Comparison of methods for establishing native grasses in pastures dominated by annual weeds

Abstract

Native grasses have potential to improve temperate pastures where introduced perennial grasses are not surviving. They are generally well-adapted to Australia’s conditions including low-fertility or acidic soils, sporadic rainfall and high summer temperatures. However, they are difficult to establish from seed because of slow seedling development and vulnerability to competition from weeds, especially fast-growing annuals. Native grass practitioners in the Mount Lofty Ranges of South Australia have successfully established native grasses but their methods were poorly documented. For this thesis, 12 practitioners were interviewed and their native grass establishment methods and the problems surrounding these were documented. From the interviews and a review of the literature, a test of concept area and two field trials were established. The test of concept area was used to determine which species to use in the trials and how and when to sow them. Four native Wallaby grasses (Rytidosperma spp. Steud), Kangaroo grass (Themeda triandra Forsk.) and Weeping rice grass (Microlaena stipoides (Labill.) R. Br. and Microlaena stipoides var. Burra) established most successfully. Weed control was least time-consuming when the grasses were sown in rows rather than randomly distributed. Management was also simplified by separating C3 and C4 grasses. Soil solarisation with polyethylene and other plastics was also tested and it was found that solarisation can control annual weeds and seed found in the top 50 mm of soil. The first field trial was at Mylor, SA. It compared 7 weed control methods to determine which method created the most bare ground; an indication for a potential establishment window for native grasses. These methods were: removal of 50 mm of topsoil; soil solarization; soil inversion; till and harrow; herbicide; burning and harrowing. It was found that soil solarisation with polyethylene and topsoil removal were the most effective treatments with about 75% (± 3%) bare ground. There was least bare ground with burning (23% ± 4%) and herbicide (28% ± 4%). Till/harrow, harrow only and topsoil inversion ranged from 46-55% (± 3%) bare ground. There was no bare ground in the control. Since polyethylene is not recyclable in South Australia, a trial comparing the effectiveness of polyethylene and a fully biodegradable plastic was conducted in the Waite Arboretum, SA. The treatments included no treatment, tillage only and tillage with polyethylene of biodegradable plastic. All treatments except the control were sprayed with herbicide. The biofilm remained intact for 27 days. During this time, the mean daily temperature under the polyethylene (41.7 ± 0.4 °C) was always higher than under the biofilm (39.8 ± 0.3 °C). Both were hotter than the tilled treatment (34.1 ± 0.3 °C) and the control (33.9 ± 0.3 °C). Despite the higher temperature no measurable treatment effect could be detected by the end of the experiment but sown native grasses established well in all treatments with 30-50% native grass cover and very little weed. The lack of treatment effect was likely due to the small plot size, the use of herbicides to control some weeds and high seed bank variability within treatments. In summary, soil seed bank management is critical to successful native grass establishment. Topsoil removal and soil solarization with low density polyethylene were the most successful weed management methods. Other methods may need 2-3 years of treatment before sowing native grasses which increases the risk of soil erosion and may degrade the soil structure. Sowing the grasses in rows made weed management easier and sowing them thickly provided maximum weed competition. The cost and availability of native grass seed will be a significant barrier to the adoption of native grasses for pasture applications but on-farm seed production areas are one solution to this problem.