Effects of Ozonation on the Speciation of Dissolved Iodine in Artificial Seawater

EFFECTS OF OZONATION ON THE SPECIATION OF DISSOLVED IODINE IN ARTIFICIAL SEAWATER

Johanna Sherrill, D.V.M., M.S., Brent R. Whitaker, M.S., D.V.M., and George T. F. Wong, Ph.D.

Abstract: Iodine in the form of iodide is required for synthesis of tri-iodothyronine and thyroxine in ®sh. Iodine chemical speciation in aliquots of raw arti®cial seawater mix was measured before, during, and after exposure for ®xed time periods to air only and to concentrations of ozone required to achieve oxidation±reduction potentials typical of a protein skimmer (400 mV) and an ozone contact chamber (800 mV). Chemical species of iodine were also measured in tank water from a large, recirculating, ozonated aquarium system that has a low-grade incidence of thyroid lesions (e.g., thyroiditis, hyperplasia, adenoma, and adenocarcinoma) in its ®sh. With increasing exposure to ozone, concentrations of iodide and dissolved organic iodine (DOI) decreased, whereas iodate levels increased. As a result of exposure to 400 mV, iodide concentration dropped to less than half the amount found in raw arti®cial seawater mix. After exposure to 800 mV, initial iodide levels decreased by 67%, and DOI became undetectable, whereas iodate concentration increased by 155%, with no remarkable change in total iodine concentration. These results indicate ozone-induced conversions from iodide to iodate, and DOI to iodide or iodate (or both). Iodide and DOI were not detectable in the aquarium system's water samples. Ozonation of arti®cial seawater may alter the relative concentrations of iodine species in a closed tank system, so that iodide supplementation of the diet or tank water of captive teleosts and elasmobranchs living in ozonated seawater is advisable. Key words: Fish, arti®cial seawater, iodine, ozone, thyroid disease, goiter.

INTRODUCTION have been documented in a variety of ®sh within Thyroid function and hormone regulation in ®sh natural and arti®cial systems, and a lack of iodine have been reviewed in relation to hormone concen- has been implicated as a causative agent in these 9,34 trations and conversions, growth stages, migration conditions. A gold®sh (Carassius auratus) goiter patterns, osmoregulation, and natural epizootics of was successfully treated with an organic iodine thyroid hyperplasia (goiter).4,10,12±14,17,23,24,31,34 In gen- compound at a concentration of 10 ppm.8 eral, ®sh thyroid hormone metabolism mimics that Fish in closed seawater aquaria may experience of most other vertebrates.32,35 In bony and cartilag- gradual or sudden decreases in the bioavailability inous ®sh, thyroid metabolism and thyroid disease30 of iodine, leading to the development of IϪ de®- are affected by dietary and environmental9,22 iodine ciency, thyroid disease, and even death.29,35 During levels.23,32 Studies of the chemical form of iodine the past several years, thyroid hyperplasia and neo- that is used in teleost or cartilaginous ®sh thyroid plasia have been detected in necropsied ®sh from metabolism are scarce.7,17,24 Iodide (IϪ) is required several different tanks at the National Aquarium in for the synthesis of thyroid hormones tri-iodothy- Baltimore (NAIB), including the Atlantic Coral 4,17,24 ronine (T3) and thyroxine (T4) in ®sh and other Reef (ACR) exhibit. The ACR is a three-story, 1.3 vertebrates.7 Iodine in an aqueous environment is ϫ 106±L habitat designed to look like a natural cor- more ef®ciently absorbed by diffusion across ®sh al reef that contains a variety of tropical bony and gill membranes than through the alimentary tract.33 cartilaginous ®sh species. Iodine in water provided rainbow trout (Oncorhyn- Speci®c lesions found in NAIB ®sh include thy- chus mykiss) with up to 84% of their plasma IϪ.22 roid hyperplasia in a neon goby (Elacatinus ocean- Thyroid hyperplasia9,21,30,34 and neoplasia2,29,30,41 ops) and in several bluestriped grunts (Haemulon sciurus), a thyroid adenoma penetrating major ar- From the Department of Animal Health, Smithsonian terial blood supply in a lookdown ®sh (Selene vo- National Zoological Park, 3001 Connecticut Avenue, mer), an expansive thyroid adenoma with marked, Washington, D.C. 20008, USA (Sherrill); the Department multifocal, granulomatous thyroiditis in a yellow- of Biological Programs, National Aquarium in Baltimore, headed jaw®sh (Opistognathus aurifrons), a thyroid Pier 3, 501 East Pratt Street, Baltimore, Maryland 21202, adenocarcinoma in®ltrating overlying epidermis in USA (Whitaker); and the Department of Ocean, Earth and an angel®sh (Pomancanthus sp.), and an invasive Atmospheric Sciences, Old Dominion University, 4600 Elkhorn Avenue, Norfolk, Virginia 23529-0276, USA thyroid adenoma causing pericardial compression (Wong). Present address (Sherrill): P.O. Box 5051, Rolling (tamponade) in a trumpet ®sh (Aulostomus macu- Hills Estates, California 90274, USA. Correspondence lates). Cases of thyroid disease in ®sh inhabiting should be directed to Dr. Whitaker. the ACR may be caused by environmental or die- 347 348 JOURNAL OF ZOO AND WILDLIFE MEDICINE tary iodine de®ciency. Identi®cation of factors as- seawater mix from the NAIB mixing vat. All water sociated with functional iodine de®ciency in this samples to be used for subsequent experiments aquarium exhibit should aid in the prevention of were taken from either of these two containers. The thyroid disease in future resident ®sh and may ap- noncommercial, proprietary formula for NAIB ar- ply to similar closed systems in other aquaria. ti®cial seawater mix contains 0.5 ␮MofIϪ or 0.08 As a trace element in the natural marine envi- mg/L (ppm) based on the addition of 38 g of po- ronment, dissolved iodine occurs either as inorgan- tassium iodide (KI) per 486,400 L of arti®cial sea- Ϫ Ϫ ic I and iodate (IO3 ) or as dissolved organic io- water. 