Fertiliser phosphorus (P) and, more recently, fertiliser nitrogen (N) are regularly applied to intensively grazed dairy pastures in south-western Australia. However, it is not known if applications of fertiliser N change pasture dry matter (DM) yield responses to applied fertiliser P. In three Western Australian field experiments (2000–04), six levels of P were applied to large plots with or without fertiliser N. The pastures were rotationally grazed. Grazing started when ryegrass plants had 2–3 leaves per tiller. Plots were grazed in common with the lactating dairy herd in the 6-h period between the morning and afternoon milking.

A pasture DM yield response to applied N occurred for all harvests in all three experiments. For the two experiments on P deficient soil, pasture DM yield responses also occurred to applications of P. For some harvests when no fertiliser N was applied, probably because mineral N in soil was so small, there was a small, non-significant pasture DM response to applied P and the P × N interaction was highly significant (P < 0.001). However, for most harvests there was a significant pasture DM response to both applied N and P, and the P × N interaction was significant (P < 0.05–0.01), with the response to applied P, and maximum yield plateaus to applied P, being smaller when no N was applied. Despite this, for the significant pasture DM responses to applied P, the level of applied P required to produce 90% of the maximum pasture DM yield was mostly similar with or without applied N. Evidently for P deficient soils in the region, pasture DM responses to applied fertiliser P are smaller or may fail to occur unless fertiliser N is also applied. In a third experiment, where the soil had a high P status (i.e. more typical of most dairy farms in the region), there was only a pasture DM yield response to applied fertiliser N.

We recommend that fertiliser P should not be applied to dairy pastures in the region until soil testing indicates likely deficiency, to avoid developing unproductive, unprofitable large surpluses of P in soil, and reduce the likelihood of P leaching and polluting water in the many drains and waterways in the region. For all three experiments, critical Colwell soil test P (a soil test value that was related to 90% of the maximum pasture DM yield), was similar for the two fertiliser N treatments.