Arch. Environ. Contam. Toxicol. 43, 296–300 (2002) ARCHIVES OF DOI: 10.1007/s00244-002-1203-6 Environmental Contamination and Toxicology © 2002 Springer-Verlag New York Inc.

Control of Fouling , leucophaeata (Conrad), with Sodium Hypochlorite

S. Rajagopal,1 M. van der Gaag,1 G. van der Velde,1 H. A. Jenner2 1 Department of Ecology and Ecophysiology, Section Ecology, Faculty of Science, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands 2 KEMA Environmental Research and Sustainables, P.O. Box 9035, 6800 ET Arnhem, The Netherlands

Received: 18 October 2001/Accepted: 4 March 2002

Abstract. Though the Conrad’s false mussel, Mytilopsis leu- 1898), Germany (Nordostseekanaal in 1932), and more re- cophaeata, is an important fouling animal in industrial cooling cently in Britain (South Wales in 1998) (Nyst 1835; Vorstman water systems, there are no published reports on the tolerance 1935; Oliver et al. 1998). It is a highly euryhaline of this species to chlorination. A series of experiments was (Siddall 1980; Van der Velde et al. 1998) and occurs mostly in conducted to determine the effects of mussel size (2–20 mm brackish waters (Boettger 1932). This species is reported to shell length), season (breeding versus nonbreeding), nutritional have originated from the Atlantic coast of the United States and status (fed versus starved) and acclimation temperature (5– the (Marelli and Gray 1983). M. leucophaeata 30°C) on the mortality pattern of M. leucophaeata under con- is a biofouling and nuisance organism, causing problems in tinuous chlorination (0.25–5 mg/L). The effect of mussel size industrial cooling water systems (Jenner et al. 1998). Settle- on M. leucophaeata mortality in the presence of chlorine was ment densities as high as 6.5 million/m2 have been reported significant, with 10 mm size group showing greater near power station intakes in the Noordzeekanaal (The Neth- resistance. At 0.25 mg/L residual chlorine, 2 mm size group erlands), the canal which connects the harbors of Amsterdam mussels took 89 days to reach 100% mortality, whereas 10 mm with the (Rajagopal et al. 1995). A comparison of size group mussels took 109 days. M. leucophaeata collected values in the literature with those obtained for M. during nonbreeding season (December–April) was more toler- leucophaeata (37 kg/m2 dry weight) from the Noordzeekanaal ant to chlorine than those collected during breeding season shows that the biomass build-up there is probably one of the (June–October). Nutritional status of the mussel had no signif- highest reported from the temperate waters. icant influence on the chlorine tolerance of the mussel: fed and Chlorination of cooling water, using chlorine gas or hypo- starved mussels succumbed to chlorine at equal rates. The chlorite from electrolytic process, is commonly used for mussel effect of acclimation temperature on M. leucophaeata mortality control in power stations (Jenner et al. 1998; Bidwell et al. in the presence of chlorine was significant. At 0.5 mg/L resid- 1999). However, there are disadvantages in the use of chlorine ual chlorine, mussels acclimated at 5°C required 99 days to as an antifouling agent in once-through cooling systems. Chlo- reach 95% mortality, whereas mussels acclimated at 30°C rine by-products are potential pollutants of receiving waters required 47 days. A comparison of present data with previous that can impact nontarget organisms (Jenner et al. 1998; Allo- reports suggests that resistance of M. leucophaeata to chlori- nier et al. 1999). A total residual chlorine level of 1 mg/L is nation is higher than other mussel species causing fouling normally used for mussel control in Europe during breeding problems in The Netherlands (Mytilus edulis and periods; during nonbreeding periods considerably lower chlo- polymorpha). rine levels (0.2–0.5 mg/L) are used (Jenner et al. 1998). There- fore, a study on the response of M. leucophaeata to chlorine has great importance in terms of the optimization of chlorine treatments in the cooling systems. Though there is a large and The brackish water mussel, (Conrad scattered literature on the chlorine response of bivalves in 1831) (syn. Congeria cochleata Kickx in Nyst 1835) was first general, and for Dreissena polymorpha (Pallas) and Mytilus recorded in The Netherlands by Maitland in 1895 in the Amstel edulis L. in particular (see Jenner et al. 1998 for a review), (Vorstman 1935). In Europe, the occurrence of M. leucophae- there are no published studies on the response of M. leucophae- ata has been also reported from Belgium (River Schelde near ata to chlorination. In the present study, identical experimental Antwerp in 1835; first record in Europe), France (near Cae¨n in setups were used for conducting chlorine toxicity studies with M. leucophaeata, M. edulis and D. polymorpha to make results comparable and to assist utilities in planning chlorine regimes for controlling these mussels. The objectives of this study were Correspondence to: S. Rajagopal; email: [email protected] (1) to find out responses of M. leucophaeata at different chlo- Response of M. leucophaeta to Chlorination 297 rine concentrations along with other common fouling mussel of exposed mantle tissues to prodding. The number of dead in species M. edulis and D. polymorpha and (2) to examine each experiment was recorded along with their shell lengths for each potential sources of variation in the responses of M. leucopha- observation time. eata in chlorine bioassays with respect to (a) animal size (shell length), (b) season (breeding and nonbreeding), (c) mainte- nance regime (fed versus starved), and (d) acclimation temper- Effects of Mussel Size atures. To determine if size had any significant effect on the survival of M. leucophaeata in bioassays, three size groups (2 mm, 10 mm, and 20 Materials and Methods mm shell length) were tested at six different chlorine concentrations (0.25, 0.5, 1, 2, 3, and 5 mg/L). Four replicates of each chlorine treatment concentration and of controls were used for all toxicity Experimental Animals studies. Altogether, 1,680 mussels were used for the mussel size mortality experiments (20 mussels in each experiment ϫ 7 chorine ϫ ϫ ϭ The response of mussels to different chlorine concentrations was doses including control 3 size groups 4 replicates 1,680 studied in the laboratory, using animals collected from the Noordzee- mussels). kanaal, Velsen (near Ijmuiden), The Netherlands. Mussels attached on stones were collected from the (sub)littoral zone and transferred to the laboratory within 4 h. The experimental mussels for size comparisons Effects of Season were collected in winter (December), when water temperature was 5°C (Rajagopal et al. 1995). However, mussels for breeding versus non- M. leucophaeata in the Noordzeekanaal actively reproduce between breeding experiments were collected during summer (June) and winter June and October (Rajagopal et al. 1995). The mussels for experiments (December), respectively. In summer, the temperature was 20°C (Ra- were collected at two different times of the year viz., June (breeding) jagopal et al. 1995). Water collected from the field site was used to and December (nonbreeding) to determine any significant variation in acclimate mussels to the laboratory conditions. Prior to the experi- survival in bioassay. Mussels of 10 mm shell length were tested at ments, mussels were acclimated at 20°C for minimum of 2 weeks in a three (1, 2, and 3 mg/L) different chlorine concentrations (20 mussels temperature-controlled room. in each experiment ϫ 2 breeding and nonbreeding ϫ 4 chorine doses including control ϫ 1 size group ϫ 4 replicates ϭ 640 mussels).