42 dine (DOI). The presence of molecular iodine (I2) Parameters of a separately obtained aliquot of the and of hypoiodous acid has been hypothesized in NAIB arti®cial seawater mix were analyzed by in- 42 seawater but neither has been observed directly. house techniques, adapted from standard methods The distribution of iodine chemical species, as well for the examination of wastewater.26 In brief, pH as the biologic or abiologic processes that may was measured with a combination electrode cause transformations among IϪ,IOϪ, and DOI, 3 (ThermoOrion Model 920Aplus pH/mV/ISE meter, have been a major focus of studies concerning the ThermoOrion, Beverly, Massachusetts 01915, biogeochemistry of iodine in natural marine envi- USA), calibrated at three points daily. Salinity was ronments.42±44 Evidence for the biologic use of IO Ϫ 3 measured directly with a 0.475-cm, four-electrode or the simultaneous production of IϪ by marine phytoplankton has been reported in laboratory cul- cell conductivity meter (ThermoOrion Model tures.28,44 Iodide may react with naturally occurring 150Aplus, ThermoOrion). Phosphate was measured reactive transient oxidants, including ozone and hy- in seawater by spectrophotometric ascorbic acid digestion (Hach DR-4000 UV/VIS, Hach Co., Love- drogen peroxide, to form I2 and hypoiodous acid, and these products may further react with organic land, Colorado 80539-0389, USA). Simple titration molecules to form DOI.39,40 using a weak acid as the titrant and bromocresol as Large aquaria often use ozone, a potent oxidant, an indicator was used to measure alkalinity. to remove organic debris such as bacteria and vi- For each step of the experiment, a 1.5-L aliquot ruses3 from tank water. Ozone treatment of seawa- of NAIB raw arti®cial seawater was equilibrated to ter is an important part of water quality and life 25.5ЊC in a 2-L glass beaker set within an open support for aquatic animal exhibits1 and aquaculture water bath. Throughout the study, the temperature systems.37 Effects of ozonation on the chemical of sample water was monitored using a glass ther- speciation and therefore bioavailability of iodine in mometer immersed in the beaker of interest. The arti®cial seawater are important to measure because thermometer was rinsed copiously with deionized of the potential for altered iodine metabolism and water between uses. effects on thyroid health in ®sh housed in ozonated systems. Ozone measurement In an effort to eliminate several variables, the The ORPs were measured in millivolts using a effects of ozonation on speciation of dissolved io- portable ORP meter (Model 2000, VWR Scienti®c dine in arti®cial seawater made at NAIB were stud- Products, South Plain®eld, New Jersey 07080, ied experimentally. The result of adding ozone to USA) as an estimate of ozone concentration in the seawater is measurable in millivolts as the oxida- seawater samples. The probe stayed immersed in a tion±reduction potential (ORP). We postulated that pH-balanced probe solution when not in use and ozone concentrations required to achieve ORP lev- was rinsed thoroughly with deionized water be- els approximating those in an ozone contact cham- tween measurements of seawater samples. ber (800 mV) and a protein skimmer (400 mV) would measurably alter concentrations of iodine Sample description chemical species in arti®cial seawater, resulting in a de®ciency of bioavailable iodine for resident ®sh. Iodine chemical speciation in aliquots of NAIB Because the ACR exhibit has a low-grade, docu- raw arti®cial seawater mix at 25ЊC was compared mented incidence of thyroid lesions in its ®sh, an before, during, and after exposure for ®xed time additional goal of our study was to measure the periods to air only or to concentrations of ozone existing levels of iodine chemical species in this needed to achieve ORPs of 400 and 800 mV. Three closed, ozonated aquarium system. separate laboratory trials were run to perfect methods. Water samples generated in the third trial were MATERIALS AND METHODS then analyzed for concentrations of iodine species. Sample water Samples from the ACR were collected from both Two clean, empty 11.4-L plastic containers were the tank surface and the water supply downstream rinsed three times and then ®lled with raw arti®cial from the ozone contact chamber and were placed SHERRILL ET AL.ÐOZONATION EFFECTS ON IODINE IN ARTIFICIAL SEAWATER 349

