Once you have completed this form Send by post to: Environmental Protection Authority, Private Bag 63002, Wellington 6140 OR email to: submissions@epa.govt.nz

SUBMISSION FORM

www.epa.govt.nz

Once your submission has been received the submission becomes a public document and may be made publicly available to anyone who requests it. You may request that your contact details be kept confidential, but your name, organisation and your submission itself will become a public document.

Submission on application number:

APP201051

Name of submitter or contact for joint submission:

Rodolphe QUEROU

Organisation name (if on behalf of an organisation):

Dow Chemical

Postal address: 371 Rue Ludwig Van Beethoven, F-06560 Valbonne, FRANCE

Telephone number: +33 4 93 95 54 03

Email: rquerou@dow.com

I wish to keep my contact details confidential

The EPA will deal with any personal information you supply in your submission in accordance with the Privacy Act

1993. We will use your contact details for the purposes of processing the application that it relates to (or in

exceptional situations for other reasons permitted under the Privacy Act 1993). Where your submission is made

publicly available, your contact details will be removed only if you have indicated this as your preference in the tick

box above. We may also use your contact details for the purpose of requesting your participation in customer

surveys.

The EPA is likely to post your submission on its website at www.epa.govt.nz. We also may make your submission

available in response to a request under the Official Information Act 1982.

2

Submission Form

September 2012 EPA0190

I support the application

I oppose the application

I neither support or oppose the application

The reasons for making my submission are1:

Dow Microbial Control is a subsidiary of The Dow Chemical Company and is the supplier of the active substance 4,5-Dichloro-2-octyl-3(2H)-isothiazolone (DCOIT) for use in Antifouling Paints. As per question 12.1 in the Application for reassessment of antifouling paints (APP201051), Dow doesn't agree with the recommendation made for AFPs containing the active substance DCOIT and would like to provide additional evidence to support its view:

1. To provide additional information in order to refine the environmental risk assessemtn of DCOIT, see details in appendix 1a and 1b.

2. To provide additional information on the benefits of use of DCOIT as an antifouling agent in AFPs for commercial boats, see details in Appendix 2

I wish to be heard in support of my submission (this means that you can speak at the hearing)

I do not wish to be heard in support of my submission (this means that you cannot speak at the hearing)

I wish for the EPA to make the following decision:

To retain AFPs containing DCOIT as an active substance with additional controls R1-R2-R3-R4 and R5.

1 Further information can be appended to your submission, if you are sending this submission electronically and attaching a file we accept the

following formats – Microsoft Word, Text, PDF, ZIP, JPEG and JPG. The file must be not more than 8Mb.

Appendix 1b

Additional information supporting the use of the SSD

Approach for the determination of the PNEC of DCOIT (and answer to New Zealand EPA questions)

1. Additional information on the mode of action of DCOIT as an active

substance in antifouling products 4,5-dichloro-2-octyl-2H-isothiazol-3-one (DCOIT) is a broad spectrum antimicrobial biocidal active substance from the well isothiazolinone family. It exhibits rapid inhibition of growth at very low levels and cidal effects at higher levels or for longer contact periods. DCOIT may function as a broad spectrum antifouling agent for preventing the growth and settlement of soft fouling (ex., bacteria, fungi, algae) and hard fouling (ex., barnacles) organisms on submerged surfaces. DCOIT reacts with the proteins of organisms that come in contact with the coating surface. This results in interruption of the metabolic processes that utilize these proteins. Fouling organisms initiate specific physiological activities involved in attaching to solid surfaces that are disrupted by DCOIT. As a result the organisms do not successfully colonize the treated surfaces and biofouling is minimized. Structure and physico-chemical properties of the molecule DCOIT is a covalent organic molecule that does not dissociate into ionic species. Therefore, the measurement of dissociation constant is not applicable to this active substance. This observation was confirmed by studying the effect of pH on the water solubility of DCOIT.1

The water solubility at 20°C was determined to be 4.26 and 3.47 mg/L at pH 5 and 7 respectively, which demonstrates no significant effect of pH on solubility, an observation consistent with a covalent organic molecule with no ionizable moieties.

At pH 9, DCOIT rapidly hydrolyzes due to the presence of the nucleophile OH-, which will result in a nucleophilic attack of the liable N-S bond and subsequent cleavage of the isothiazoline ring. DCOIT has hydrophobic properties, the water solubility is low, and the molecule is adsorptive, this is confirmed by the measure of the adsorption coefficient Kocsoil = 5662-7865; Kocsediment = 15441 (see EU assessment report).