Bioassay Preparation Starved Versus Fed Mussels Mussels were tested at six different chlorine concentrations (0.25, 0.5, 1, 2, 3, and 5 mg/L). The selected chlorine concentrations include To determine differences between starved and fed mussels, bioassays usual ranges used in power stations during normal and shock-dose were conducted at 1 mg/L residual chlorine with 10 mm size group. chlorination regimes (Jenner et al. 1998). Water collected from the One set of mussels was fed with mixed algal cells (20,000 cells were field was used for the experiment, after a day’s storage. Factors that added per ml of the brackish water in the reservoir) during both the Ϯ may influence the response of mussels such as salinity (mean SD; acclimation period and experiment, while other set of mussels were Ϯ Ϯ Ϯ 5.2 0.1‰), pH (7.8 0.2), temperature (20.1 0.3°C), dissolved starved (20 mussels in each experiment ϫ 2 starved and fed ϫ 2 Ϯ Ϯ oxygen (6.3 0.6 mg/L), suspended particulate matter (46.4 3.8 chorine doses including control ϫ 1 size group ϫ 4 replicates ϭ 320 Ϯ ␮ Ϯ mg/L), chlorophyll a (33.4 4.6 g/L), and flow rate (100 2.6 mussels). ml/min) were kept constant in each of the experimental treatments. The experiments were conducted in continuous once-through flow systems following procedures outlined by Rajagopal et al. (2002a). Brackish water was stored in a 100-L aquarium tank (reservoir) and Effects of Acclimation Temperature chlorine solution prepared from sodium hypochlorite was stored in a volumetric flask. An appropriate mix of the two were used to maintain Toxicity experiments were also conducted at 0.5 mg/L residual chlo- the desired chlorine concentration in 20-L experimental tanks having rine with 10 mm size group mussels. They were acclimated at six an outlet at the 17-L mark using a peristaltic pump. Mixing of the different acclimation temperatures (5, 10, 15, 20, 25, and 30°C) for 14 water with sodium hypochlorite was facilitated by the use of mixing days to determine if mussels living at different water temperatures pumps. A continuous flow at a rate of 100 ml/min was maintained would exhibit significant variation in survival in bioassays (20 mussels throughout the test. in each experiment ϫ 6 temperatures ϫ 2 chorine doses including control ϫ 1 size group ϫ 4 replicates ϭ 960 mussels). During the acclimation period, mussels were not fed. Bioassays

