Using a modelling approach to evaluate two options for improving animal nitrogen use efficiency and reducing nitrous oxide emissions on dairy farms in southern Australia

Ruminant livestock are generally considered inefficient converters of dietary nitrogen (N) into animal product. Animal nitrogen use efficiency (NUE) is a measure of the relative transformation of feed N into product and in dairy systems this is often expressed as milk N per unit of N intake (g milk N/100 g N intake). This study was a theoretical exercise to explore the relative potential efficacy and value proposition of breeding versus feeding to improve NUE, reduce urinary N excretion and associated environmental impact in pasture-based dairy systems. The biophysical whole farm systems model DairyMod was used across three dairying regions of south-eastern Australia representing a high-rainfall cool temperate climate (HRCT), a high-rainfall temperate climate (HRT) and a medium-rainfall temperate climate (MRT) to examine the two theoretical approaches of (1) maintaining the same amount of N exported in milk from a reduced N intake; and (2) increasing the amount of N exported in milk for the same amount of dietary N intake. Sixteen scenarios were explored for each site; these include four supplementary feed N (SN) concentrations (ranging from 1% to 4% N) combined with four milk N (MN) concentrations (ranging from 0.50% to 0.65% N). Reducing the SN concentration from 4% to 1% increased the 30-year mean model-predicted NUEs from ~16 g milk N/100 g N intake at all three sites to between 23 and 28 g milk N/100 g N intake, with the least and greatest improvements in NUE occurring for the HRCT and MRT sites, respectively. Corresponding to this improved NUE through reduced SN concentrations, model-predicted N2O emissions declined from 3.0 to 1.3 t carbon dioxide equivalents (CO2-e)/ha.annum for the HRCT site, from 4.2 to 2.1 t CO2-e/ha.annum for the HRT site and from 4.4 to 2.1 t CO2-e/ha.annum for the MRT site, representing a decline of between 50% and 57%. In contrast, increasing the MN concentration from 0.50% to 0.65% increased the 30-year mean model-predicted NUEs from 17 to 22 g milk N/100 g N intake for the HRCT site, from 18 to 23 g milk N/100 g N intake for the HRT site and from 18 to 24 g milk N/100 g N intake for the MRT site. Corresponding to the improved NUE through increased MN concentrations, model-predicted N2O emissions declined from 2.3 to 2.0 t CO2-e/ha.annum for the HRCT site, from 3.3 to 3.1 t CO2-e/ha.annum for the HRT site and from 3.4 to 3.2 t CO2-e/ha.annum for the MRT site; representing a decline of between 7% and 11%. These results suggest that improving animal NUE to reduce associated N2O losses holds much more promise if achieved through a reduction in the amount of N in supplementary feed than through increasing N exported in milk. This is an important finding for the Australian dairy industry, since manipulation of dietary N to better balance the energy to protein ratio would be much easier to implement than manipulation of N concentration in milk through genetics.