1 EU Assessment report for DCOIT in Product Type 8. February 2011.

Mechanism of action ReferenceThe information below is based on the following reference (attached as appendix):

:

• Terry Williams (2007) “The mechanism of Action of Isothiazolone Biocides.”, published in PowerPlant Chemistry 2007, 9(1).

At the surface of the antifouling coatings, because of its hydrophobic properties, DCOIT will rapidly associates and binds with cells of aquatic organisms, within a few minutes. It is then postulated that at environmentally relevant concentrations, the mechanism of transport of isothiazolone through membranes is an active specific process. It occurs within minutes, as the inhibition of the central (intracellular) metabolic pathway (respiration and energy production) is observed to start within minutes. DCOIT utilizes a two-step mechanism of action, involving rapid growth inhibition leading to a loss of viability. Growth inhibition is the result of rapid disruption of the central metabolic pathways of the cell by inhibition of dehydrogenase enzymes. Key physiological activities that are rapidly inhibited in microbial cells are growth (reproduction) and respiration (oxygen consumption and carbon dioxide production). These processes are critical in bacteria, algae, fungi, and invertebrates, which explains why DCOIT is such a broad spectrum biocide. Inhibition of cellular activity is rapid (within minutes), whereas, cell death (cidal activity) is observed after several hours contact. Thus, the time course for efficacy is minutes to hours. Cell death results from the progressive loss of protein thiols in the cell from one of multiple pathways. As cell metabolism is disrupted, free radicals are known to be produced within cells and this is a likely contributor to the cidal mechanism. Overall, the higher the concentration of biocide, the shorter the contact time required for more complete kill. The interaction, of DCOIT with proteins and in particular thiol proteins is associated with an opening of the isothiazolone ring, which means that efficacy (target oraganisms) or toxicity (non target organisms) are directly related to a biodegradation of the molecule. The basis for the relationship of efficacy and time is stoichiometric, with the number of molecules per cell affecting both efficacy and time course. Instances of low biocide concentration and higher cell densities, effectively reducing the number of biocide molecules per cell, would therefore reduce efficacy [and toxicity] and lengthen the time course of action. This unique mechanism results in the broad spectrum of activity of DCOIT biocide, low use levels for microbial control, and difficulty in attaining resistance.

2. The EU CAR document did not include any SSD refinement, could you justify why the Rapporteur Member State didn’t accept these additional studies and instead used the lowest NOEC?

The draft EU Competent Authority Report which has been made publicly available on the website of the EU Commission by the Rapporteur Member State, Norway, is a draft, dated December 2010. Dow Microbial Control has commented on this draft and between other comments, provided the SSD refinement to Norway in June 2011, which explains why it does not appear in the draft CAR. However, during 2011 and 2012, the methodology to perform the risk assessment for environment has been modified in order to harmonize the evaluation of DCOIT with the evaluation of other antifouling active substance. The results of these refinements were that DCOIT passed all the environmental risk assessment scenarios for commercial harbor. Because DCOIT is supported only for commercial boats, passing the Marina scenario for pleasure crafts was not necessary and a detailed marina assessment was not performed. As a consequence, refining the PNEC was not needed anymore. Considering this, and the fact that the SSD refinement was provided relatively late in the EU evaluation process, Norway indicated to Dow Microbial Control that they were not favorable to re-open the discussion on the PNEC with other member states to include proposed SSD refinement. Indeed this would have delayed the evaluation and was not formally needed because all the commercial harbor scenario were passed with a comfortable margin of safety. Dow Microbial Control agreed with Norway and this is the reason why the SSD approach was not further evaluated by Norway which does not mean it was rejected. It is also very important to understand that the EU CAR available on the EU website was an early draft from December 2010. Based on the extensive discussion which took place in 2011 and 2012, Norway is preparing a revised version of the CAR that will be used to support proposed inclusion of DCOIT in Annex 1 of the BPD. Based on the information available to us today, in the final CAR, DCOIT shall pass the assessment for commercial harbor. This information can be verified with the Norwegian competent authorities (Biocider@klif.no).

3. Some studies used in the SSD are non-GLP studies and sometimes with

no analytical measurements (like the study on sea urchin). We are not convinced about the use of a geometric mean of the nominal and 0.1x the nominal concentrations rather than measured data. With DCOITs rapid degradation, this could result in an overestimate of the exposure concentration. We are unsure why the endpoints in the algal studies are selected for less than the duration of the study, i.e. if a 96 h study was done, why work with the 24 NOErC? At least both 24 and 96 h values should be presented. It’s not entirely clear from the paper, but we think that 5 of the 7 algal studies in the table were accepted by the Rapporteur and 2 others were found in the literature. The issue of duration of the endpoint being shorter than duration of the study affects both of the additional literature studies and 3 of the other studies.

