A field experiment was established in northern Victoria in the autumn of 1999 to quantify the effects of renovating a 15-year-old, irrigated perennial pasture with a high paspalum content. The treatments were: (i) control, the existing pasture; (ii) oversown, the existing pasture grazed, topped and direct drilled; and (iii) resown, the existing pasture sprayed, cultivated and a new pasture sown. The grass species used in both renovation treatments were perennial ryegrass, Italian ryegrass and tall fescue. The treatments were grazed by dairy cows. The botanical composition, tiller density and plant frequency are reported in this paper.

Oversowing with either perennial or Italian ryegrass increased the sown grass content by an average of 4–8% DM, and by up to 20% DM during winter and spring, but did not affect the average white clover, volunteer species or dead contents. Oversowing did not affect the sown grass or paspalum tiller densities but annual oversowing with either perennial or Italian ryegrass increased the ryegrass plant frequency and, in 2 of the 4 years, reduced the white clover growing point density.

Resowing increased (P < 0.05) the content of sown grasses (by 12% DM for perennial ryegrass and by 20% DM for tall fescue) and white clover (by 8% DM) and decreased (P < 0.05) the content of volunteer species (by 18% DM) and, in years 1 and 2, dead material. The plant frequency of tall fescue increased over time (from 75 to 83% of quadrats), whereas that of perennial ryegrass declined (from 87 to 72% of quadrats). This resulted in the resown tall fescue having a lower (P < 0.05) white clover content and growing point density in years 3 and 4, and a tendency for a lower volunteer species content and tiller density in year 4, than the resown perennial ryegrass. Tall fescue is thus better able to resist the invasion of summer-active species and is less likely to require either oversowing or resowing than perennial ryegrass-based pastures. These results suggest that tall fescue should be considered as an alternative to perennial ryegrass when sowing pastures. The use of nitrogen did not improve the sown grass content, tiller density or plant frequency and will not therefore overcome problems resulting from a low ryegrass content.