and levels. As a consequence, results obtained from a particular scale or level may be quite different from those obtained by a simple extrapolation of results from a smaller scale and lower level of organization.