Use of non GLP-studies

The SSD approach is based on 16 chronic endpoints for 16 separate species. Amongst these, only 5 are non GLP tests, extracted from literature publications (Arrhenius 2006; Bellas 2006; Braithwaite 2005). A quality assessment was conducted on these studies indicating that statistical methods are reported, increasing concentrations were used, and a dose-response curve was observed. It was therefore considered that these studies were reliable although non GLP.

Use of endpoints based on shorter duration than test duration and based on nominal concentrations (algae and aquatic plant studies)

Considering the rapid biodegradation of DCOIT, most of the GLP tests provided in the DCOIT dossier for EU evaluation have been done in flow-through conditions, and therefore the NOEC from these studies can be used without further refinements. This is true for most of the tests in fish and in invertebrates. However, in algae it is not possible to perform tests in flow-through conditions, and therefore static tests have been conducted and showed a rapid decrease of DCOIT concentration in culture medium (in general the concentration at the NOEC level was below the Limit of Detection after 24-h to 48-h). This decrease in concentration was associated sometimes with an increase of the NOEC over time. This is also true with the aquatic plant Lemna gibba. The NOEC in function of time for algae and aquatic plant studies is shown in the Table 1 below.

Table 1

: NOEC values over time for several algal species used in the SSD document (all in µg/L).

Species 24-h 48-h 72-h 96-h 120-h 7-days Navicula pelliculosa 0.34 0.77 1.4 2.2 NA NA

Skeletonema costatum2 0.48 NA NA 1.44 NA NA

Skeletonema costatum3 2 3.9 3.9 3.9 NA NA

Selenastrum capricornutum

31 31 31 7.8 NA NA

Anabaena flos-aquae 12 12 6.4 2.8 NA NA

Lemna gibba NA NA 4.54 NA NA 11.8

2 2002 study 3 2007 repeated study

These results indicates that in several instances, in particular in the most sensitive species (Navicula pelliculosa, Skeletonema costatum and Lemna gibba), the decrease in DCOIT concentration over time (below the LOD) was associated with an increase in the NOErC based on nominal concentration. This explains why for the most sensitive species, the NOEC reported for these studies were for exposure duration shorter than the duration of the experiment. Considering this, and considering that Navicula pelliculosa was the most sensitive species, 2 options were available to determine the NOEC for these studies, either to use a geometric mean approach, or to use shorter term endpoints based on nominal concentration. The geometric mean approach is based on the calculation of a geometric mean concentration between the nominal concentration and the concentration at the end of the experiment (in algae relevant time is in general considered to be 72-h). However, the geometric mean approach has some cons-argument. indeed, when the active ingredient is not detectable at the end of the experiment, it is necessary to calculate a geometric mean based on the Limit Of Detection which is itself depending on the analytical method and can therefore be relatively independent from the toxicity. Additionally, it was considered that because the mode of action of DCOIT is very rapid (within minutes), it was more relevant to use short term toxicity endpoints based on nominal concentration. Dow Microbial Control developed specific arguments to support this view which was raised to Norway and to other EU competent authorities. This argument was accepted and the EU authorities agreed to consider the 24-h NOErC based on nominal concentration as the relevant endpoint. Extracts of the argument are reproduced below.

Argument to use nominal short-term (24-h or 48-h) values for NOErC in algae

Dow Microbial Control advocates the use of the time zero initial measured DCOIT concentrations to derive the toxicity endpoints. The rationale is based on the specific mode of action of DCOIT in bacteria, fungi and algal cells and its rapid time course. Dow Microbial Control suggests that the use of the calculated geometric mean for the derivation of the toxicity endpoints and subsequently the PNEC for the aquatic risk assessment is highly conservative and does not reflect the unique mode of action of DCOIT. A recent Navicula pelliculosa study (Sinderman et al. 2007a) was conducted according to the current OECD Guidelines, specifically OECD 201, adopted 23 March, 2006 and is considered to be entirely valid.

