|dc.description.abstract||A lysimeter study was conducted to determine the fate of ¹⁵N-labelled urine applied to a pasture soil. The influence of subsoiling upon the fate of the nitrogen was also measured and the accuracy of selected soil nitrogen models tested.
A total of twelve undisturbed soil monolith lysimeters (800 mm diameter x 1100 mm deep) were extracted from an established pasture. Six of the lysimeters were sampled from an area which had previously been subsoiled to an average depth of 450 mm. The other six lysimeters were sampled from a non-subsoiled area of the same paddock. All of the lysimeters were installed in an underground lysimeter laboratory, which enabled normal environmental conditions to be maintained during experimentation. The edge of each lysimeter was sealed to prevent preferential water flow or root growth.
On 11th July 1990, a two litre solution of synthetic urine, containing the equivalent of 500 kg N ha⁻¹ labelled with 5 atom % ¹⁵N was applied evenly to the surface of each individual lysimeter, thus simulating a dairy cow urination event. Simulated rainfall (10 mm) was applied immediately after the urine to ensure that significant volatilisation losses of ammonia did not occur.
During the following year, the pasture was harvested periodically in order to determine the efficiency of urine nitrogen recovery by the pasture plants. Leachate samples were collected from each lysimeter after each 0.05 pore volume of drainage had occurred. A high leaching scenario was created by supplementing the received natural rainfall to ring the total water inputs up to the 75th percentile of the 100 year rainfall distribution over the winter and spring. During the summer months, border-dyke flood irrigation was simulated as per common district practice.
One full calendar year following the urine application, each lysimeter was carefully dissected into 50 mm depth increments and the amount of ¹⁵N remaining in the soil and roots was determined. A complete urine- ¹⁵N mass balance was therefore possible.
Subsoiling had no significant effect upon the amount of recovery of applied nitrogen by the pasture plants. Over the year of the experiment, an average of approximately 40% of the applied nitrogen was recovered by the pasture on both treatments.
A non-linear regress ion model, the logistic function, was used to test for significant differences between treatments for both the amount of drainage and the amount of nitrogen leached from each lysimeter. A significantly greater amount of water was found to have drained from the subsoiled (SS) as opposed to the non-sub oiled (NS) lysimeters (SS = 648 mm ; NS = 517 mm). Almost twice as much ¹⁵N was recovered in the leachate of the subsoiled lysimeters when compared with the non-subsoiled lysimeters (approximately 16% and 8% of the applied nitrogen respectively. A significantly higher amount of ¹⁵N was also recovered in the soil in the subsoiled lysimeters at the end of the experiment (SS = 26%; NS = 20%).
Overall, a greater amount of (he applied 15N was able to be accounted for at the end of the experiment in the subsoiled lysimeters (SS = 81%; NS = 72%). A smaller amount of the applied ¹⁵N was therefore presumed to have been lost by denitrification in the subsoiled lysimeters (SS = 19% ; NS = 28%). This was attributed to the increased aeration status which resulted from the subsoiling operation. Difference in drainage rates and in soil water potentials supported this hypothesis.
The accuracy of prediction of four selected simulation models was tested by comparing nitrate leaching losses measured in the experiment with those predicted by the models.||en