Lanthanum Modified Bentonite (LMB; Phoslock®) is used to mitigate eutrophication by binding phosphate released from sediments. This study investigated the fate of lanthanum (La) from LMB in water, sediment, macrophytes, and chironomid larvae in Lake Rauwbraken (The Netherlands). Before the LMB application, water column filterable La (FLa) was 0.02 µg L-1, total La (TLa) was 0.22 µg L-1. In sediment the total La ranged 0.03-1.86 g m-2. The day after the application the maximum FLa concentration in the water column was 44 µg L-1, TLa was 528 µg L-1, exceeding the Dutch Maximum Permissible Concentrations (MPC) of 10.1 µg L-1 by three to fourfold. TLa declined below the MPC after 15 days, FLa after 75 days. After ten years, FLa was 0.4 µg L-1 and TLa was 0.7 µg L-1. Over the post-application years, FLa and TLa showed statistically significant downward trends. While the LMB settled homogeneously on sediment, after 3 years it redistributed to 0.2-5.4 g La m-2 within shallow zones, and 30.7 g m-2 to 40.0 g La m-2 in deeper zones. In the upper 20 cm of sediment, La concentrations were 7-6702 mg kg -1 dry weight (DW) compared to 0.5-7.0 mg kg-1 before application. Pre-application anaerobic sediment release of FLa was 0.006 mg m-2 day-1. Three months after the application it was 1.02 mg m-2 day-1. Three years later it was 0.063 mg m-2 day-1. Before application La in plants was 0.8-5.1 mg La kg-1 DW, post-application values were up to 2925 mg La kg-1 DW. In chironomid larvae, La increased from 1.7 µg g-1 DW before application to 1421 µg g-1 DW after one month, 3 years later it was 277 µg g-1 DW. Filtration experiments indicate FLa is not truly dissolved free La3+ cations.

The combination of a low-dose coagulant (polyaluminium chloride—‘Floc’) and a ballast able to bind phosphate (lanthanum modified bentonite, LMB—‘Sink/Lock’) have been used successfully to manage cyanobacterial blooms and eutrophication. In a recent ‘Floc and Lock’ intervention in Lake de Kuil (the Netherlands), cyanobacterial chlorophyll-a was reduced by 90% but, surprisingly, after one week elevated cyanobacterial concentrations were observed again that faded away during following weeks. Hence, to better understand why and how to avoid an increase in cyanobacterial concentration, experiments with collected cyanobacteria from Lakes De Kuil and Rauwbraken were performed. We showed that the Planktothrix rubescens from Lake de Kuil could initially be precipitated using a coagulant and ballast but, after one day, most of the filaments resurfaced again, even using a higher ballast dose. By contrast, the P. rubescens from Lake Rauwbraken remained precipitated after the Floc and Sink/Lock treatment. We highlight the need to test selected measures for each lake as the same technique with similar species (P. rubescens) yielded different results. Moreover, we show that damaging the cells first with hydrogen peroxide before adding the coagulant and ballast (a ‘Kill, Floc and Lock/Sink’ approach) could be promising to keep P. rubescens precipitated.