The influence of initial plant density on the changes in the populations of 3 perennial pasture species, lucerne (Medicago sativa L.), wallaby grass (Austrodanthonia richardsonii (Cashm.) H.P. Linder), and phalaris (Phalaris aquatica L.), over a 3-year pasture phase was examined in the wheatbelt of southern New South Wales. The perennials were sown at 5 rates in combination with subterranean clover (Trifolium subterraneum L.) at 2 locations, Kamarah [430 mm average annual rainfall (a.a.r.)] and Junee (550 mm a.a.r). The range in initial plant populations for lucerne, phalaris, and wallaby grass was 4–74, 8–94, and 2–20 plants/m2, respectively, at Kamarah and 11–120, 9–149, and 6–48 plants/m2 at Junee.
When sown at higher densities, the density of lucerne and phalaris declined curvilinearly over the 3 years at both sites. At the 3 lower densities, phalaris populations remained constant at both sites. Lucerne, in contrast, declined over all densities at both sites except at the lowest density at the wetter site (Junee). The rate of decline in lucerne was negatively related (R2 = 0.75) to initial density at Junee, but not at Kamarah. The density of the native grass, wallaby grass, increased with time at both sites through seedling recruitment.
The invasion of experimental plots by the summer weed Eragrostis cilianensis (All.) Vign. ex Janchen (stinkgrass) was restricted by lucerne and phalaris, with a negative curvilinear relationship between perennial density and E. cilianensis seedlings in both environments (R2 = 0.65–0.70). In contrast, wallaby grass was ineffective at suppressing E. cilianensis.
By the third year, phalaris had significantly higher herbage yields in spring than lucerne and wallaby grass at both sites and phalaris yield was independent of density. Lucerne yields at this time increased with density only at the wetter site (R2 = 0.64), but wallaby grass yields responded to increasing density at both sites (R2 = 0.27–0.59).
The experiment demonstrated that establishing higher initial perennial populations of lucerne and phalaris, which did not recruit during the experiment, will result in the maintenance of higher populations over the life of a 3–4 year pasture phase despite proportionally higher rates of plant loss. The size of the initial population was less critical for wallaby grass, which was able to increase in density through recruitment. High initial populations are likely to be an advantage for suppressing weeds in swards of perennial species with limited ability to increase their basal area, such as lucerne and wallaby grass, but will be less beneficial for species such as phalaris, which can compensate by greatly increasing its basal area at lower densities.