A field study was conducted (1994–97) to investigate defoliation time and frequency effects on a newly sown stand of phalaris (Phalaris aquatica cv. Sirosa). Defoliation treatments were applied from spring 1994 to autumn 1995 and effects on plant frequency and basal cover monitored until June 1997. The experimental design was a randomised split-plot in three replicates, with 45 main plots and 90 subplots. Main plots included three types of defoliation (two cutting heights or grazing with Merino sheep) applied at four different times (early stem elongation, anthesis, end of summer and end of autumn) and an undefoliated control. The main plots were divided into two subplots for frequency of defoliation (once only or cut every 8 weeks).

In September 1995, none of the treatments had a significant effect on plant frequency, basal cover, green tiller weight and weight of individual tillers. However, there were significant (P < 0.05) effects of defoliation type on green tiller number, defoliation time on basal bud weight and water soluble carbohydrates and defoliation type and frequency on basal bud water soluble carbohydrates.

In February and April 1996, there were significant effects of defoliation time on green tiller number and weight, basal bud weight and water soluble carbohydrates. In 1996, Sirosa phalaris plant frequency and basal cover values were generally higher (P < 0.05) in the undefoliated control compared with those defoliated at early stem elongation. However, despite fertiliser application and moderate grazing, Sirosa phalaris plant frequency had declined to a mean of 23.6% by February 1997 and to <10% by June 1997.