Abstract of a paper based on work funded at least in part by the Australian Flora Foundation
Smith, M. A., Bell, D. T. and Loneragan, W. A.
Department of Botany, University of Western Australia, Nedlands, WA 6907, Australia.
Australian Journal of Ecology. 1999. 24: 35-42
CAB Abstract 990604063
A study was conducted during 1995 in the Yule Brook Botany Reserve, Perth, Australia, to investigate the germination ecology of the 2 grasses Austrostipa compressa and Ehrharta calycine in the understorey of Banksia woodland. The study compared soil seed bank densities at 11 sites in regions differing in age-since-last-fire (14 or 25 months or 19, 23, 37 or 45 years) and determined the influence of separate and multiple environmental cues on germination to compare the germination ecology and invasive ability of the 2 species. At each site 4 sample cores were taken and seeds/m2 calculated. Topsoil samples were placed in trays and subjected to one of 4 treatments (control, heat, smoke or heat + smoke) and germination and seedling viability assessed. Light quality cues for germination and the effect of exposure to smoked water on seed germination were also analysed. A. compressa was stimulated to germinate under a range of temperatures, in the presence of light and exposure to smoke-water. This combination of environmental cues resulted in winter-maximum germination in immediate postfire and disturbed-soil environments of this Mediterranean-type climate. In contrast, Ehrharta calycina (an introduced perennial grass from southern Africa that has invaded Banksia woodlands) germinated under a wide range of temperature and light conditions, but showed no promotive response to smoke-water. Although A. compressa seeds tolerated heat shock better than E. calycina, the self-burial mechanism of A. compressa seeds ensured protection from fire. It is suggested that high-intensity fire could have a greater impact on E. calycina, as the seeds of this species tend to accumulate in the top of the soil profile, and that survival of mature individuals by postfire resprouting ensures continued survival in native woodlands. Estimates of soil seed bank densities showed extreme variability, but some recently burnt areas of the reserve contained up to 8000 seeds/m2 of A. compressa and nearly 75 000 seeds/m2 of E. calycina. Viable soil seed bank densities of A. compressa were reduced with time-since-last fire, but areas of greater than 45 years since the last fire still contained up to 119 seeds/m2. In both species, about half their soil seed bank germinated following fire, thus ensuring the potential for later recruitment. It is concluded that massive soil seed populations of E. calycina in native Banksia woodlands pose a major problem to the management of this plant community type.