DYNAMICS OF PHOSPHORUS-COMPOUNDS IN A LOWLAND RIVER SYSTEM - IMPORTANCE OF RETENTION AND NONPOINT SOURCES

Intensive measurements of the fluxes of phosphorus (P) and of P retent ion were carried out in a Danish lowland watershed (Gjern Angstrom) du ring two hydrological years (June 1987 to May 1989). Seasonal and shor t-term variations of P concentrations were measured by intensive autom atic sampling covering P fluxes from the entire Gjern Angstrom watersh ed and two subcatchments: the Gelbaek and Lake Sobygard. Moreover, inf requent sampling was performed in all major tributaries of the river s ystem. Transport of P compounds (dissolved and particulate P) exposed a seasonal pattern which was highly affected by flow conditions, instr eam processes, as well as point and non-point sources. Stormflow P tra nsport during the two study years (1987-8 and 1988-9) constituted 56 a nd 61% in the Gelbaek, respectively, compared with 25 and 23% in the G jern Angstrom, respectively. Particulate P transport in the Gelbaek co nstituted 56 and 75% of annual total P transport during the two years compared with 77 and 75% in the Gjern Angstrom. The magnitude of month ly baseflow and stormflow total P loads was significantly correlated w ith the mean monthly discharge during baseflow and stormflow in the Ge lbaek, whereas only stormflow loads were correlated to discharge in th e Gjern Angstrom. In situ measurements of retention showed that the ma ximum amount of P stored in the Gelbaek during summer (June to August) was 22 g P m(-2) (1988) compared with 27 g P m(-2) (1987) in the Gjer n A. Lake Sobygard was a P source during summer but a P sink during wi nter and on an annual basis the net P released was 810 and 1420 kg, wh ich corresponded to 7 and 20% of the annual P export from the watershe d, respectively. Retention constituted about 30% of gross P transport in the Gelbaek during summer compared with 20% in the Gjern Angstrom. Resuspension of retained P during stormflows in September 1987 constit uted 94% of the stormflow P transport and 54% of the total P transport . Monthly total P and dissolved reactive P (DRP) mass balances for the main channel of the Gjern Angstrom revealed a significant DRP retenti on over 17 months (p < 0.05) caused by P adsorption on sediments and b iological uptake. On an annual basis the main channel was a sink of DR P (1220 and 1660 kg P) but a source of total P (3440 and 1000 kg P), i mplying that the channel is an important source of particulate P. Bank erosion proved to be a significant P source in the main channel of Gj ern Angstrom, whereas P delivery from soil erosion was possibly of min or importance. The annual non-point P export from two intensively samp led catchments was significantly higher (0.89 and 0.45 kg P ha(-1) yr( -1)) than from the five extensively sampled subbasins (0.43 and 0.17 k g P ha(-1) yr(-1)). No significant relationship could be established b etween non-point P export acid the proportion of arable land and soil type. Only for the wet year of 1987-8 was a significant relationship e stablished between non-point P export and the P load from scattered dw ellings outside sewage areas in five subbasins.