| Abstract: | Field germination behaviour and seedbed environment of the annual legume Townsville stylo (Stylosanthes humilis) and the vigorous annual grass weed (Digitaria ciliaris) were studied at Katherine, Northern Territory. Previous studies of the population dynamics of these pastures had indicated the importance of the germination-establishment phase in maintaining S. humilis as a pasture component. Analysis of seed samples at the time of sward establishment showed that near complete germination of viable seed (in the case of D. ciliaris) and of soft seed (in the case of S. humilis) occurred in the early storm period before the onset of the wet season. Swards of both species generally resulted from a single, major germination wave often after an isolated rainfall. The seedbed environment at the time of major germination waves was studied in detail for 2 years and environmental models were used with historical climate records to calculate probability distributions of temperature and moisture over 16 years. The main determinants of the seedbed environment were storms (most often occurring in the period from late afternoon to midnight) and high solar radiation levels (70% > 20 MJ m-2 day-1 in the days after rainfall. Surface soil generally dries to -1.5 MPa by noon. Minimum night surface temperatures showed little variation (～25°C) and were followed by high surface temperatures the following day (> 35°C). The use of a simple water balance model showed a high chance (80% of events) of the top 40 mm drying to – 1.5 MPa at the end of the second day following rainfall. Despite the low variation of annual rainfall at Katherine (cv of 20%) these self regenerating annual species have a high risk of seedling death. Germination and seedling establishment take place at a time of rapid fluctuations in surface moisture and seedbed temperature. The analysis of seedbed conditions indicated that intensive sampling of the seed population would be required in the early storm period to explain differential survival of the species and that laboratory work should concentrate on factors controlling speed of germination over short periods of available moisture (12–20 h). |
