Water Circulation and Marine Environment in the Antarctic Traced by Speciation of 129I and 127I

Emissions of anthropogenic 129I from human nuclear activities are now detected in the surface water of the Antarctic seas. Surface seawater samples from the Drake Passage, Bellingshausen, Amundsen, and Ross Seas were analyzed for total 129I and 127I, as well as for iodide and iodate of these two isotopes. The variability of 127I and 129I concentrations and their species (127I−/127IO3 −, 129I−/129IO3 −) suggest limited environmental impact where ((1.15–3.15) × 106 atoms/L for 129I concentration and (0.61–1.98) × 10−11 for 129I/127I atomic ratios are the lowest ones compared to the other oceans. The iodine distribution patterns provide useful information on surface water transport and mixing that are vital for better understanding of the Southern Oceans effects on the global climate change. The results indicate multiple spatial interactions between the Antarctic Circumpolar Current (ACC) and Antarctic Peninsula Coastal Current (APCC). These interactions happen in restricted circulation pathways that may partly relate to glacial melting and icebergs transport. Biological activity during the warm season should be one of the key factors controlling the reduction of iodate in the coastal water in the Antarctic.

prevents the direct oxidation of iodide to iodate 13 . However, some biological processes including bacteria, enzyme and plankton activity have been suggested to control the formation of iodide 14 . Therefore, the transformation of iodine species, and in particular the 129 I species, can reflect the change of marine primary production and marine environment 14,15 .
The only stable isotope of iodine is 127 I, and the most long-lived radioisotope (T 1/2 = 15.7 My) is 129 I with a pre-nuclear era natural ratio ( 129 I/ 127 I) of less than 1.5 × 10 −12 in the marine system 16 . Human nuclear activities including nuclear weapons test, nuclear fuel reprocessing and nuclear accidents have released a large amount of 129 I to the Erath's environments, and elevated 129 I level by 1-5 orders of magnitude, up to values ( 129 I/ 127 I) exceeding 10 −7 . Due to uniqueness of 129 I resources and the relatively long residence time of iodine (~300 ky) compared with the water turnover time (~1000 y) in the ocean 17 , 129 I is a useful tracer for investigation of ocean circulation and water mass exchange 14,[18][19][20] . These studies have mainly focused on the Northern Hemisphere, especially the North Atlantic and Arctic, the highest value up to 4 × 10 12 atoms/L for 129 I concentration, and 3 × 10 −6 for 129 I/ 127 I atomic ratio occurred in the English Channel and the North Sea 14 . Only a few data from the Southern Hemisphere (40-75°S) were available 21,22 , where the lowest value down to (1-3) × 10 6 atoms/L for 129 I concentration, and (6-13) × 10 −12 for 129 I/ 127 I atomic ratio were reported in the Antarctic water 23 .
It is generally known that pollutants are transported to the Antarctic through atmosphere dispersion and ocean currents. We have reported the dispersion and pathway of gaseous pollutant to the Antarctic from Northern Hemisphere and lower latitude region in our previous study 23 . The investigation presented here aims to examine levels and distribution of iodine species ( 129 I and 127 I) in the surface seawater in the Drake Passage, Bellingshausen, Amundsen, Marie Byrd and Ross Seas in the Antarctic sector. The information is used to trace the sources and transport pathways of different species of iodine isotopes in the Antarctic water and to contribute for the better understanding of circulation and movement of the water masses and its effects on the marine environment.

