21bb3034 Bijlage 1 behorende bij 21bb437 - Phoslock toxicity evaluation

https://doi.org/10.1016/j.watres.2016.11.045 Waters-Hart, K.E., 2020. Assessing aquatic macrophyte recovery responses following lanthanum-modified bentonite applications. University of Stirling. Watson-leung, T., 2009. Phoslock Toxicity Testing with Three Sediment Dwelling Organisms ( Hyalella azteca , Hexagenia spp . and Chironomus dilutus ) and Two Water Column Dwelling Organisms ( Rainbow Trout and Daphnia magna ). Etobicoke, Ontario. WHO, 2003. Guidelines for safe recreational water environments - Coastal and Fresh waters new Guidelines for Safe Recreational Water. Wood, S.A., Heath, M.W., Holland, P.T., Munday, R., McGregor, G.B., Ryan, K.G., 2010. Identification of a 30

benthic microcystin-producing filamentous cyanobacterium (Oscillatoriales) associated with a dog poisoning in New Zealand. Toxicon 55, 897–903. https://doi.org/10.1016/j.toxicon.2009.12.019 Yamada-Ferraz, T.M., Sueitt, A.P.E., Oliveira, A.F., Botta, C.M.R., Fadini, P.S., Nascimento, M.R.L., Faria, B.M., Mozeto, A.A., 2015. Assessment of Phoslock® application in a tropical eutrophic reservoir: An integrated evaluation from laboratory to field experiments. Environ. Technol. Innov. 4, 194–205. https://doi.org/10.1016/J.ETI.2015.07.002 Yasseri, S., 2008. Report on the Application of Bentophos on the Silbersee. Ottersberg, Germany. Yasseri, S., Epe, T.S., 2016. Analysis of the La:P ratio in lake sediments – Vertical and spatial distribution assessed by a multiple-core survey. Water Res. 97, 96–100. https://doi.org/10.1016/j.watres.2015.07.037

31

Appendix Box 1A – Exclusion of studies using soluble lanthanum salts As pointed out in section 2.1.1 the use of different lanthanum compounds in ecotoxicological assays has produced a high variability in the test results, showing higher toxic effects for the soluble lanthanum salts compared to the insoluble lanthanum forms and to Phoslock® (Herrmann et al., 2016 and references there in). Furthermore, even for studies conducted with the soluble salts and using the same test species, substantial differences for the dose descriptor values were reported between the studies (Spears et al., 2013a). The discrepancy of the dose descriptor values for the soluble lanthanum compounds reported in the literature is mainly due to the implementation of different study designs and the difference in the chemical composition of the aqueous medium used in the assays (Copetti et al., 2016). Lanthanum has a strong preference for oxygen-containing anions like carbonate and phosphate with which it can form stabile compounds. Therefore, the different ionic strength and ionic composition of the test medium used in the various ecotoxicological assays highly influenced the test results, even among studies conducted with the soluble salts.

This inconsistency in tests results with the soluble lanthanum salts does not allow a consolidated comparison with tests conducted with Phoslock® itself and was one reason why these studies were excluded in the present toxicity assessment. In addition, with regard to the use of Phoslock® for eutrophication management, it is important to consider the speciation of any Phoslock® derived lanthanum in the natural environment to which it is introduced. Phoslock® is specifically designed to overcome any ecotoxicological effects associated with lanthanum (Robb et al., 2003). The lanthanum is strongly adsorbed within the interlayer space of the clay matrix and is not released in its ionic form in moderate to high alkalinity water. Any filterable lanthanum species in natural water is found to be in a colloidal form, either as lanthanum phosphate or associated with humic acids (Reitzel et al., 2017). Thus, due to the different chemical properties and toxicological features of Phoslock® in the aqueous environment the direct comparison with tests conducted with the soluble lanthanum salts is inappropriate and the second reason why these studies are excluded for the toxicological assessment in this report.

At least, as the early experimental formulation of Phoslock® was a liquid form which differs considerably from the actual dry granular formulation produced since the end of 2006, studies undertaken during the early development phase of Phoslock® were not considered for the present document. The sole exception to this is the study described by Stauber & Binet (2000). These authors demonstrated that after thorough testing by CSIRO, modifications to the formulation avoid excessive lanthanum leaching from the liquid form. Therefore the results from this study allows a somewhat comparability with the studies undertaken with the actual dry granular form and thus were included in the toxicity evaluation presented here.

