Measurement and modelling of runoff and phosphate movement from seasonally dry hill‐country pastures

Abstract

In seasonally dry hill pastures on a Waipawa stony loam (Pallic Soil) in Hawke's Bay, a range of methods was used to measure phosphorus (P) movement and losses in runoff water. Measurements with a rainfall simulator and micro‐plots showed that dissolved reactive P (DRP) and total dissolved P (TDP) concentrations in surface runoff were directly related to soil Olsen P (R2= 0.93 and R2 = 0.90 respectively) status. The distance that surface runoff travelled, as measured by P fertiliser transport, was about 5 m when rainfall in autumn had not occurred for 12–22 days, but decreased sharply in subsequent rainstorms. Over a 12‐month period most natural surface runoff occurred mostly during early summer to late autumn and was inversely related to soil moisture content (R2= ‐0.38 for all data). Most (55–84%) of total P (TP) losses occurred in three of 12 storms measured, with the majority as DRP from low P (60–85%) and high P (70–93%) soils. Continuous flows from twin catchments of 12.6 ha (north, with low soil P status), and 12.8 ha (south, with a moderate to high soil P status) were recorded. For the north catchment, total flow ranged from 18 to 26% of annual rainfall, and for the south catchment from 18 to 21%, over 2000 to 2002. About half of this occurred in July‐August. Over a 7‐month period, surface runoff was equivalent to only about 3% of rainfall. The difference in soil P status of the two catchments provided an approximate index of the difference in TP export. The DRP and TDP concentrations in catchment flow were about 10‐fold less than those recorded in surface runoff from associated pastures. The TOPMODEL runoff simulation model accurately estimated total flow from two catchments (R = 0.89 and 0.87 for north and south catchments respectively), but not the contribution by surface runoff.