After 2 weeks of acclimation, mussels were allowed to colonize Statistical Analysis naturally to polystyrene artificial substrates (10 ϫ 10 cm square) of their own (Martin et al. 1993). The artificial substrates with 20 mussels The differences in mortality of different size groups of M. leucophae- of a known size group were introduced into the experimental tanks (20 ata at various chlorine doses were tested by analysis of covariance L) containing known chlorine concentration. The chlorine levels were (ANCOVA). The post hoc differences between size groups in each measured periodically at different locations in these experimental chlorine dose were compared by Student t tests after Bonferroni tanks to ensure a uniform distribution of chlorine residuals. Total corrections (Zar 1984). Before analysis, all data were tested for nor- residual chlorine (sum of free chlorine and combined chlorine) mea- mality and homogeneity of variation and log transformed, if necessary. surements were done using the iodometric method (White 1972). The The group-wise comparisons (starved versus fed, breeding versus criterion for mortality of mussels was valve gaping with no response nonbreeding) of mussel mortality at different chlorine concentrations 298 S. Rajagopal et al. were tested by Tukey’s pair-wise multiple comparison tests (Zar 1984). The data obtained on mortality of mussels at different chlorine doses were subjected to probit analysis, yielding the statistic LT50 and LT95 (Litchfield and Wilcoxon 1949). All analyses were performed using a Statistical Analysis Systems package (SAS 1989).

Results

No mortality occurred in control tanks. The time to 100% mortality of M. leucophaeata decreased with increasing chlo- rine concentration (Figure 1). The exposure time required to 100% mortality was analysed by ANCOVA for all six concen- trations of chlorine (0.25–5 mg/L) and was found to be highly Fig. 1. Time required for 100% mortality of different size groups of significant (chlorine dose effect: F ϭ 756.90, p Ͻ 0.001). Mytilopsis leucophaeata at different chlorine concentrations. Mortality (5,1673) Ϯ ϭ The three size groups of M. leucophaeata (2, 10, and 20 mm data are expressed as mean SD (n 80) of four replicate experi- ments (n ϭ 20 in each experiment). Mortality of M. leucophaeata was shell length) showed 100% mortality at significantly different monitored at 6-h intervals. The criterion for mortality of mussels was exposure times between 0.25 mg/L and 5 mg/L chlorine con- shell valve gape with no response of exposed mantle tissues to external ϭ Ͻ centrations (mussel size effect: F(2,1673) 31.74, p 0.001). stimuli For example, at 1 mg/L residual chlorine, 2 mm and 10 mm size group mussels took 35 and 46 days, respectively, to achieve 100% mortality (t tests, p Ͻ 0.05). No significant differences were found between replicate experiments (repli- ϭ Ͼ cates: F(3,1673) 0.42, p 0.05). For comparison, 100% mortality data of D. polymorpha and M. edulis are shown in Figure 2. Test methods and mortality determinations used for D. polymorpha and M. edulis were similar to M. leucophaeata. The exposure time required for 100% mortality of M. leuco- phaeata at different chlorine concentrations were much higher than that required for D. polymorpha and M. edulis.