32

Figure 1A. free La ions concentration estimated as a function of La concentration added using CHEAQS (Verweij, 2017). The input was based on data provided by HHSK of CO3 (170 mg/L), Ca (80 mg/L) - as a proximity of alkalinity- and PO4 (0.6 mg/L) concentration, pH value was fixed at 7. This estimation is a worst-case scenario because we considered that 1% of the La added would be leached from the clay. The values actually measured are below 0.11% (Lürling and Tolman, 2010; NICNAS, 2001; Mucci et al., in review.). The CHEQS may also overestimate possible free La3+ concentration as La ions might also bind dissolved organic carbon ameliorating any possible toxic effect (Li et al., 2020; Spears et al., 2013a).

Figure 2A. Fish stock in Lake de Kuil (The Netherlands) before and after a Phoslock application (Data modified from Waajen et al., 2016a).

33

100

PO4 and La species (%)

80

Free PO43-

H2(PO4)H3(PO4)(aq) La(PO4)(s)

60

H4(PO4)+

H(PO4)2La3+ La(OH)3(s)

40

La(OH)-4 20

0 1

2

3

4

5

6

7

pH

8

9

10

11

12

13

14

Figure 3A. La and PO4 species (in % of abundance) at different pH predicted from chemical equilibrium modelling (CHEAQS -Verweij, 2017)) using as input 4 µM of La and P. Note that La(PO4) is stable between pH 4 and 12. Subscript s in the legend means solids. Grey rectangle is the pH range with predominantly the insoluble La(PO4) (s). (graph from Mucci et al., in prep).

34

A

120

Survival (%)

100 80

60

60 40 40 20

-1

Survival Filterable La (mg L-1)

80

Lanthanum (mg L )

100

140

20

0

0 0

100

330

1000

3300

10000

-1

La-modified bentonite concentration (mg L )

B

Figure 4A. (A) Experiments exposing Chironomus sp. to different Phoslock® concentration (100 to 10000 mg/L) for seven days. The bars represent survival rate (%) and circles filterable lanthanum concentration (µg/L). (B) picture of the experiment (Lürling, unpublished data).

35

Table 1A. List of all the 29 peer-reviewed studies that evaluated whole field Phoslock® application and/or enclosures experiments together with the main ecological changes observed in the studied systems. Table continues on next pages.

Reference

Main ecological changes observed

Barçante et al., 2020

Phoslock® reduced phosphorus concentration, increased Secchi disk and noncyanobacteria biodiversity. Those changes were short due to the high large amount of nutrients that still enters in the reservoir by the tributaries. No fish kill or any adverse effect of Phoslock is mentioned.

van Oosterhout et al., 2020

It showed that after almost 10 years of Phoslock® application in lake Rauwbraken the La added was most presented in the sediment (91%), small percentage was presented in macrophytes. No fish kill or any adverse effect of Phoslock® is mentioned.

Waajen et al., 2019

Several restoration measures were implemented in Lake Kleine Melanen (Fish removal, dredging, removal trees and shrubs, construction of marsh area and sanding capping with Phoslock®). After the in-lake measures phosphorus and chlorophyll-a suffered significant reduction. Still the target of a clear water state was not completely met due the external P load caused by storm water runoff. No fish kill or any adverse effect of Phoslock® is mentioned

Bishop and Richardson, 2018

Phoslock® decreased water column total P levels and shifted mobile sediment P fractions to the residual fraction in two golf ponds. Total N/P ratios (by mass) significantly increased. Consequent changes in the algal assemblage included decreases in dominance and overall density of cyanobacteria as well as a shift away from scum-forming genera to planktonic forms. No fish kill or any adverse effect of Phoslock® is mentioned

Lürling et al., 2017

Enclosure experiment in Heesch pond showed that a combination of dredging and Phoslock® was the treatment most effected; still, Phoslock® was not able to reduce as expected the P concentration which could be connect to the presence of humic acids, however the experiment was running for 59 days. Later Dithmer et al., 2015 showed that humi acids might cause a kinetic interference so probably if the experiment were longer a further reduction in P would be expected. Also showed that the filterable La concentration was higher in the treatment with Phoslock® (However, this is not expected to be La3+ ). Zooplankton abundance increase in all the enclosure and no adverse effect of Phoslock® was observed.

Nürnberg, 2017

The study described the application and some results from 4 Phoslock® application in Canada and gives advice that for a successful aplicaiton of Phoslock the external P lead should be smaller than the internal loading. No fish kill or any adverse effect of Phoslock® is mentioned

Epe et al., 2017

Long term (9y) monitoring showing that Phoslock® treatment in Bärensee was able to reduce total P concentration and chlorophyll-a improving water quality and allowed the lake to be used continuously for swimming and recreation acitivity. Due to the constant input of nutrients from the recreational activities small maintance applications is needed. No fish kill or any adverse effect of Phoslock® is mentioned