The definitive study dose regimen was defined following the conduct of a range finder study. The method limit of quantification [LOQ] was defined as 0.80 µg a.i. L-1 calculated as the product of the lowest calibration standard [20.0 µg a.i. L-1] and the dilution factor of the matrix blank samples [0.040]. DCOIT is a very effective biocide due to its rapid mode of action and its favorable environmental fate characteristics. DCOIT utilizes a two-step mechanism involving rapid growth inhibition leading to a loss of viability. Growth inhibition is the result of rapid disruption of the central metabolic pathways of the cell by inhibition of dehydrogenase enzymes. Key physiological activities that are rapidly inhibited in microbial cells are growth (reproduction) and respiration (oxygen consumption and carbon dioxide production). These processes are critical in bacteria, algae, fungi, and invertebrates, which explains why DCOIT is such a broad spectrum biocide. DCOIT rapidly associates with microbial [and similarly algal] cells. Inhibition of cellular activity is rapid (within minutes), whereas, cell death (cidal activity) is observed after several hours contact. Thus, the time course for efficacy is minutes to hours. Cell death results from the progressive loss of protein thiols in the cell from one of multiple pathways. As cell metabolism is disrupted, free radicals are known to be produced within cells and this is a likely contributor to the cidal mechanism. Overall, the higher the concentration of biocide, the shorter the contact time required for more complete kill. The basis for the relationship of efficacy and time is stoichiometric, with the number of molecules per cell affecting both efficacy and time course. Instances of low biocide concentration and higher cell densities, effectively reducing the number of biocide molecules per cell, would therefore reduce efficacy [and toxicity] and lengthen the time course of action. This unique mechanism results in the broad spectrum of activity of DCOIT biocide, low use levels for microbial control, and difficulty in attaining resistance. It is noteworthy that the relationship between DCOIT concentration and cell density is pertinent with regard to in vitro culture conditions. Under environmentally realistic circumstances the environmental fate characteristics of DCOIT and the interactions with natural benthic and lentic assemblages tend to mitigate potential toxicity. It is likely that the reduction in DCOIT concentrations in algal cell cultures is the result of several concurrent processes, namely adsorption and uptake of DCOIT into cells as well as some degree of biodegradation. It is likely that the dissipation of the test material from the cell cultures due to adsorption to organic carbon is a relatively small fraction of the total available. However, as noted previously, the mode of action of the biocide is via cellular uptake and subsequent interference with the dehydrogenase enzymes of the energy metabolism pathways of the exposed cells. According to the mechanism of action of DCOIT, uptake through the cell wall and membrane of the algae occurs rapidly, within hours and facilitates the activity of the biocide. The cell wall of the diatom Navicula is comparatively unrestrictive with regard to movement of molecules compared to the complex cell wall structure of bacteria and blue green algae. Therefore, uptake is relatively unencumbered and rapid. Based on the ratio of the initial inoculum versus DCOIT concentration, uptake

can be significant with regard to total mass. Subsequent to uptake and enzymatic inhibition, the isothiazolone ring is cleaved rendering the molecule inactive and undetectable with methods developed for parent. This means that the inhibitory effect on algae will also result in a degradation of the molecule by the algae. With regard to biodegradation (by bacteria), Dow Microbial Control contend that the extent of bacterial biodegradation in the algal cultures is likely to be relatively insignificant, this conclusion is based on several reasons: 1. while the algal cultures are not sterile, they are however relatively free of bacteria; contamination with significant bacteria would foul cultures and 2. DCOIT is bacteriostatic at concentrations at which the algal studies were conducted [1.4 – 9.6 µg L-1]. Thus, the Dow Microbial Control contends that bacterial biodegradation is likely to be a relatively minor dissipation route for DCOIT in algal cultures. Dow Microbial Control has previously provided data indicating that DCOIT is stable over time with regard to abiotic degradation. Therefore, the dissipation of DCOIT from the test media is linked to the presence of the algal inoculum and the degradation of DCOIT specifically by the algae. The mechanism of degradation is closely linked to the mode of action of DCOIT. Based on the proceeding supporting information Dow Microbial Control suggests that that the toxicity endpoints should be based on the time zero initial measured toxicant concentrations.