The time to 50% mortality (LT50) of mussels was investi- gated by probit and regression analysis, and also shows that significant chlorine dose effect (ANCOVA, p Ͻ 0.001) and Ͻ size effect (ANCOVA, p 0.001) on LT50 of M. leucophae- ata. Comparisons of the six means of LT at different chlorine Fig. 2. Comparison of exposure times to reach 100% mortality of 50 Mytilopsis leucophaeata (shell length in mm Ϯ SD; 10.3 Ϯ 0.7), residuals (0.25–5 mg/L) showed that 0.25 mg/L exposure was Dreissena polymorpha (20.1 Ϯ 0.7), and Mytilus edulis (21.2 Ϯ 1.6) significantly less toxic than other chlorine concentrations. at different chlorine concentrations. Mortality data are expressed as The effect of breeding season on M. leucophaeata mortality mean Ϯ SD (n ϭ 80) of four replicate experiments (n ϭ 20 in each ϭ Ͻ due to chlorine toxicity was significant (F(1,638) 81.64, p experiment). Test methods and mortality determinations were similar 0.001). At all chlorine concentrations, time taken to reach in all toxicity studies of species 100% mortality in mussels collected during nonbreeding sea- sons was higher than that for mussels collected during breeding season (Figure 3). At 3 mg/L residual chlorine, nonbreeding Discussion mussels took 25 days to reach 100% mortality whereas breed- ing mussels took 19 days, a difference of 29% (see Figure 3). The present study is the first report of chlorine toxicity in the The fed and starved mussels showed 100% mortality at important biofouling species M. leucophaeata. The influence of similar exposure times at 1 mg/L residual chlorine (Tukey’s mussel size on tolerance to chlorine is significant in M. leuco- test, p Ͼ 0.05). For example, at 1 mg/L residual chlorine, fed phaeata. The tolerance is maximum in medium size mussels and starved mussels took 46 and 47 days, respectively, to reach (10 mm), with smaller and larger mussels (2 mm and 20 mm) 100% mortality. showing greater sensitivity (Figure 1). These results are in The effect of acclimation temperature on M. leucophaeata contrast with the results reported for M. edulis, where the mortality (Figure 4) in the presence of chlorine was significant tolerance linearly increases with shell size (Rajagopal et al. ϭ Ͻ (acclimation temperature effect: F(4,795) 27.18, p 0.001). unpublished data). Kilgour and Baker (1994) and Rajagopal et Time to 95% mortality of M. leucophaeata exposed to 0.5 al. (2002b) have reported that in the case of D. polymorpha mg/L at 5°C (99 days) was substantially higher than at 30°C shell size has no effect on chlorine toxicity. Therefore, com- (47 days). However, above 35°C, there was no significant parison of these three species proves that the relation between difference between combined use of temperature and chlorine mussel size and chlorine toxicity is not similar among different (0.5 mg/L residual chlorine) and temperature alone (Tukey’s mussel species and that generalizations regarding the size ef- test, p Ͼ 0.05). fect should be made carefully. Response of M. leucophaeta to Chlorination 299

Fig. 4. Comparison of exposure times to reach 95% mortality of Mytilopsis leucophaeata and Dreissena polymorpha at 0.5 mg/L re- sidual chlorine depending on the acclimation temperature. Rectangles: data of Rajagopal et al. (2002b); circles: Rajagopal et al. (1994); plus: data of Rajagopal et al. (1995, temperature alone); triangles: present study (Ϯ confidence limits); lines are linear regressions

