Managing competitive interactions to promote regeneration of native perennial grasses in semi-arid south-eastern Australia

摘要: Treatments to reduce available soil nitrogen and achieve specified levels of weed control were evaluated for their capacity to promote regeneration of native perennial grasses in a degraded semi-arid woodland in central-western New South Wales. Treatments were factorial combinations of nitrogen-reduction levels and weed-control levels. The four levels of nitrogen reduction were no intervention, and oversowing of an unfertilised summer crop, an unfertilised winter crop or an unfertilised perennial grass. The three weed-control levels were defined by the outcome sought rather than the chemical applied and were nil, control of annual legumes and control of all annual species (AA). Regeneration of perennial grasses, predominantly Enteropogon acicularis, was promoted most rapidly by the AA level of weed control with no introduction of sown species. Sown species negated the benefits of weed control and limited but did not prevent the regeneration of native perennials. Sown species also contributed substantially to biomass production, which was otherwise severely limited under the AA level of weed control, and they were effective in reducing soil nitrogen availability. Sown species in combination with appropriate herbicide use can therefore maintain or increase available forage in the short–medium term, permit a low rate of native perennial grass recruitment, and condition the system (by reducing soil mineral nitrogen) for more rapid regeneration of native perennials should annual sowings be discontinued or a sown grass fail to persist. Soil nitrate was reduced roughly in proportion to biomass production. High levels of soil nitrate did not inhibit native perennial grass regeneration when biomass was suppressed by AA weed control, and may be beneficial for pastoral production, but could also render sites more susceptible to future invasion of exotic annuals. The need for astute grazing management of the restored grassland is thus emphasised. This study was conducted on a site that supported a remnant population of perennial grasses. Use of the nitrogen-reduction techniques described may not be appropriate on sites where very few perennial grass plants remain.