Table 1. Sample descriptions and corresponding ORPs of nine samples of arti®cial seawater from the NAIB collected during an experiment designed to investigate the effects of ozonation on the chemical speciation of dissolved iodine. Samples 1±7 were generated by a laboratory experiment, and samples 8 and 9 were collected from the ACR, a system with a recent history of thyroid disease in some of its ®sh.a

No. Sample description ORP (mV)

1 NAIB raw arti®cial seawater mix, aerated for 2 min to mix sample before collection 175 2 NAIB raw arti®cial seawater mix, aerated for 14 min at 1 L/min 185 3 NAIB raw arti®cial seawater mix, aerated for an additional 5 min of aeration 185 4 Ozonated samples: Ozone 400. Ozonated to simulate a protein skimmer average ORP level 500 of 400 mV 5 Ozone 400, degassed. After additional aeration to normalize ORP 277 6 Ozone 800. Ozonated to simulate maximum ozone contact chamber ORP level of 800 mV 786 7 Ozone 800, degassed. After additional aeration to normalize ORP 269 8 ACR ®ltration, downstream from ozone contact chamber. Postozonation before delivery into 239 main exhibit. Expected ORP was 400 mV 9 ACR display tank surface. Approximately 0.5±1 m below surface, where ®sh often swim 235

a ORP, Oxidation±reduction potential; NAIB, National Aquarium in Baltimore; ACR, Atlantic Coral Reef. directly into plastic sampling bottles as described ent regions of the ACR, an inhabited system with below. Laboratory and ACR water samples were many variables. Sample 8 was collected from a pipe analyzed simultaneously to initially compare iodine immediately downstream from the ozone contact chemical species' concentrations in laboratory-gen- chamber of the ACR. The expected or intended erated samples with those occurring in a working ORP at this location was approximately 400 mV, aquarium system. but the actual level at time of sampling was much Descriptions and corresponding ORPs of the nine less (239 mV). Sample 9 was collected from the samples collected for analysis are presented in Ta- top of the ACR exhibit where ®sh routinely swim ble 1. Points of sampling were based on laboratory near the surface (depth range 0.5±1.0 m with an simulation of arti®cial seawater traveling from its ORP of 235 mV). initial source, through ozonation and ®ltration pipe- All samples were stored in the dark at Ϫ25ЊC lines, and into the display portion of the ACR. Sam- from the time immediately after collection until ple 1 consisted of NAIB raw arti®cial seawater analysis 3 mo later. Storage times of 3 mo or less mixed by aeration only for a total of 2 min. Sample have minimal effect on determination of iodine 2 was NAIB raw arti®cial seawater subjected to 14 chemical species levels, and cold storage is pre- min of aeration only, approximating time required ferred.6 Samples from the ACR were stored simi- for water to travel through the ACR pipelines and larly after a 15-min transit time from tank to freez- ozonation chamber. Sample 3 was collected after er. an additional 5 min of aeration to mimic degassing of seawater while traveling from the ozone contact Equipment chamber to the display portion of the ACR. In the New polyethylene tubing was used to connect next phase of the experiment, NAIB raw arti®cial the various components of the test apparatus. A ta- seawater was ozonated in two separate ways. Sam- bletop ozone generator (120 V, 60 Hz, 30 W; Model ple 4 consisted of an aliquot collected after 4 min D-06, Aqua-Flo, Inc., Baltimore, Maryland 21206, of ozone exposure at an ORP of approximately 400 USA), powered by a small air compressor (5±7 L/ mV, intended to simulate ORP levels occurring as min maximum output), was linked by approximate- water travels through a typical protein skimmer. Af- ly 0.5 m of tubing to a ¯ow meter (1±5 L/min ca- ter the ozonated seawater stabilized to an ORP of pacity). Another 0.5-m length of tubing connected 290±315 mV, the normal range at the display level the ¯ow meter to a wooden rectangular airstone of a tank system at NAIB, sample 5 was taken to (approximately 4 ϫ 2 ϫ 1 cm) by way of a conical measure iodine chemical species after degassing plastic adaptor. Approximately 1.0 m of tubing ex- from 400 mV. Samples 6 and 7 were taken after 4 tended from the air compressor to the outside of min at maximum ozone contact chamber ORP level the fume hood to serve as an inlet for laboratory of approximately 800 mV and after degassing to an room air. For each stage of the experiment, the test ORP range of 290±315 mV, respectively. beaker was ®lled with a fresh aliquot of approxi- Samples 8 and 9 were collected from two differ- mately 1.5 L of NAIB raw arti®cial seawater, main- 350 JOURNAL OF ZOO AND WILDLIFE MEDICINE tained at 25.5ЊC using an open water bath. A cen- larger, and a detection limit of approximately 0.02 tigrade glass thermometer was used to continuously ␮M, according to proven methods.19,45,46 The con- monitor the temperature of the seawater. centration of DOI in each sample was determined Care was taken to minimize organic material on indirectly using previously developed methods.43 In the laboratory equipment and containers used in the brief, DOI and IϪ were quantitatively oxidized to Ϫ experiment. The laboratory trial was executed in a IO3 by ultraviolet irradiation exposure for 3 hr us- clean, chemical fume hood, with the exhaust run- ing a 700-W mercury vapor lamp in a photochem- ning