Waajen et al., 2017b

Phoslock® application in Lake de Kuil increased water quality and the La concentration content in macrophytes, chironomids, fish liver bone, muscle and skin. No acute and chronic effects of La accumulation were observed and human health risks are considered negligible. No fish kill or any adverse effect of Phoslock® is mentioned in the paper. 36

Waajen et al., 2017a

The effect of Phsolock® combined with a low dose of FeCl3 on macroinvertebrates were assessed by a microcosm experiment; and by evaluating the macroinvertebrate community in lake de Kuil before and after the intervention. Results show a reduction in number and taxa after on month of the application, however, after one year the macroinvertebrate numbers and taxa exceeded the pre-application situation and Gastropoda and Oligochaeta prospered. The atuhors affirmed that the effects one month after the treatment are most likely due to the combination of physical impacts of the use of bentonite and chemical impacts of the use of FeCl3, while effects after one year are likely attributed to the shift in trophic state of the lake.

Waajen et al., 2016a

Phoslock together with fish removal reduced sediment P release and reduced chlorophylla concentration which established and kept a clear-water state in two urban ponds. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

Spears et al., 2016

The study evaluated 18 lakes after a Phoslock® application (some of them together with a flocculant some of them solo Phoslock). The meta-analysis showed that total phosphorus concentration decreased significant across the lakes, also significant reduction of phosphate. In addition to reduction of summer chlorophyll-a concentration and increase in Secchi depth, relatively weak but still significant. Aquatic macrophytes species number and maximum colonization depths increased followed by a Phoslock® treatment. In general, Phoslock® application improved water quality and the authors stress the need for comprehensive pre- and post-application assessment.

Waajen et al., 2016b

Phoslock® application, together with a low dose of FeCl3 (as a flocculant) reduced sediment P release and effectively precipitated the cyanobacterial bloom presented in Lake de Kuil. The treatment increased secchi depth, decrease chlorophyll-a and TP, increased the macrophyte coverage. Overall, the treatment improved water quality for six summer changing the Lake to a more mesotrophic state. However, due to the constant external P load the lake the sediment P-release has gradually increased in recent years. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

After a solo Phoslock® application in Swan lake, the TP concentration and phytoplankton biomass reduced in the year after. Secchi transparency increased and no cyanobacterial Nürnberg and Lazerte, bloom nor cyanotoxin was identified. Overall, the authors observed a lack of response in 2016 the first treatment year because Phoslock® was applied late in the season (end of April and March), when the P released from the sediment was already been taken up by phytoplankton. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

Yamada-Ferraz et al., 2015

In an enclosure experiment in a reservoir, Phoslock® increased turbidity but after 12h turbidity was lower than the control. A reduction in phosphate was also observed after a Phoslock® application. An increase in phytoplankton community was also observed in the enclosure. After 29 days of treatment a reduction in richness of the benthic community was observed, the authors stressed that a natural recovery process might occur in the reservoir after the restoration period, enabling the appearance of a great diversity of organisms linked to healthier aquatic conditions. The authors also evaluate through microcosm experiments the effect on one species of zooplankton and benthic organisms, showing an effect during the treatment on C. silvestrii and no mortality after 984 hours. 42% of the benthic organisms died in the control which is higher than the treatment (33%), because of the high mortality rate in the control the experiment cannot conclude anything towards Phoslock® toxicity.

37

Dithmer et al., 2016

The study evaluated sediments from 10 Europeans lakes treated with Phoslock®. They showed that Lanthanum phosphate mineral (rabdophane and monazite) were formed in all of the treated lakes. Also that the ratio between La and P were above 1 and that La was vertically mixed in the sediment in a depth up to 10 cm. Phosphorus release experiment done also indicating effective control of sediment P release (2 to 9 years after the treatment).

Gunn et al., 2014

Phoslock® improved water quality in Loch Flemington (described in other study) and this paper discussed how the aquatic macrophyte community in Loch Flemington responded to Phoslock®. The Aquatic plant colonization depth increased and plant coverage of the lake bed extended as a result of Phoslock® application. However, the predominant plant was Elodea canadensis and species richness remains similar to previously condition. The authors discussed that the P reduction is not sufficient by itself to rapidly restore aquatic macrophyte species diversity of a lake, so the P reduction will not solve a pre-existing problem with aquatic invasive species.

Moos et al., 2014

The paper evaluate the diatoms assemblage through Paleolimnology studies in 4 system in which Phoslock was applied. The conclusion is that Phoslock reduced TP concentration in all of them and in 2 Canadian system there were a short-lived changes in the diatom assemblage while in the two Australian sites the diatom assemblage were similar to preapplication which seems that to be driven by season rainfall or climate. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

Bishop et al., 2014

Phoslock® was applied in Laguna Niguel Lake in 2011, as a result phosphorus concentration decreased, Secchi depth improved and phytoplankton diversity increased. No fish kill or any adverse effect of Phoslock is mentioned in the paper.