Results
Distribution of 127 I and 129 I in the Antarctic surface seawater. The concentrations of 127 I in the Antarctic surface waters (Fig. 1a, Supplementary Table S1) range from 0.20 μmol/L to 0.60 μmol/L, with an average of 0.32 μmol/L. The data show big variation in the concentrations along the sampling area. Relatively high 127 I concentrations (0.35-0.60 μmol/L) occurred in the Drake Passage, central Bellingshausen Sea, central Amundsen Sea and its coastal area, central Marie Byrd and Ross Sea coast. The concentrations of 129 I (Fig. 1b, Supplementary Table S1) in the analyzed seawater range from 1.15 × 10 6 atoms/L to 3.15 × 10 6 atoms/L, with an average of 2.14 × 10 6 atoms/L, which is lower than that in the Northern Hemisphere (>1.0 × 10 7 atoms/L) 24 by a factor of more than 3. As the case with 127 I, high 129 I concentrations ((2.5-3.15) × 10 6 atoms/L) occur in the Drake Passage, eastern Bellingshausen Sea, central Amundsen Sea and its coastal area, eastern Marie Byrd and central Ross Sea. Considerable variability is also observed in the 129 I/ 127 I atomic ratios with values ranging from  (Supplementary Table S1, Supplementary Fig. S1). These values are more than 4 times higher than the pre-nuclear level of 129 I/ 127 I in the marine system (1.5 × 10 −12 ) 16 .

Discussion
The concentrations of 129 I and the 129 I/ 127 I ratios in the Antarctic surfaces seawater show the lowest values compared to those measured in surface waters of other oceans (Fig. 3). The 129 I concentrations in the Antarctic surface seawater are 3-5 orders of magnitude lower than what was found in the surface water of the Nordic Seas and the Arctic Ocean 7,14,20,26 . Sources of the considerably high 129 I in the North Atlantic and the Arctic Oceans and related seas were attributed to dispersion of marine discharges from the nuclear reprocessing plants at La Hague (France) and Sellafield (UK). Even the relatively low 129 I values ((0.6-0.9) × 10 7 atoms/L) measured in seawaters in the Indian Ocean 22 are 2-8 times higher than those in the Antarctic. The only reported values of 129 I concentrations ((0.5-0.9) × 10 6 atoms/L) and 129 I/ 127 I ratio ((0.3-0.6) × 10 −11 ) in the shallow seawater, which are slightly lower than those in Antarctic surface seawater (59-77°S) measured here, were observed in the southern South Pacific Ocean (47-62°S) 21 . The low 129 I concentrations in the Pacific Ocean waters might result from an underestimation of 129 I concentration because organic 129 I was not separated using solvent extraction and excluded in the measured 129 I 27   Irrespective of the 129 I concentration variability in the world's oceans, all the data shown in Fig. 3 indicate values above the natural concentration in the pre-anthropogenic era seawater. Fallout from nuclear weapons tests, discharges from Nuclear fuel reprocessing plants (NFRPs) and nuclear accidents such as at Chernobyl and Fukushima are the major anthropogenic sources of 129 I to the environment. Each of these sources has dispersed in the Earth's surface environment through particular spatial and temporal mode. Dispersion in the atmosphere was the main mode of the nuclear weapons testing derived radioactive releases, while the main pathway of the NFRPs releases is through marine waters. Both of these sources have globally spatial and a few decades temporal extension compared to the restricted regional effects of the nuclear accidents. The NFRPs in the Southern Hemisphere are the ones in Argentina, Brazil and South Africa. Antarctica marine waters lie far from any direct discharges from these NFRPs and the nuclear weapons testing sites. Consequently, it is expected that 129 I in the Antarctica marine waters originates from remote sources and was transported through ocean currents and/or atmospheric dispersion. 129 I released from nuclear weapons testing under seawater may directly enters the seawater, while the atmospheric fallout contributes to 129 I produced by atmospheric weapons testing, including the close-in (tropospheric) fallout for tests conducted in small islands in the Pacific Ocean. Some of the atmospheric nuclear weapons tests conducted at the Pacific Proving Grounds (PPG) mainly in the Marshall Islands in 1946-1962, yielded about 108.5 Mt TNT, corresponding to about 24.7% of the total yield of global nuclear weapons tests 28 . Close-in fallout of the PPG was preferentially deposited in the sea since the tests were conducted on small islands. It has been demonstrated that close-in fallout of radionuclides (e.g. plutonium) from the PPG has been transported to the northwestern Pacific Ocean by the North Equatorial Current and Kuroshio Current 29,30 . There is no evidence, however, indicating that the radionuclides of close-in fallout from the PPG were transported through sea currents to the lower latitude region of the Northern Hemisphere and across the equator to the Southern Hemisphere.