at approximately 50% capacity during the ex- ical platform reactor (Ace Glass, Inc., Vineland, periment. Glass beakers and the two ACR sample New Jersey 08362-0688, USA) in the presence of bottles (250 ml, dark brown, polyethylene; Nalge- hydrogen peroxide. Next, the concentration of total Ϫ ne, Nalge Nunc International, Rochester, New York I in the sample was measured as IO3 , with a pre- 14625, USA) were prepared by rinsing with nitric cision of Ϯ0.01 ␮M.38 Finally, the concentration of acid and deionized water and then steam autoclav- DOI was calculated as the difference between total Ϫ ing for 20 min at 834 kPa (121 psi). Seven sample I concentration and total inorganic iodine (IO3 ϩ bottles (175 ml, opaque, polyethylene; Nalgene, IϪ), with a precision of Ϯ0.02 ␮M. Thus, the con- Ϫ Nalge Nunc International) were used for the bench- centrations of total I, IO3 , and DOI are reported to top experiment. Each bottle was soaked in distilled, two decimal places, whereas IϪ is given to three deionized water (type I reagent-grade water) for 24 decimal places. Ϫ hr, cleaned with laboratory detergent, rinsed with One separate subsample was analyzed for IO3 , copious amounts of distilled, deionized water (to IϪ, and total I individually. However, for the deter- remove any residual detergent), leached with 10% mination of each iodine species, the polarogram Ϫ (v/v) hydrochloric acid (American Chemical Soci- (for the determination of IO3 and total I) or vol- ety grade) for 24 hr, and then rinsed thoroughly tammogram (for the determination of IϪ) of each with distilled, deionized water (to remove any re- subsample was repeated at least three times and the sidual acid). average of these readings was used for calculating The glass thermometer and the wooden airstone the concentration of each dissolved iodine species. were soaked for 20 min in a cold sterilization solution (Cidex-Plus, Johnson & Johnson Medical RESULTS Inc., Arlington, Texas 76004, USA) and rinsed gen- Water quality parameters of raw arti®cial sea- erously with deionized water after each use. Sterile water mix from the NAIB, measured during the 35- and 60-ml plastic syringes (Monoject௢, Tyco same week as the experiment, were as follows: pH Healthcare Group LP, Mans®eld, Massachusetts 7.87, phosphate ϭ 50 ϫ 10Ϫ6 ␮mol (0.05 mg/L), 02048, USA) were used to draw up and transfer alkalinity ϭ 4 mEq/L (200 mg/L), and salinity ϭ samples of 120±150 ml each into the plastic bottles. 33 g/L, all considered within normal limits. The Ϫ Ϫ After initial use, the syringes were subsequently concentrations of IO3 ,I, DOI, and total I mea- cleaned for each stage of the experiment by a min- sured in an aliquot (sample 1) of NAIB raw arti®- imum of three deionized water rinses followed by cial seawater mix (0.18, 0.137, 0.11, and 0.43 ␮M, sterilization in a steam autoclave for 20 min at 834 respectively) were within the ranges of typical lev- kPa (121 psi). els found in the surface seawater of natural marine systems.42 Sample analysis Concentrations of the iodine chemical species in Sample bottles were transferred from the labo- the nine samples of NAIB arti®cial seawater gen- ratory freezer at NAIB to a dry-ice shipper and sent erated by one experimental run are presented as by overnight courier to GTFW's research labora- Figure 1. Samples 1±7 were collected experimen- tory where concentrations of iodine species in each tally. The concentration of total I in these samples sample were individually analyzed. Concentration ranged from 0.43 to 0.48 ␮M with an average of of total iodine (total I) per sample consisted of the 0.47 Ϯ 0.02 ␮M. After a sample had been treated sum of the concentrations of inorganic plus organic with aeration or ozonation (samples 2, 4, and 6), Ϫ Ϫ iodine species, i.e., [IO3 ] ϩ [I ] ϩ [DOI]. Iodide further degassing of each sample (samples 3, 5, and was determined directly using cathodic stripping 7) resulted in only minor changes in the concentra- square wave voltammetry with a precision of Ϯ3% tions of the dissolved iodine species. Aeration of or Ϯ0.005 ␮M, whichever was larger, according to the NAIB raw arti®cial seawater alone also had established methods.25,45,46 Iodate was determined minimal to undetectable effects on the iodine spe- directly using differential pulse polarography with cies' concentrations. The concentration differences Ϫ Ϫ a precision of Ϯ3% or Ϯ0.01 ␮M, whichever was in IO3 , DOI, and I with and without aeration SHERRILL ET AL.ÐOZONATION EFFECTS ON IODINE IN ARTIFICIAL SEAWATER 351

Figure 1. Iodine chemical speciation in arti®cial seawater made and used at the National Aquarium in Baltimore: initial, after aeration only, and after ozonation. Concentrations (in ␮M) of iodine chemical species in samples of arti®cial seawater were measured before and after exposure to air and then to two different concentrations of ozone (400 and 800 mV). Concentrations of the sum of iodide and iodate were only minimally different between aerated samples and the aquarium's raw arti®cial seawater mix. This in vitro experiment reveals a systematic decrease in concentrations of iodide and dissolved organic iodine and a concomitant increase of iodate with increasing exposure to ozone.