and Baker 1994). They reported that mussels maintained on diet of Chlorella are consistently more sensitive to hypochlo- rite than starved mussels and attributed this to an increased tendency of these mussels to filter water, which increases exposure to chlorine. Clearly mussels which are fed with microalgae are likely to filter more water than those which are unfed. This explanation is not valid for the results obtained in present study because the chlorine concentration used (1 mg/L residual chlorine), acts as a strong suppressant of filtration activity. This has been conclusively shown by Rajagopal et al. (1997) using Mussel-monitor௡. Shell valve movement of M. Fig. 3. Cumulative mortality (%) of breeding and nonbreeding leucophaeata tested with unfiltered brackish water from the Mytilopsis leucophaeata at different chlorine concentrations (TRC ϭ Noordzeekanaal showed little or no filtration in presence of 1 total residual chlorine). Eighty mussels were used at each chlorine mg/L residual chlorine. Obviously, presence of microalgae dose. Mortality of M. leucophaeata was monitored at 6-h intervals would have no significant effect on M. leucophaeata in pres- ence of 1 mg/L residual chlorine. Therefore, feeding has no effect on chlorine toxicity in the M. leucophaeata. Interest- Mussels collected during the breeding season and the non- ingly, Rajagopal et al. (2002b) have also shown that filtration breeding season behaved quite differently with respect to their activity in D. polymorpha stops almost completely at 0.5 mg/L sensitivity to chlorine (Figure 3). Mussels collected during the residual chlorine. In view of results reported by Kilgour and breeding season were less tolerant to chlorine, whereas those Baker (1994), shell valve activity of mussels under chlorination collected during the nonbreeding season were more tolerant. needs to be reexamined in presence of microalgae. The difference in tolerance between the two groups was nearly The present study indicates that mussels acclimated to dif- 29%. Kilgour and Baker (1994) and Jenner et al. (1998) have ferent temperatures show a significantly different tolerance to reported similar results for D. polymorpha. They have attrib- chlorine (Figure 4). The decrease in acclimation temperature uted greater tolerance of mussels during the nonbreeding sea- from 30°Cto5°C increased chlorine tolerance by 52 days. son to low metabolic rates and reduced filtration rates, which Such increase in chlorine tolerance at lower acclimation tem- would result in reduced exposure to the toxicant. Lower ener- peratures have also been reported for other mussel species, like getic demands during nonbreeding seasons may be the reason D. polymorpha (Van Benschoten et al. 1995; Rajagopal et al. for reduced toxicant uptake. On the other hand, mussels tend to 2002b) and M. edulis (Lewis 1985; Jenner et al. 1998). How- be weaker after spawning when they have little energy resource ever, at acclimation temperatures above 35°C, temperature has in the body (Bayne et al. 1976), with the result that they are less overriding effects when compared to chlorine (Figure 4). These tolerant to biocide. The data point to the importance of judi- results are similar to those of Harrington et al. (1997) for D. cious sampling while doing toxicity experiments using sea- polymorpha. They observed that at 36°C combined use of sonal breeders, such as M. leucophaeata. temperature and chlorine resulted mortalities similar to those Status of feeding appears to have no significant effect on the obtained with heat alone. toxicity of chlorine to M. leucophaeata. Fed and starved mus- A comparison of present chlorine toxicity data with those for sels showed similar mortality rates in chlorine toxicity. This is other important fouling mussel species (such as M. edulis and in contrast with results published for D. polymorpha (Kilgour D. polymorpha) shows that M. leucophaeata is the most toler- 300 S. Rajagopal et al. ant, followed by M. edulis and D. polymorpha (Figure 2). This water management in European power stations: biology and con- increased tolerance is probably related to its euryhaline nature trol. Hydroecol App 1–2:1–225 (Siddall 1980). In the Noordzeekanaal (The Netherlands), the Kilgour BW, Baker MA (1994) Effects of season, stock, and labora- euryhaline nature capability of M. leucophaeata allows it to tory protocols on survival of zebra mussels (Dreissena polymor- coexist with M. edulis, which is a largely marine species, and pha) in bioassays. Arch Environ Contam Toxicol 27:29–35 D. polymorpha, which is a largely freshwater species (Van der Lewis BG (1985) Mussel control and chlorination. Report no. TPRD/ L/2810/R85, Central Electricity Research Laboratories, Leather- Velde et al. 1998). head, Surrey, England, p 33 Litchfield JT, Wilcoxon F (1949) A simplified method of evaluating dose-effect experiments. J Pharmac Exp Ther 