Novak and Chambers, 2014

The paper shows the result of two Australian river before and after a Phoslock® application. It shows that phosphorus reduced over time as a result of the treatment in both rivers. In Lower vasse river despite a TP reduction from 400 to 150 ug/L still cyano blooms persisted. The authos stressed that Phoslock® application resulted in better water quality in both periods over the period was applied and also that can provide a better environment for the establishment of macrophytes, however for certains systems other intervention to assure macrophytes transplantation might be needed for instance water level manipulation, which is the case for Canning river. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

Lürling and van Oosterhout, 2013

Lake Rauwbraken was treated with Phoslock® and a low dose of PAC in 2008 the intervention caused a reduction in phosphorus and chlorophyll-a. Also, an increase in secchi depth and water transparency. Lake Rauwbraken since the intervention has not suffered cyanobacterial blooms. No fish kill or any adverse effect of Phoslock is mentioned in the paper.

Spears et al., 2013

The paper investigated the potential for negative ecological impacts from elevated La concentrations associated with the use of Phoslock® across 16 case study lakes. Surface and bottom water peak TLa concentrations during and in the month following the application of Phoslock ranged from 0.026 mg /L to 2.300 mg/ L and from 0.004 mg/ L to 0.892 mg/ L, respectively. The authors concluded besides other thoughts that the maximum reported estimates of Phoslock® in receiving waters did not exceed the EC50 values for Phoslock reported by others and based on laboratory trials. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

38

Meis et al., 2013

Phoslock® was applied to Loch Flemington, and sediment cores before and after the application were collected. The paper showed that Phoslock® increased the mass of P permanently bound to the sediment (apatite bound P). Phoslock® application also caused an increase in La concentration in the sediment. The authors also discussed if a multiple smaller dose compared to a single high dose is likely to increase cost-effectiveness. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

Lürling and van Oosterhout, 2012

Lake Het Groene Eiland (clastrum) was treated with Phoslock® as a result the mean phosphate and total P concentrations were only marginally lower in the claustrum compared with the reference site. The cyanobacteria chlorophyll-a concentrations were significantly lower in the claustrum than in the reference water. The clastrum, however also experience gradual increase in P concentrations. Already after 1 year of Phoslock® application, a reapplication was needed. The authors stressed that interference with oxyanions and humic substances could negative influenced Phoslock. But these interferences were later proved to be a kinetic and given a certain time Phoslock would bind the phosphate (Dithmer et al., 2015). The authors also mentioned that external P load from birds, or groundwater rich in P could have influenced the gradual increase in P concentration. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

Lürling and Faassen, 2012

Enclosure experiment in an urban pond showed that a combination of dredging and Phoslock® was the treatment most effected. The combined treatment reduced cyanobacterial biomass, microcystin and phosphate concentration. The authors also discussed that application of Phoslock® under high humic substances needs deeper research, which nowadays this interferences was study by Dithmer et al. (2015) and showed that the interference is kinetic and does not change Phoslock® binding sites. No fish kill or any adverse effect of Phoslock is mentioned in the paper.

Meis et al., 2012

This paper studied the short term alterations in sediment composition and sediment Pfractions based on sediment cores taken 2 days before and 28 days following a Phoslock® application in Clatto reservoir. After the intervention La concentration increased in the sediment. There was an increase in residual P fraction but the P mobile did not vary significantly over a period of 28 days, it is also arguable that 28 days is still too short to see a difference in the Pmobile as the P is not yet available to be bound and over time the change will occur as observed in other field applications. Experimental P adsorption studies showed that Phoslock® can increase the P-sorption capacity of the sediments.

van Oosterhout and Lurling, 2011

Phoslock® and a low dose of PAC was applied in Lake Rauwbraken. The intervention reduced P and cyanobacterial biomass, also improved water transparency. They showed that right after the application D. galeata disappears from the water column for at least 3 months. However, after 3 months daphnids recovered to similar abundance than before the treatment. The authors discussed that several factors may have caused the disappearance of D. galaeta following this treatment, for instance physical effects due to flocks (created by the addition of PAC), grazing inhibition by flocks and clay, very low food concentration and absence of predation refuge.

Robb et al., 2003

Phoslock® was applied in Vasse river and Canning River. In Vasse River, P and chlorophyll-a reduction were observed, however in Vasse River those changes were less clear. The authors stress that Phoslock® effectively intercepts the P released form the sediments and can strip phosphate from the water column, however, it will only work in a longer term in combination with external source control. No fish kill or any adverse effect of Phoslock® is mentioned in the paper.

39

Liu et al., 2009

Document not available

Kaczorkiewicz and Heese, 2008

Document not available

40

41