(samples 1 and 2) were 0.02 Ϯ 0.02, 0.06 Ϯ 0.04, (0.45 ␮M) measured in the NAIB raw arti®cial sea- and 0.03 Ϯ 0.01 ␮M, respectively. water mix. In contrast, ozonation led to readily discernible changes in the concentrations of iodine species. The DISCUSSION concentrations of IϪ and DOI decreased, whereas Ϫ In the seven samples collected from the labora- the concentration of IO3 increased systematically with increasing exposures to ozone (samples 1, 4, tory experiment, any variation in the concentration and 6). After exposure to 800 mV of ozone, 0.09 of total I from the average value was similar to the ␮M (67%) of the initial IϪ disappeared and DOI minimal analytical uncertainties of the experimental became undetectable, whereas the concentration of methods used. This suggests that during the exper- Ϫ IO3 increased by 0.29 ␮M (155%). The concom- iment, no appreciable gain or loss of dissolved io- Ϫ itant increase in the concentration of IO3 exceeded dine to the gaseous or particulate phase occurred. the decrease in the concentration of IϪ. This result- Changes in the concentrations of the dissolved io- ed in an increase in the concentration of inorganic dine species were due to interconversions among iodine (i.e., the sum of the concentrations of IϪ and the chemical species. Furthermore, the concentra- Ϫ IO3 ) of 0.18 ␮M (56%). tions of the dissolved iodine species were suf®- The water samples (11 and 12) from the ACR ciently stable that they were not changed during were devoid of IϪ and DOI. Concentration of total degassing periods; thus, interaction between a sam- I from the ACR samples averaged 0.11 ␮M, which ple and the surrounding laboratory environment, is approximately a fourth of the amount of total I such as the overlying laboratory room air, did not 352 JOURNAL OF ZOO AND WILDLIFE MEDICINE contribute materially to any observed changes in water are complex.1,20,26 Ozonation of seawater and the speciation of iodine. subsequent effects on important ions, such as bro- Dissolved iodine occurs either as inorganic spe- mide, chloride, and IϪ, have been studied exten- Ϫ Ϫ 16,18,36 cies, I and iodate IO3 , or as DOI in surface sea- sively. Reactions between ozone and iodine water.42 These three species of iodine were found species are well known because the classical io- in the raw seawater mix (sample 1), and their con- dometric methods for the determination of ozone in Ϫ centrations were similar to those found typically in water samples involve the oxidation of I to I2 and surface seawater.42 On aeration, induced concentra- hydrolized organic iodine (HOI) under alkaline tion changes were minor and could be attributed to conditions.5 These iodine species may further react inherent minor analytical uncertainties alone. This with organic molecules to form DOI.39,40 Further- further indicates that the changes in the concentra- more, because I2 and HOI are volatile, these inter- tions of the iodine species on ozonation must have mediates may be lost to the gaseous phase. In one resulted from their reactions with ozone. The sys- laboratory-based study investigating the interaction tematic and conspicuous rise in the concentration of ozone with iodine in surface seawater samples, Ϫ of IO3 and the concomitant decrease in the con- ozonation induced the release of iodine vapor, a centration of IϪ on increasing ozonation of NAIB process postulated to contribute to iodine cycling arti®cial seawater in vitro clearly demonstrate between the atmosphere and the ocean's.16 Ϫ Ϫ ozone-induced conversions of I into IO3 in a con- The rate of decomposition of ozone is affected trolled laboratory setting. The increase in the con- by pH, temperature, organic and inorganic sub- centration of inorganic iodine and the accompany- strates, amount of mixing, and competing oxida- ing decrease in the concentration of DOI on expo- tion±reduction reactions.36 The formation of highly sure to ozone indicate the conversion of DOI to reactive oxidants, such as hydroxyl radical (OH•), Ϫ Ϫ IO3 or I . formed during the autodecomposition of aqueous The concentrations of IϪ, DOI, inorganic iodine, ozone has been described.36,37 Oxidant radicals pre- and total I in the water samples from the ACR sys- sent in ozonated seawater may react with dissolved tem containing ®sh (samples 8 and 9) were all ap- iodine, thereby affecting the concentration of iodine preciably lower than those in the source water species in a closed system like the ACR. (NAIB raw arti®cial seawater mix). In fact, the con- De®ciency of iodine has been proposed as a po- centrations of IϪ and DOI dropped to undetectable tential etiology of thyroid hyperplasia (goiter) in levels, whereas those of inorganic iodine and total ®sh.17,23,33,35 Other factors, including the presence of I were approximately a third and a quarter of those goitrogens, may alter iodine metabolism and avail- in the NAIB raw arti®cial seawater mix, respec- ability and subsequently affect thyroid health.9,21,30,34 tively. Thus, in addition to the interconversions An aquatic environment can contain a variety of among the dissolved iodine species, there was also suspected goitrogenic