With any grazing method, the grazing pressure applied to an individual plant is a site, stock density and time dependent variable and the diet selection hierarchy of grazing animals is to the disadvantage of the most palatable and actively growing pasture components. The greater the differences in palatability and abundance among the components of a sward, and the lower the stock density, the greater the variation in the grazing pressure exerted. These effects are heightened when animals are set-stocked under adverse environmental conditions. This paper reports the comparative effects of cell grazing and continuous grazing on pasture composition on three properties on the Northern Tablelands of New South Wales. The basal diameters, relative frequency and contribution to dry weight of the most desirablelpalatable species at each site were found to remain constant or to increase under cell grazing, while declining significantly under continuous stocking. The converse was true for the least palatable components of the pasture, which declined significantly under cell grazing but changed little under continuous grazing. Percentage ground cover was significantly higher after two years of cell grazing than under continuous grazing. These changes in pasture composition may have long-term benefits with respect to erosion control, nutrient cycling, hydrological function and the stability of animal production at the cell grazed sites.

Summary. Four experiments were conducted to determine the effects of temperature, light and leaf extract solutions on the germination of Giant Parramatta grass [GPG, Sporobolus indicus (L.) R. Br. var. major (Buse) Baaijens] collected from a population on the North Coast of New South Wales. In the first experiment, seeds were subjected to one of a range of temperature combinations immediately after collection and again after 8 and 27 weeks. Germination was restricted to a narrow range of alternating temperatures with a peak at 35°C day/15°C night when seeds were tested immediately after collection. More seeds germinated when the samples had been stored, although germination remained depressed at constant temperatures. These data indicate that freshly collected GPG seeds are subject to primary dormancy and that few would germinate in the field immediately after seed fall. In a second experiment, seeds were buried beneath leaf litter in a pasture immediately after collection. After 7 months, the seeds were exhumed and subjected to either constant (20°C) or alternating (35/15°C) temperatures in either full light, reduced red:far-red (R : FR) light or dark treatments. Over 95% of GPG seeds germinated when subjected to alternating temperatures, regardless of light treatment. At constant temperatures, 97% of seeds germinated under full light, 59% at reduced R : FR light and <1% in dark treatments. A germination response to alternating temperatures and/or light treatments has been reported in pasture weeds and may be an adaptation to detecting gaps in the pasture canopy. Consequently, the germination of GPG in a pasture may be manipulated to some extent by altering the amount of pasture cover using grazing management, mowing and fertiliser applications. In experiment 3, leaves from a range of coastal grasses were mixed with water and the solutions were used to germinate GPG seeds. Solutions extracted from setaria (Setaria sphacelata) leaves completely inhibited GPG germination while 27% of GPG seeds germinated when imbibed with kikuyu leaf extract solution. Solution extracted from carpet grass (Axonopus affinis) leaves had the least effect on GPG germination. In experiment 4, the effects of solutions that had been leached from the leaves of either setaria or carpet grass on seed germination, and root and shoot lengths of GPG seedlings were compared. Germination was less inhibited by leachate solutions compared with the extract solutions used in experiment 3. Seedlings in setaria leachates had significantly shorter roots and shoots than both those germinated in carpet grass leachates and control seedlings. This may explain, at least in part, why carpet-grass-based pastures are readily infested with GPG while setaria-based pastures are relatively resistant to infestation. The potential for allelopathic interactions between GPG and setaria to be fully utilised to reduce the abundance of GPG in coastal New South Wales pastures is discussed.

Inputs and losses from Giant Parramatta grass [GPG, Sporobolus indicus (L.) R. Br. var. major (Buse) Baaijens] soil seed banks were quantified on the North Coast of New South Wales. Monthly potential seed production and actual seed fall was estimated at Valla during 1991-92. Total potential production was >668 000 seeds/m2 for the season, while seed fall was >146000 seeds/m2. Seed fall >10000 seeds/m2.month was recorded from January until May, with further seed falls recorded in June and July. The impact of seed production on seed banks was assessed by estimating seed banks in the seed production quadrats before and after seed fall. Seed banks in 4 of the 6 sites decreased in year 2, although seed numbers at 1 damp site increased markedly. Defoliation from mid-December until February, April or June prevented seed production, reducing seed banks by 34% over 7 months. Seed banks in undefoliated plots increased by 3300 seeds/m2, although seed fall was estimated at >114 000 seeds/m2. Emergence of GPG seedlings from artificially established and naturally occurring, persistent seed banks was recorded for 3 years from bare and vegetated treatment plots. Sown seeds showed high levels of innate dormancy and only 4% of seeds emerged when sown immediately after collection. Longer storage of seeds after collection resulted in more seedlings emerging. Estimates of persistent seed banks ranged from 1650 to about 21260 seeds/m2. Most seedlings emerged in spring or autumn and this was correlated with rainfall but not with ambient temperatures. Rates of seed bank decline in both bare and vegetated treatment plots was estimated by fitting exponential decay curves to seed bank estimates. Assuming no further seed inputs, it was estimated that it would take about 3 and 5 years, respectively, for seed banks to decline to 150 seeds/m2 in bare and vegetated treatments.