Eutrophication of coastal areas is a global problem. A full-scale coastal remediation project was initiated in Björnöfjärden bay in the Stockholm archipelago in 2011. Measures to reduce external nutrient inputs from the surrounding catchment (15 km2) targeted agriculture, on-site wastewater treatment facilities, and horse keeping. The effects were evaluated at 22 water quality monitoring stations over 11 years (2012-2022) to determine temporal trends in nutrient concentrations, spatial correlations within and between monitored sub-catchments, and effects of individual mitigation measures at local and catchment scale. The effect of individual measures varied from no significant effect to significant nutrient decreases (21% reduction in dissolved P concentrations in one lime filter) or increases (11% higher concentrations in total P in one constructed wetland). However, few significant trends were detected at sub-catchment outlet stations. Tailored placement, design, dimensioning, and maintenance of implemented mitigation measures are needed to improve their nutrient retention effect.

Nordic agriculture faces big challenges to reduce phosphorus (P) loss from land to water for improving surface water quality. While understanding the processes controlling P loss and seeking for P mitigation measures, Norwegian and Swedish researchers have substantially benefited from and been inspired by Dr. Andrew Sharpley's career-long, high-standard P research. Here, we demonstrate how Sharpley and his research have helped the Nordic researchers to understand the role of cover crops in cold environmental conditions, best manure P management practices, and ditch processes. His work on critical source area (CSA) identification and site assessment tool development have also greatly inspired our thinking on the targeting of mitigation measures and the contextualizing tools for Nordic climate, landscape, and soils. While reflecting on Sharpley's legacy, we identify several needs for Norwegian and Swedish P research and management. These include (1) tackling the challenges caused by local/regional unevenness in livestock density and related manure management and farm P surpluses, (2) identifying CSAs of P loss with high erosion risk and high P surplus, (3) obtaining more high-resolution mapping of soils with low P sorption capacity both in the topsoil and subsoil, (4) improving cross-scale understanding of processes and mitigation measures and proper follow-up of applied mitigation measures, and (5) increasing collaborations of researchers with farmers and farmers' advisory groups and watershed groups by developing high-quality educational courses and extension materials. The needs should be addressed in the context of the challenges and opportunities created by climate change.

Abstract. Agricultural headwater streams are important pathways for diffuse sediment and nutrient losses, requiring mitigation strategies beyond in-field measures to intercept the transport of pollutants to downstream freshwater resources. As such, floodplains can be constructed along existing agricultural streams and ditches to improve fluvial stability and promote deposition of sediments and particulate phosphorus. In this study, we evaluated 10 remediated agricultural streams in Sweden for their capacity to reduce sediment and particulate phosphorus export and investigated the interplay between fluvial processes and phosphorus dynamics. Remediated streams with different floodplain designs (either on one side or both sides of the channel, with different width and elevation) were paired with upstream trapezoidal channels as controls. We used sedimentation plates to determine seasonal patterns in sediment deposition on channel beds and floodplains and monthly water quality monitoring. This was combined with continuous flow discharge measurements to examine suspended sediment and particulate phosphorus dynamics and reduction along reaches. Remediated streams with floodplains on both sides of the channel reduced particulate phosphorus concentrations and loads (−54 µg L−1, −0.21 kg ha−1 yr−1) along reaches, whereas increases occurred along streams with one-sided floodplains (27 µg L−1, 0.09 kg ha−1 yr−1) and control streams (46.6 µg L−1). Sediment deposition in remediated streams was five times higher on channel beds than on floodplains and there was no evident lateral distribution of sediments from channel to floodplains. There was no effect from sediment deposition on particulate phosphorus reduction, suggesting that bank stabilization was the key determinant for phosphorus mitigation in remediated streams, which can be realized with two-sided but not one-sided floodplains. Further, the overall narrow floodplain widths likely restricted reach-scale sediment deposition and its impact on P reductions. To fully understand remediated streams' potential for reductions in both nitrogen and different phosphorus species and to avoid pollution swapping effects, there is a need to further investigate how floodplain design can be optimized to achieve a holistic solution towards improved stream water quality.

Erosion, soil loss and consequent nutrient fluxes impair water quality and can degrade arable soils. Erosion rates in Sweden are generally low but episodic losses of suspended solids (SS) can affect water quality. Identifying critical source areas (CSAs) and “hot moments” is essential to reduce erosive losses from arable land. Here we use a distributed, dynamic high-resolution erosion model that simulates the sum of all transported material, i.e., erosion within the soil profile, on the soil surface and transport through drainage systems. We simulate monthly SS transport in six small agricultural catchments with varying soil texture over 8 years. Kling-Gupta Efficiency (KGE) was used as model performance statistics, and calibration (KGE = 0.45–0.78) and validation (KGE = 0.64–0.83) showed acceptable model performance for all catchments, with mean annual SS losses between 2.1 and 31.5 t km-2yr-1. Equifinality could be minimised by using more precise initial parameter values. We suggest that the model can be applied to comparable unmonitored catchments to identify erosion-sensitive periods and CSAs.

Eutrophication is an important threat to aquatic ecosystems world-wide, and reliable identification of areas vulnerable to phosphorus (P) losses from diffuse sources is essential for high efficiency of mitigation measures. In this three-step study we investigated (i) relationships between the agronomic (Olsen-P and P-AL) and environmental soil P tests (P-CaCl2) with molecular techniques (31P NMR and XANES) followed by (ii) rainfall simulation experiment on topsoil lysimeters and (iii) comparison to long-term field measurements of water quality. Soil samples were collected from seven sites indicated to be vulnerable to nutrient losses due to underlying geology. High P release correlated to standard agronomic P tests (Olsen P, r = 0.67; and P-AL, r = 0.74) and low P sorption capacity (r = − 0.5). High content of iron-bound P compounds indicated more labile P and higher release of dissolved P (r = 0.67). The leaching experiment showed that three out of four soils with high initial soil P status had both higher P leaching concentrations before fertilization (0.83–7.7 mg P l−1) compared to soil with low initial soil P status (0.007–0.23 mg P l−1), and higher increase in P concentrations after fertilization. Higher soil P sorption capacity reduced P leaching losses. Finally, long-term monitoring data show no significant trends in P losses in a field with low initial P content and moderate P fertilization rates whereas high and over time increasing P losses were recorded in a field with high initial soil P content and repetitively high P fertilization rates.