Applicability of the argument described above for literature data

Three chronic NOEC have been extracted from the literature reference Bellas (2006), on the mollusc Mytilus edulis, the echinoderm Paracentrotus lividus and the ascidean Ciona intestinalis. Dow Microbial Control believes that the argument develop above advocating for the use of short term NOEC based on nominal concentration rather than the geomean approach, are also valid for these 3 studies. The test with Mytilus edulis was performed on very early stage of development, namely from fertilized eggs until the veliger larvae stage, so on organisms with very simple cellular structure. Similarly in Paracentrotus lividus and in Ciona intestinalis, tests were conducted on very early stage of development from fertilized eggs until the initial embryo stages. Under these conditions, it can be expected that the mechanism of action of DCOIT previously described for algae and microbial cells (binding to cell followed by cellular uptake and inhibition of specific intracellular enzymes) would be very similar in such early stages of development with simplified cellular structure. It is also expected that these toxicity tests on very sensitive stages of development provide a conservative estimation of the chronic NOEC for these species. We can expect a very rapid uptake of DCOIT by embryo cells, and subsequent to uptake an enzymatic inhibition, during which the isothiazolone ring is cleaved rendering the molecule inactive. This means that the inhibitory effect observed in embryos will also result in a degradation of the molecule by the aquatic organisms. With regard to biodegradation by microorganisms, Dow Microbial Control believes that like in algae, the extent of bacterial biodegradation in the embryo cell cultures is likely to be relatively insignificant, especially considering that the author performed the test using filtered sea water (0,22 µm), which certainly limited the presence of microorganism and therefore the biodegradation of DCOIT prior to binding to the cells. This means that from our point of view if any decrease in concentration of DCOIT had been observed in these tests, this would have been associated mostly with the interaction of DCOIT with proteins in non target embryo cells (associated with the inhibitory effect) rather than from a biodegradation by microorganisms outside the embryo cells. It has been shown that abiotic degradation at environmental relevant pH is not a major degradation pathway of DCOIT. Therefore, in filtered seawater, the degradation of DCOIT is expected to be low. This has already been observed in the past when using artificial sea water. In such conditions, considering nominal values should be acceptable, the use of a geometric mean of the nominal and 0.1x the nominal concentrations is even over-conservative, considering the mechanisms described above. The test on the macroalgae Fucus serratus Linnaeus (Braithwaite 2005) was also performed on very early stages of development (germination of zygotes) over 24, 48 and 72-h. Therefore the same reasoning as described above based on mechanism

of action also applies. The NOEC in this study increased over time in particular between 48-h and 72-h, confirming that the 24-h or 48-h results aree more relevant than the 72-h results. In the last study from the literature (Arrhenius, 2006), on Scenedesmus vacuolatus, the test was performed over 24-h and DCOIT concentration was reported. From all these observations, Dow Microbial Controls concludes that for the literature studies used in the SSD, the use of short-term NOEC based on nominal concentration provide a sufficiently conservative estimate of the chronic NOECs in the tested species.

4. Comparison of marine and freshwater sensitivity

The mode of action of DCOIT has been described in details in the section 1 above.

To summarize, the mode of action includes the following steps:

- Cell binding - Transport across membranes - Inhibition of intracellular proteins including dehydrogenases and thiol-proteins

Cell binding

It has been shown that the

is a passive physical process resulting from hydrophobic interactions, which will depend only from the concentration of DCOIT into water, which is itself independent from salinity and pH because DCOIT is a non ionizable organic covalent molecule.

transport

The inhibition of intracellular protein is associated with a growth inhibition resulting from rapid disruption of the central metabolic pathways of the cell by inhibition of dehydrogenase enzymes. Key physiological activities that are rapidly inhibited in microbial cells are growth (reproduction) and respiration (oxygen consumption and carbon dioxide production). These processes are not specific to fresh or marine water organisms.

of DCOIT across membranes is an active process energy-dependant. It is therefore also expected to be independent from the conditions of the media, but only related to the presence of receptors on the membrane. Considering the broad spectrum of action of DCOIT on fouling species, the presence of such receptors is expected to be common in aquatic organisms from different taxonomic groups.

These observations are consistent with the results obtained with DCOIT toxicity testing in aquatic organisms, which don’t indicate any specific difference in fresh and marine water species.

More generally speaking, the ECETOC Technical Report on Risk Assessments in Marine Environmnets4

indicates that there is no evidence to suggest increased sensitivity in marine species when compared to equivalent phyla from the freshwater environment. This conclusion is supported and endorsed within the text of the EU Technical Guidance for New and Existing Substances (TGD).

5. Comparison of Mesocosm NOEC with HC5 data

Dow Microbial Control agrees that the mesocosm NOEC from Larson et al (2003) has been established based on photosynthetic activity (as a surrogate for cells activity).

In another study on periphyton (Arrhenius, 2006), the effect of DCOIT was observed on photosynthetic activity and reproduction activity.

The objective of the comparison of DCOIT NOEC from toxicity testing (SSD analysis) and with available mesocosm studies was to show consistency between the results and an indication that the SSD-derived HC5 is protective enough.

Conclusion

Based on the additional information provided in this document, Dow Microbial Control believes that the SSD approach developed in Appendix 1a is a valid approach and provides a realistic worst case estimation of DCOIT PNEC (0,11 µg/L) that can be used in environmental risk assessment.

4 Risk Assessment in Marine Environments. European Centre for Ecotoxicology and Toxicology of Chemicals. Technical Report No.: 82. ISSN -0773-8072-82. Brussels, December 2001.