Phosphorus in surface runoff and drainage water affected by cultivation practices

To improve the water quality of Finnish lakes, rivers and coastal areas, knowledge is needed of management practices aiming to reduce losses of phosphorus (P) from agriculture. In the present study, the P load on the aquatic environment through surface runoff and drainage water was estimated by investigating water and soil samples taken from field experiments. The aim of the study was, first, to investigate errors associated with determinations of different P fractions in surface runoff and drainage water samples and, second, to highlight some cultivation practices and processes in runoff/soil interaction that might induce high P losses from agricultural soil to watercourses.

The experiments were carried out on soils originally quite poor in plant-available P at two experimental sites, one at Jokioinen, southwestern Finland, and one at Toholampi, western Finland. The Jokioinen site was situated on a heavy clay soil (Vertic Cambisol) with a 2% slope and the Toholampi site on a fine sand soil (Gleyic Podzol) with a 0.5% slope. The two 2-2.6 ha fields, with plots of 0.11-0.16 ha, had facilities to measure and collect surface runoff and drainage water, and allowed farming practices to be carried out with normal size machinery. The effect of different fertilizing and cropping practices on P losses was studied on both soils but the influence of subsurface drainage improvement only on the heavy clay soil. Water samples for investigating errors in P determinations were obtained from the heavy clay soil.

Suspended soil particles, which are typical of spring and autumn flow from cultivated clay soils, were the principal source of errors in determinations of different P fractions in the water samples. In peroxodisulphate digestion, which is used for determining total P, adding more sulphuric acid than is recommended in current standard methods increased the values by 5-7 % but they were still 6-11% lower than those obtained by HF digestion. In dissolved orthophosphate P determination, Nuclepore (0.2 and 0.4 um) and Sartorius Pen (0.45 um) filters produced the same results although the release of orthophosphate from colloids passing through the filters was not totally eliminated, even with Nuclepore 0.2 um filters. The presence of suspended particles led to considerable overestimation (77%) of orthophosphate if filtrates were stored and sulphuric acid was used as a preservative.

On both heavy clay and the fine sand, grass ley decreased particulate P loss compared with barley although the proportion of surface runoff was increased in grass ley cultivation. However, the dissolved othophosphate P losses were higher from ley, mostly due to surface application of fertilizers, which resulted in high dissolved orthophosphate P concentrations in surface runoff, indicating direct losses from applied P. On the fine sand, immediate ploughing after application of slurry in autumn reduced P losses efficiently. On the heavy clay, peak concentrations of orthophosphate P in drainge water indicated fast preferential movement of orthophosphate from the surface through macropores, while the podzolized fine sand profile efficiently retained P. Study of the fine sand showed that surface application of fertilizers on grass ley increased the amount of inorganic P and soil test P values in the top layer (0-5 cm), and the mean flow-weighted concentration of dissolved orthophosphate P in surface runoff rose linearly with the values of soil test P, thereby multiplying the P losses. Very shallow soil sampling (< 5 cm) is needed to assess the P loading potential in a soil with surface-applied P.

On the heavy clay, the contribution of drainage water to total discharge (drainage+surface runoff) rose from 10-40% to 50-90% when the subsurface drainage system was improved. The drainage water contained relatively high amounts of suspended soil particles and particulate P both before and after the drainage improvement, surface soil being the most probable origin of the particles. Contrary to wood chips, topsoil as backfill material was able to reduce particulate P losses from the field. The 25% lower losses of dissolved orthophosphate P after the improvement indicated efficient sorption of orthophosphate in the new drain trenches.