compounds, including am- a loss of iodine to the particulate or gaseous phase monia, urea, nitrates, nitrites, and established chem- on repeated circulation of the arti®cial seawater mix ical toxins.2,11,35 In one study comparing hypothy- through the ACR system. Although this study was roidism in two different populations of captive whi- not designed to fully investigate the fate of iodine tetip reef sharks (Triaenodon obesus), thyroid hy- species in the ACR system, these changes are not perplasia in the sharks was linked to low levels of inconsistent with what is known about the biogeo- IϪ and high levels of nitrates acting as goitrogens chemistry of iodine in the natural marine environ- in the environment.9 Thyroid lesions in teleost and ment. Both biologic and abiologic processes have cartilaginous ®sh can result in a variety of second- been proposed for mediating the transformations ary effects, including destruction of thyroid-adja- Ϫ Ϫ 6,42±44 among I ,IO3 , and DOI. Dissolved organic cent cardiorespiratory tissues, or metastasis to other iodine can be converted to IϪ by photochemical re- vital organs, leading to respiratory compromise, 43 Ϫ 2,29 actions. Evidence for the biologic use of IO3 or emaciation, and potentially death. At NAIB, ®sh the simultaneous production of IϪ by marine phy- with suspected thyroid disease receive a complete toplankton has been reported in laboratory cul- postmortem examination, including histopathology. 28,44 Ϫ Ϫ tures. In the process, IO3 is converted to I and Evaluation of gross and microscopic lesions direct- dissolved iodine can be incorporated into the par- ly applies to the development of preventive care ticulate phase. strategies for resident ®sh. Iodide reacts with naturally occurring reactive Our results indicate that the repeatedly circulated transient oxidants, including ozone and hydrogen and ozonated water in the ACR system of the NAIB 42 peroxide, to form I2 and hypoiodous acid. The re- contained 0.11 ␮M of dissolved iodine. The iodine Ϫ actions between ozone and the constituents of sea- was exclusively in the form of IO3 , which is not SHERRILL ET AL.ÐOZONATION EFFECTS ON IODINE IN ARTIFICIAL SEAWATER 353 readily bioavailable to resident ®sh. Minimum IϪ duce stress.15 In one report, toxic effects of ozone levels of 0.10 and 0.15 ␮M have been suggested as were detected in gills and erythrocytes and acute guidelines to prevent goiter in multiple species of death occurred in Japanese charr (Salvelinus leu- captive whitetip reef sharks9 and teleost ®sh,35 re- comaenis) at ozone concentrations less than 0.7 mg/ spectively. In a report using the same iodine mea- L and greater than 0.7 mg/L, respectively.15 Dele- surement methodologies as in this study, hypothy- terious effects of atmospheric ozone can be poten- roidism and goiter occurred in whitetip reef sharks tiated by diet or preexisting disease;27 however, housed in seawater that had low environmental IϪ these are lessened in seawater because of certain (Ͻ0.005 ␮M) but not in a separate group of sharks characteristics of aqueous ozone, such as lower to- kept in seawater with high IϪ (0.60 ␮M).9 Thus, tal doses, partial solubility in water, and a short there is apparently a severe iodine de®ciency in the half-life (usually less than 1 hr).1,36,37 ACR system for the resident ®sh. A substantial aug- mentation of IϪ coupled with a consistent monitor- CONCLUSIONS ing program for dissolved iodine content in the ar- Ozonation of arti®cial seawater may alter the rel- ti®cial seawater supply at both source and exhibit ative concentrations of iodine chemical species pre- levels may be required to prevent an iodine de®- sent in a closed tank system, such that IϪ and DOI ciency and subsequent thyroid disease in resident Ϫ levels decrease and IO3 levels increase. Samples ®sh. from the ACR exhibit were depleted of IϪ and DOI. Alternatively, iodine de®ciency may be remedied It is possible that losses of IϪ and DOI through with oral dietary supplementation, which is com- oxidation by ozone or biologic processes is linked mon practice in many aquaria, especially for carti- to thyroid diseases documented in ACR ®sh over laginous ®sh. Guidelines addressing oral iodine re- recent years. Thus, it may be advisable to supple- quirements are primarily formulated for food ®sh, ment the diets of captive teleosts and elasmo- particularly salmonids. For example, Chinook salm- branchs living in ozonated seawater with IϪ or to on (Oncorhynchus tshawytscha) should be fed 0.6± add IϪ directly to tank water downstream from the 1.1 mg of iodine per kg of ration based on actual entry of ozonated water into the system. Ideally, 33 feed analysis. Certain NAIB sharks diagnosed weekly to biweekly testing of tank water for IϪ, with thyroid disease have been successfully treated Ϫ IO3 , and total dissolved iodine levels is recom- with a daily calcium iodate supplement (Mazuri௢ mended to detect ¯uctuations in available iodine, to Vita-Zu Sharks/Rays, Mazuri Feeds, Purina Mills- maintain a desired IϪ range of 0.10±0.15 ␮M, and L.L.C., St. Louis, Missouri 63144, USA) added to to aid in the calculation of appropriate