LAUSR

Farm-gate balances for nitrogen (N) and phosphorus (P) were used to determine the potential nutrient surplus and nutrient-use efficiency of three grassland-based milk production systems. These were either full-time grazing (FG) or part-time grazing with indoor feeding of fresh herbage with low (IF) or substantial concentrate allowance (IFplus). Data collected over 3 years were used (a) from 3 × 8 commercial dairy farms, and (b) from a controlled experiment. On FG, IF and IFplus farms, the N surplus (net input minus net output) was 78.6, 75.1 and 138.2 kg N ha − 1 a − 1 , respectively, and N use efficiency (NUE; net output/net input) was 0.33, 0.44 and 0.34, respectively. Annual P inputs and outputs were relatively balanced, with a higher P use efficiency (PUE) found for FG (1.19) and IF (0.97) farms than for IFplus farms (0.57). The most important input sources, apart from biological N 2 fixation and atmospheric N deposition, were mineral fertilizer N, concentrate N/P in IFplus, and animal manure P. Results from the controlled experiment confirmed most of these findings. In general, the IF provided an efficient nutrient use at a relatively low surplus. Accounting for off-farm losses associated with feed inputs and manure export reduced the system NUE and PUE substantially. The large variability in nutrient-use efficiency among the commercial farms indicates a clear potential for improvement. The average N surplus could be reduced by approximately 40% in reference to the best performing farms in each system.