An Appraisal of Potential for Sowing of Nasturtium officinale into Streams to Mitigate Nutrient Pollution in Eastern Scotland

Abstract

This study examines a farmer-led initiative to sow watercress (Nasturtium officinale) in field ditches. The objective was to assess the potential of this practice to mitigate summer nutrient loads in rivers. Two ditches—one seeded, the other unseeded—on a mixed-livestock farm in Eastern Scotland were monitored during the spring-summer of 2014–2016. The un-replicated trial design limited statistical analysis. However, changes in N and P concentrations along the two ditches were measured. In the watercress-seeded ditch, N retention of 0.092 g/m2/d (p < 0.001, SE = 0.020) and P retention of 0.0092 g/m2/d (p = 0.001, SE = 0.0028) occurred, while total organic C in the water increased along the ditch. Retention was close to zero for the unseeded ditch. The seeded ditch was also found to have more dry matter production and lower stream temperature. The impact of plastic covering (to increase spring temperature) on vegetation and nutrient removal was also assessed on replicate 5-m sections of the ditches. No significant impact on N and P removal was found; however, the release of C increased significantly in the plastic-covered sections. The rise in air temperature (up to > 30 °C) promoted a greater growth of opportunist species (nettle (Urtica), rush (Juncus), and grasses. These observations were used to make a simple assessment of the potential catchment scale impact of seeding watercress into first and second order streams in the nearby Lunan Water catchment. It was concluded that this could make a significant contribution to the reduction of nutrient loads.

An evolutionary Monte Carlo method for the analysis of turbidity high‐frequency time series through Markov switching autoregressive models

“…For key pollutants (e.g., suspended sediment or particulate phosphorus), such evidence is difficult to obtain, because of the cost of sampling and chemical analysis of storm event driven changes in concentrations and loads in streams and rural drainage features. Previous work has investigated the use of continuous automated turbidity as a proxy to estimate particulate phosphorus, fine sediment or hydrophobic pollutant loads (e.g., Minella et al, 2008;Rügner et al, 2014;Stutter et al, 2017) using site specific calibrations of turbidity versus the pollutant of interest, with some success (Lannerg ȧrd et al, 2019;Vinten & Bowden-Smith, 2020). However, the approach is limited by only moderately good correlations, and therefore methods which enhance the fit, for example between turbidity and phosphorus loads, are needed.…”