Nitrate leaching from a drained, sheep-grazed pasture. II. Modelling nitrate leaching losses

Abstract

Nitrate (NO^-₃ ) concentrations in 0·5-mm increments of drainage from adjacent mole- and pipe-drained paddocks of a silt loam soil under pasture near Palmerston North, New Zealand, were measured during 2 winters. The data were simulated using a simple analytical transfer function model (TFM). Urea fertiliser applied at the rate of 120 kg N/ha to one paddock was treated as a pulse input to the pool of resident soil NO^-₃. A source{sink term was included for plant uptake and net mineralisation (including any effect of denitrification).

During the first winter (1990), a TFM using either a 1-parameter Burns probability density function (pdf) for solute travel, or a 2-parameter lognormal pdf, satisfactorily simulated the NO^-₃ concentration trends and predicted the total amounts of N leached. The pdf parameters were derived from previous chloride leaching data for this site. The best-fit value for the transport volume &thgr;_st, the key parameter in the Burns pdf, was set at 0·37 m³ /m³ in 1990, as used in previous modelling of sulfate leaching. However, a value of 0 ·25 m³ /m³ in the Burns pdf gave better simulations of the 1991 data. This was probably due to more intense rain events during the early part of the drainage season in 1991 compared with 1990, which resulted in more preferential flow through the soil and a lower value for &thgr;_st. The simulations for both years showed that ≥50% of the total leachable NO^-₃ was retained in the soil, despite normal winter drainage of about 300 mm.

Ideally, the appropriate value of st should be determined by independent measurement. It may need to be adjusted according to the likely incidence of preferential flow early in the winter when NO^-₃ concentrations are highest. Provided the average initial soil NO^-₃ concentration can be estimated and a net source{sink term defined, the amount of NO^-₃ leached in drained soils can be satisfactorily modelled using the TFM approach with a 1-parameter pdf. Duplex soils which have a fluctuating watertable in the A horizon over an impermeable B horizon may prove to be an analogous system.