Beschreibung:
<jats:title>Abstract</jats:title><jats:p>Ants often appear to be important post‐dispersal seed predators, particularly in Australia where they are exceptionally abundant and apparently can remove large quantities of seeds from the ground. Rates of seed removal by ants usually are measured by recording removal from artificial seed baits, but the reliability of this approach has not been tested, nor have there been many attempts to integrate the results with the activity of seed‐eating ants. This paper describes the rates of seed removal, estimated using a baiting technique that is tested for its reliability, by the seed‐eating ants in adjacent heath and woodland sites at Wilson's Promontory, Victoria. Ants removed up to 100% of seeds, but rates varied according to seed species, size of seed clumps, season, time of exposure, and other aspects of the baiting technique. Methodological guidelines are provided to make baiting conditions approximately those likely to occur in nature.</jats:p><jats:p>Seed‐eating ants, particularly species of<jats:italic>Rhytidoponera, Chelaner and Pheidole</jats:italic>, were by far the most important post‐dispersal seed predators, and patterns of seed removal were directly related to their composition, abundance and foraging behaviour. Lygaeid bugs were also observed eating seeds, but there was no evidence of seed predation by rodents or birds.</jats:p><jats:p>The results suggest that seed predation by ants can substantially deplete seed reserves: however, its actual effect on seedling recruitment is likely to depend on many factors including seed size, crop size, weather, timing and location of seed fall, availability of alternative food sources, patterns of seedling mortality, and fire, none of which have been adequately investigated.</jats:p>