Nutrient balance at chain level: a valuable approach to benchmark nutrient losses in milk production?

A nutrient balance (NB) approach is often used to quantify losses of nutrients, such as nitrogen (N) and phosphorus (P), that contribute to environmental problems such as eutrophication. An NB generally is computed at farm level, whereas nutrient losses related to off-farm processes are neglected. Using an NB at farm level to compare systems that differ in, for example, amount of purchased concentrates, however, may lead to biased conclusions. We, therefore, analysed 19 Irish grass-based dairy farms (IR) and 13 Dutch concentrate-based dairy farms (NL). For each farm, we computed N and P losses at farm and chain level, expressed per ton fat-and-protein-corrected milk (FPCM) to investigate whether an NB at chain level provides better insights than an NB at farm level. Data were analysed using an independent T-test, a Wilcoxon-Mann-Whitney test and regression analysis. Results show that on average, IR farms had higher N losses per ton FPCM than NL farms, both at farm (IR=20; NL=8 in kg N/ton FPCM) and chain level (IR=22; NL=11 in kg N/ton FPCM). P losses per ton FPCM, on the other hand, did not differ between IR and NL farms at farm (IR=0.3; NL=0.1 in kg P/ton FPCM) or chain level (IR=0.8; NL=1.0 in kg P/ton FPCM). Results of the regression analysis showed that the NB at chain level could be accurately predicted from the NB at farm level (R2=0.992 for N; R2=0.910 for P), whereas in case of P the slope tended to differ between IR and NL farms (p