Variation of soil P and other nutrients in a long-term grazed grassland P experiment field

This is the authors' abstract. We don't add key points for this paper.

Field-scale variation of soil phosphorus (P) information is very important for P fertilizer application and its soil sampling design in grassland. A total of 108 soil samples were collected from a long-term (41 years) grazed grassland P experiment field at Teagasc, Johnstown Castle, Wexford, Ireland, in March 2009. There were six P treatments (P0-0, P0-30, P15-15, P15-5, P30-30, and P30-0) since 1968, with changes since 1999. Each treatment had 6 replicate plots (a total of 36 plots, 3 soil samples per plot). The samples were analyzed for available (Morgan’s) P, potassium (K), magnesium (Mg), lime requirement (LR), and pH. The highest mean available P concentration was found in the P30-30 (30 kg P ha−1 pre- and post-1998) plots, and the lowest mean available P concentration was found in the P0-0 (no P fertilizer since 1968) plots. Significant differences of mean P, Mg, LR, and pH values in different treatments were observed. There was a positive proportional effect for both the 36 plots and the 6 treatments for the P data: the local standard deviation increased with the increase of local mean. The proportional effect should be considered in order to optimize sampling design. Fewer samples can reflect soil P status in fields with low soil P levels, while more attention should be paid to the fields with high P levels in order to reduce environmental consequences of uniform applications.