supplement their diet at the suggested dosage of one tablet per dosages for resident ®sh. Complete postmortem ex- 0.23 kg of ®sh fed. amination including histopathology should be per- Ozone dosage ranges from 0.5 to 1.0 mg/L and formed in ®sh with suspected or proven thyroid dis- 1.0 to 2.0 mg/L, with a programmed ozone cham- ease, and this may lead to improved management ber contact time of 4 min, are used in NAIB sys- strategies for captive ®sh collections. tems that contain ®sh and marine mammals, re- A host of biologic and abiologic processes, such spectively. In NAIB aquariums, the oxidative ef- as pH, temperature, organic substances,39,40 and io- fects of ozone, which can be measured as ORP, are dine metabolism by microorganisms,28,44 have been typically low (290±310 mV) compared with ozone reported to affect iodine chemical speciation in both contact chamber levels (700±800 mV). An ORP arti®cial and natural seawater systems.42,43 Future reading exceeding a range of 180±240 mV de®nes studies investigating some of these processes in the presence of oxidants or ozone in a water sam- NAIB tank systems, as well as identi®cation and ple, whereas low ORP readings (Ͻ180 mV) signify quanti®cation of potential goitrogens or thyrotox- that few to no oxidants are present.1 Residual ozone ins, and any correlation with thyroid disease inci- (0.1±0.4 mg/L3 or an ORP Յ450 mV)1 is an im- dence in resident ®sh are warranted. portant component of water disinfection processes because it has destructive oxidative effects on bac- Acknowledgments: We thank the skilled staff at teria and viruses.3,5,26 the National Aquarium in Baltimore, especially It is feasible for residual ozone or oxidant by- Ryan Bromwell, April Smith, Jill Arnold, and Liz products in tank water to cause adverse effects on Neely (laboratory support); Jill Arnold and Andy ®sh housed in ozonated systems, either through di- Aiken (technical and editorial support); Susie Ri- rect contact with ozone or through the formation of denour (library services); Valerie Lounsbury (edi- oxidation products,37 e.g., hypohalites of chlorine torial input); and Michele Martin and Ian Walker and bromine18 can damage gills or mucosa and in- (help on animal health). Financial support for io- 354 JOURNAL OF ZOO AND WILDLIFE MEDICINE dine analyses was granted through the National of gill and erythrocytes of Japanese charr (Salvelinus leu- Aquarium in Baltimore Biological Programs De- comaenis). Comp. Biochem. Physiol. C. 101: 331±336. partment. 16. Garland, J. A., and H. Curtis. 1981. Emission of iodine from the sea surface in the presence of ozone. J. LITERATURE CITED Geophys. Res. 86: 3183±3186. 17. Gorbman, A. 1969. Thyroid function and its control 1. Aiken, A. 1995. Use of ozone to improve water in ®shes. In: Hoar, W. S., and D. J. Randall (eds.). Fish quality in aquatic exhibits. Int. Zoo Yearb. 34: 106±114. Physiology (II). Endocrinology. Academic Press, Inc., 2. Blasiola, G. C., J. C. Turnier, and E. E. Hurst. 1981. New York, New York. Pp. 241±274. Metastatic thyroid adenocarcinoma in a captive population 18. Grguric, G., J. H. Trefry, and J. K. Keaffaber. 1994. of kelp bass, Paralabrax clathratus. J. Natl. Cancer Inst. Ozonation products of bromine and chlorine in seawater 66: 51±59. aquaria. Water Res. 28: 1087±1094. 3. Block, J. C. 1982. Removal of bacteria and viruses 19. Herring, J. R., and P. S. Liss. 1974. A new method by ozonation. In: Masschelein, W. J. (ed.). Ozonation for the determination of iodine species in seawater. Deep- Manual for Water and Wastewater Treatment. John Wiley Sea Res. 21: 777±783. & Sons, New York, New York. Pp. 66±68. 20. Hoigne, J. 1982. Reaction pathways and kinetics of 4. Bonga, S. E. W. 1993. Endocrinology. In: Evans, D. reactions of ozone with compounds dissolved in water. In: H. (ed.). The Physiology of Fishes. CRC Press, Inc., Boca Masschelein, W. J. (ed.). Ozonation Manual for Water and Raton, Florida. Pp. 469±502. Wastewater Treatment. John Wiley & Sons, New York, 5. Brener, L. 1986. Methods of determination of ozone New York. Pp. 57±62. in air and in water. In: Rice, R. G., L. J. Bollyky, and W. J. Lacy (eds.). Analytical Aspects of Ozone Treatment of 21. Hoover, K. L. 1984. Hyperplastic thyroid lesions in Water and Wastewater. Lewis Publishers, Inc., Chelsea, ®sh. Natl. Cancer Inst. Monogr. 65: 275±289. Michigan. Pp. 53±70. 22. Hunt, D. W. C., and J. G. Eales. 1979. Iodine bal- 6. Campos, M. L. A. M. 1997. New approach to eval- ance in rainbow trout (Salmo gairdneri) and effects of uating dissolved iodine speciation in natural waters using testosterone propionate. J. Fish. Res. Board Can. 36: 282± cathodic stripping voltammetry and a storage study for 285. preserving iodine species. Mar. Chem. 57: 107±117. 23. Leatherland, J. F. 1982. Environmental physiology 7. Carrasco, N. 1993. Iodide transport in the thyroid of the teleostean thyroid gland: a review. Environ. Biol. gland. Biochim. Biophys. Acta 1154: 65±82. Fishes 7: 83±110. 8. Conroy, D. A., and J. A. Santacana. 1979. A case of 24. Leatherland, J. F. 1994. Re¯ections on the thyroid- goiter and its treatment in a gold®sh Carassius auratus ology of ®shes: from molecules to humankind. Guelph (L.). J. Fish Dis. 2: 555±556. Ichthyol. Rev. 2: 1±68. 