A field experiment was established in a southern Australian temperate pasture to investigate the effects of identity and proximity of perennial grasses on the demography of the annual grasses Vulpia spp. ( V. myuros, V. bromoides) and Hordeum leporinum (barley grass). Annual grasses were grown either alone or in mixtures, at different distances from rows of Dactylis glomerata (cocksfoot) and Phalaris aquatica (phalaris). Dactylis had a greater suppressive effect than Phalaris on Vulpia and Hordeum. Biomass, tiller production, and panicle production of annual grasses increased linearly with increasing distance from the perennial row. Tiller and panicle production were greater for Vulpia than Hordeum. The estimated rate of population growth (lambda) for annual grasses was greater in Phalaris than in Dactylis and in Vulpia than in Hordeum, and increased with sowing distance from perennial grass rows. It was estimated that lambda, when seeds were sown directly adjacent to a row of perennial grasses, was 1 and 0.4 for Vulpia and Hordeum, respectively, within Dactylis stands, and 7 and 3, respectively, within Phalaris stands. However, 15cm from the row, lambda reached 50 and 39 for Vulpia and Hordeum, respectively, within Phalaris stands, and 39 and 16, respectively, within Dactylis stands. In grazed, dryland pastures, perennial competition alone is therefore unlikely to prevent population growth of annual grasses, especially in systems heavily disturbed by grazing or drought. However, Dactylis showed more promise than Phalaris in limiting the abundance of these weeds.
