Nutrient Speciation Dynamics in UK Rivers: Seasonality and Flow Controls

To investigate nutrient transport and transformations in UK freshwaters, water samples were collected from a range of sites on three rivers: the River Windrush in the Cotswold region; the River Cole, Wiltshire; and the River Bure, the major inflow to the Norfolk Broads. Discharge was determined from continuous records for gauging stations within each catchment. All water samples were analysed to determine suspended sediment concentrations, and the concentrations of all N species and P fractions, following the protocol outlined by Johnes & Heathwaite (1992). Samples were collected daily and during storm events at two sampling sites in the Windrush catchment over two years: 1987/88 and 1988/89. In the River Cole, samples were collected at two sites from January 1995 to December 1996, initially on a weekly basis, after August 1995 onwards daily. In the Bure Programme, samples were collected weekly at seven sites from February 1995 to February 1996.

The results from this research provide an insight into the rates and controls on nutrient cycling and transport in lowland rivers in the UK, and useful information for the design of future sampling programme. The topography of the Windrush catchment comprises a limestone plateau deeply incised by the River Windrush and its major tributaries. The relief is gently undulating, with rich soils, supporting intensive mixed arable agriculture. The river regime is characterised by a winter maximum and summer minimum. Nutrient concentrations in the Windrush are strongly controlled by both flow and the sediment delivery pathway. Nitrate and NO2- and PO43- concentrations show a strong seasonality, correlated with seasonal variations in baseflow. Suspended sediment concentrations were correlated with storm events, leading to increases in the concentrations of sediment-associated nutrient fractions (Particulate Organic N, Particulate P, NH4+) during stormflow periods. The Dissolved Organic N (DON) and Soluble Unreactive P (SUP) fractions show no strong seasonal trends, but form the dominant nutrient fractions in the summer months when NO3-, and sediment-associated nutrient concentrations are low. Overall, whilst NO3- is the dominant N species contributing to the total N load transported in the Windrush, Organic N is also a significant secondary source, contributing 40% of the total N load in 1988/89. The total P load is dominated by the delivery of Particulate P (PP) during storm events, with the remainder of the load delivered in soluble form (SRP- Soluble Reactive P, SUP). The data suggest that where nutrient control programme are planned, routine monitoring should include all N species and P fractions, if an accurate measure of the total nutrient load delivered to the system is to be determined.

These patterns are repeated in the data for the Cole catchment, which has similar topography and land use to the Windrush. The River Cole data show that the majority of the N load is delivered along subsurface throughflow pathways as NO3- and DON during winter months, with PP delivery during storm events dominating the total P load, although SRP is the dominant P fraction in the summer months when low flows from the catchment, combined with continued SRP input from Sewage Treatment Works lead to increased SRP concentrations. The switch from weekly to daily sampling highlights the extent to which nutrient loads delivered during storm events may be missed in weekly sampling programme.

The trends for the River Bure show some similarities to those for the Windrush and Cole, but also highlight the limitations of weekly sampling in deriving statistically robust measures of nutrient transport. The Norfolk region is very flat, intensively farmed, and relatively dry in the UK context. The soils are very light and friable and are easily eroded by wind and water. The flat landscape of Norfolk also means that rivers have a very low channel gradient, and in combination with low rates of discharge, this means that the rivers are slow flowing, have a high sediment trapping efficiency, with the majority of the nutrient load transport during a few extreme storm events during the year. Much of the nutrient load exported from non-point agricultural sources to the rivers is trapped in the river bed during baseflow conditions. This trapped sediment and associated nutrient load is not static, however, and during stormflow, is resuspended and transported downstream, further contributing to the total nutrient load in the water column. This means that in Norfolk rivers, current weekly water quality monitoring schemes systematically underestimate nutrient loadings, especially for P which is largely exported as PP. Where daily sampling is not possible in nutrient budgeting studies, flow-proportional sampling is essential if an accurate measure of nutrient transport is to be determined.