9. Crow, G. L., M. J. Atkinson, B. Ron, S. Atkinson, 25. Luther, G. W., C. B. Swartz, and W. J. Ullman. A. D. Skillman, and G. T. F. Wong. 1998. Relationship of 1988. Direct determination of iodide in seawater by ca- water chemistry to serum thyroid hormones in captive thodic stripping square wave voltammetry. Anal. Chem. sharks with goitres. Aquat. Geochem. 4: 469±480. 60: 1721±1724. 10. Crow, G. L., B. Ron, S. Atkinson, and L. E. L. 26. Masschelein, W. J. (ed.). 1982. Ozonation Manual Rasmussen. 1999. Serum T4 and serum T3 concentrations for Water and Wastewater Treatment. John Wiley & Sons, in captive whitetip reef sharks, Triaenodon obesus. J. Exp. New York, New York. Pp. 166±177. Zool. 284: 500±504. 27. Mehlman, M. A., and C. Borek. 1987. Toxicity and 11. Doerge, D. R., H. C. Chang, R. L. Divi, and M. I. biochemical mechanisms of ozone. Environ. Res. 42: 36± Churchwell. 1998. Mechanism for inhibition of thyroid 53. peroxidase by leucomalachite green. Chem. Res. Toxicol. 28. Moisan, T. A., W. M. Dunstan, A. Udomkit, and G. 11: 1098±1104. T. F. Wong. 1994. The uptake of iodate by marine phy- 12. Eales, J. G., and S. B. Brown. 1993. Measurement toplankton. J. Phycol. 30: 580±587. and regulation of thyroidal status in teleost ®sh. Rev. Fish 29. Nigrelli, R. F. 1952. Spontaneous neoplasms in Biol. Fish. 33: 299±347. ®shes (VI). Thyroid tumors in marine ®shes. Zoologica 13. Eales, J. G., J. A. Holmes, J. M. McLeese, and J. 37: 185±189. H. Youson. 1997. Thyroid hormone deiodination in vari- 30. Nigrelli, R. F., and G. D. Ruggieri. 1973. Hyper- ous tissues of larval and upstream-migrant sea lampreys, plasia and neoplasia of the thyroid in marine ®shes. Mt. Petromyzon marinus. Gen. Comp. Endocrinol. 106: 202± Sinai J. Med. 41: 283±293. 210. 31. Parker, S. J., and J. L. Specker. 1990. Salinity and 14. Finnson, K. W., and J. G. Eales. 1997. Glucuroni- temperature effects on whole-animal thyroid hormone lev- dation of thyroxine and 3,5,3Ј triiodothyronine by hepatic els in larval and juvenile striped bass, Morone saxatilis. microsomes in rainbow trout, Oncorhynchus mykiss. Fish Physiol. Biochem. 8: 507±514. Comp. Biochem. Physiol. C. Pharmacol. Toxicol. Endo- 32. Phillips, A. M., Jr. 1969. Nutrition, digestion, and crinol. 117: 193±199. energy utilization. In: Hoar, W. S., and D. J. Randall 15. Fukunaga, K., T. Suzuki, M. Arita, S. Suzuki, A. (eds.). Fish Physiology (I). Excretion, Ionic Regulation, Hara, K. Yamauchi, N. Shinriki, K. Ishizaki, and K. Tak- and Metabolism. Academic Press, Inc., New York, New ama. 1992. Acute toxicity of ozone against morphology York. Pp. 406±407. SHERRILL ET AL.ÐOZONATION EFFECTS ON IODINE IN ARTIFICIAL SEAWATER 355

33. Post, G. W. 1993. Nutrition and nutritional diseases 40. Truesdale, V. W., G. W. Luther, III, and C. Canosa- of salmonids. In: Stoskopf, M. K. (ed.). Fish Medicine. Mas. 1995. Molecular iodine reduction in seawater: an W. B. Saunders Company, Philadelphia, Pennsylvania. Pp. improved rate equation considering organic compounds. 343±358. Mar. Chem. 48: 143±150. 34. Sonstegard, R., and J. F. Leatherland. 1976. The 41. Wellings, S. R. 1969. Neoplasia and primitive ver- epizootiology and pathogenesis of thyroid hyperplasia in tebrate phylogeny: echinoderms, prevertebrates, and ®sh- coho salmon (Oncorhynchus kisutch) in Lake Ontario. es. Natl. Cancer Inst. Monogr. 31: 59±128. Cancer Res. 36: 4467±4475. 42. Wong, G. T. F. 1991. The marine geochemistry of 35. Spotte, S. 1992. Captive Seawater Fishes. John Wi- iodine. Rev. Aquat. Sci. 4: 45±73. ley and Sons, Inc., New York, New York. 43. Wong, G. T. F., and X. H. Cheng. 1998. Dissolved 36. Stanley, J., and D. Johnson. 1986. Analysis of organic iodine in marine waters: determination, occur- ozone in aqueous solution. In: Rice, R. G., L. J. Bollyky, rence and analytical implications. Mar. Chem. 59: 271± and W. J. Lacy (eds.). Analytical Aspects of Ozone Treat- 281. ment of Water and Wastewater. Lewis Publishers, Inc., 44. Wong, G. T. F., A. U. Piumsomboon, and W. M. Dunstan. 2002. The transformation of iodate to iodide in Chelsea, Michigan. Pp. 71±90. marine phytoplankton cultures. Mar. Ecol. Prog. Ser. 237: 37. Summerfelt, S. T., and J. N. Hochheimer. 1997. Re- 27±39. view of ozone processes and applications as an oxidizing 45. Wong, G. T. F., and L.-S. Zhang. 1992. Chemical agent in aquaculture. Prog. Fish-Cult. 59: 94±105. removal of oxygen with sul®te for the polarographic or 38. Takayanagi, K., and G. T. F. Wong. 1986. The ox- voltammetric determination of iodate or iodide in seawa- idation of iodide to iodate for the polarographic determi- ter. Mar. Chem. 38: 109±116. nation of total iodine in natural waters. Talanta 33: 451± 46. Wong, G. T. F., and L.-S. Zhang. 1992. Determi- 454. nation of total inorganic iodine in seawater by cathodic 39. Truesdale, V. W., and G. W. Luther, III. 1995. Mo- stripping square wave voltammetry. Talanta 39: 355±360. lecular iodine reduction by natural and model organic substances in seawater. Aquat. Geochem. 1: 89±104. Received for publication 10 March 2003