2:99–113 Conclusions Marelli DC, Gray S (1983) Conchological redescriptions of and Mytilopsis leucophaeta of the brackish western Atlantic. Present studies have shown that various factors can influence Veliger 25:185–193 the response of M. leucophaeata in chlorine bioassays. Among Martin ID, Mackie GL, Baker MA (1993) Control of the biofouling the four parameters tested, mussel size, breeding season, and mollusc, Dreissena polymorpha (: ), with acclimation temperature have significant effect on chlorine sodium hypochlorite and with polyquaternary ammonia and ben- tolerance of M. leucophaeata, and nutritional status has no zothiazole compounds. Arch Environ Contam Toxicol 24:381– 388 significant effect. At acclimation temperatures above 35°C, Nyst HJP (1835) Mollusques. Bull Acad Roy Sci Brux 2:235–236 temperature has a more decisive effect on mortality than chlo- Oliver PG, Holmes AM, Mettam C (1998) Mytilopsis leucophaeta rination. Future toxicity experiments involving mussels should (Conrad, 1831) (Bivalvia: ): a species new to the describe mussel size, sampling season, and acclimation tem- British fauna. J Conchol 36:13–18 perature for proper comparison of results and planning chlorine Rajagopal S, Van der Velde G, Jenner HA (1994) Biology and control regimes in industries. The exposure time required for 100% of brackish water mussel, Mytilopsis leucophaeta in the Velsen mortality of M. leucophaeata at different chlorine concentra- and Hemweg power stations, The Netherlands. Part II. Heat treat- tions was much higher than that required for M. edulis and D. ment and chlorine. Report no. 63871-KES/WBR 94-3128, KEMA polymorpha. Therefore, chlorine treatment against mussels has Environmental Services, Arnhem, The Netherlands, p 45 to be employed judiciously depending on the mussel species Rajagopal S, Van der Velde G, Jenner HA (1995) Biology and control for effective fouling control. of brackish water mussel, Mytilopsis leucophaeta in the Velsen and Hemweg power stations, The Netherlands. Part I. Biology and behavioural response. Report no. 64211-KES 95-3109, KEMA Environmental Services, Arnhem, The Netherlands, p 55 Rajagopal S, Van der Velde G, Jenner HA (1997) Shell valve move- Acknowledgments. We thank Martin Versteeg for assistance with ment response of dark false mussel, Mytilopsis leucophaeta,to toxicity studies. We also thank V. P. Venugopalan for useful discus- chlorination. Wat Res 31:3187–3190 sions. KEMA Power Generation and Sustainables, Arnhem, The Neth- Rajagopal S, Van der Velde G, Jenner HA (2002a) Does status of erlands, financially supported this research. attachment influence survival time of , Dreissena polymorpha exposed to chlorination? Environ Toxicol Chem 21: 342–346 References Rajagopal S, Van der Velde G, Jenner HA (2002b) Effects of low- level chlorination on zebra mussel, Dreissena polymorpha. Wat Allonier AS, Khalanski M, Camel V, Bermond A (1999) Character- Res (in press) ization of chlorination by-products in cooling effluents of coastal SAS (1989) SAS/STAT user’s guide, release 6.4 edition. SAS Insti- nuclear power stations. Mar Pollut Bull 38:1232–1241 tute, Cary, NC Bayne BL, Thompson RJ, Widdows J (1976) Physiology. In: Bayne Siddall SE (1980) Early development of Mytilopsis leucophaeta (Bi- BL (ed) Marine mussels: their ecology and physiology. Cam- valvia: Dreissenacea). Veliger 22:378–379 bridge University Press, Cambridge, pp 121–206 Van Benschoten JE, Jensen JN, Harrington DK, DeGirolamo D (1995) Bidwell JR, Cherry DS, Farris JL, Petrille JC, Lyons LA (1999) Zebra mussel mortality with chlorine. J Am Water Wks Assoc Effects of intermittent halogenation on settlement, survival and 87:101–108 growth of the zebra mussel, Dreissena polymorpha. Hydrobiol Van der Velde G, Van der Gaag M, Rajagopal S, Jenner HA (1998) 394:53–62 Where do the exotic mussels Dreissena polymorpha and Mytilop- Boettger CR (1932) U¨ ber die Ausbreitung der Muschel Congeria sis leucophaeta meet in the brackish Noordzeekanaal, The Neth- cochleata Nyst in europa¨ischen Gewa¨ssern and ihr Auftreten im erlands? In: Eighth international zebra mussel and other aquatic Nordostseekanal. Zool Anz 101:43–48 nuisance species conference, 16–19 March 1998, Sacramento, Conrad TA (1831) Description of fifteen new species of recent, and CA,p54 three of fossil shells, chiefly from the coast of the United States. Vorstman AG (1935) Biologische Notizen betreffs der sessilen Fauna J Acad Nat Sci Philad 6:256–268 im Hafen der Stadt Amsterdam. Zool Anz 104:76–80 Harrington DK, Van Benschoten JE, Jensen JN, Lewis DP, Neuhauser White GC (1972) Handbook of chlorination. Van Nostrand Heinhold, EF (1997) Combined use of heat and oxidants for controlling adult New York, pp 228–277 zebra mussels. Wat Res 31:2783–2791 Zar JH (1984) Biostatistical analysis. Prentice-Hall, Englewood Cliffs, Jenner HA, Whitehouse JW, Taylor CJL, Khalanski M (1998) Cooling NJ p 718