Pointing out a major source of 129 I to the Antarctica marine water has critical used to investigate water circulation and glacier balance. If the major source is through the marine circulation, then it can elucidate where this happening is and how this can be connected to water circulation. Alternatively, if the source is dominated by atmospheric dispersal, then much of iodine can also be trapped in the ice cap and can provide indications of glacier contribution to the marine waters. Transport of the 129 I released from NFRPs which accounts to more than 90% of the 129 I in the present environment, mainly occurs via marine circulation and less restricted dispersion via the atmosphere 6, 26 . These dispersion patterns have resulted in a dramatically elevated 129 I level in the Northern Mt TNT corresponding to about 1.83% of the total yield 28 . It has been estimated that an approximate rate of 0.17 and 0.28 g of 129 I per kiloton TNT equivalent is produced from fission of 235 U and 239 Pu, respectively in a nuclear explosion. It is estimated that the French and British nuclear weapon tests have released a total of 2.1-5.1 kg 129 I, which accounts for about 5% of the total 129 I (~57 kg) 34 released from the global atmospheric nuclear weapons tests. Therefore, the 129 I released from these weapons tests is an important source of 129 I in the Southern Hemisphere seawaters. 129 I in the Antarctic surface seawater should also originate from remote areas through atmospheric dispersion in the stratosphere. Occurrence of tritium, 137 Cs, 36 Cl, 90 Sr, 238, 239, 240, 241 Pu and 241 Am in the ice cores and snow in the Antarctica was attributed to nuclear weapons testing [35][36][37][38] , although fallout levels of these radionuclides are more than 2 orders of magnitude lower than those observed in the Northern Hemisphere.  35,38 . Although no seawater samples were collected at that time for analysis of 129 I, it is expected that like other radionuclides mentioned above, 129 I has also been dispersed through the stratosphere and deposited in the Antarctic. A possible evidence supporting direct influence of atmospheric fallout of the nuclear weapons tests is the 129 I concentrations ((6-8) × 10 6 atoms/L and 3.4 × 10 6 atoms/L) in seawater of the southern Indian Ocean (43-46°S) 22 and in rivers and lakes of New Zealand (35-45°S) 19 respectively. These values are higher than the average concentrations measured in Antarctic surface seawater presented here and also higher than 129 I concentrations ((0.53-0.90) × 10 6 atoms/L) in seawater in the South Atlantic Ocean (45-62°S) 21 . This feature suggests that much of the 129 I in Antarctic surface seawater originates from fallout of atmospheric nuclear weapons tests in the Marshall Islands in 1950s through ocean currents transport along the South Pacific.
Utilization of the isotope concentration to interpret water circulation and environmental impact relies on the characteristics of the system in Antarctic waters (Fig. 4). A major control on the surface water circulation is related to the Antarctic Circumpolar Current (ACC) which is driven by large-scale diagonal compression field moving always eastwards and migrates to the Drake Passage 39 . The ACC interacts with the Antarctic Peninsula Coastal Current (APCC) that generally brings fresher water partly fed by glacial melting. The signatures of this interaction are marked by decrease in 129 I concentrations from >2.6 × 10 6 atoms/L at locations 1, 2 and 3 to 1.5 × 10 6 atoms/L at location 7 in Bellingshausen Sea. This distribution pattern of 129 I in the surface water of the Drake Passage (Fig. 4) starting with the high 129 I value is in agreement with the circulation pattern in the region where the ACC converges southwards along the Drake Passage to form the Antarctic Peninsula Coastal Current (APCC) into the Bellingshausen Sea. The decreased 129 I concentration should be attributed to the dilution of relative high 129 I in the ACC by the low 129 I water in the Antarctic Peninsula coast.
The concentrations of 127 I (0.41 μmol/L) and 129 I (2.95 × 10 6 atoms/L) from location 9 in the eastern Bellingshausen Sea are similar to those from location 2 in the Drake Passage. This similarity indicates that an important branch of ACC that carried the high 127 I and 129 I migrates southward into the Bellingshausen Sea  40 , the branch of ACC at location 9 is combined with the APCC at location 7, forming a pathway that continues westwards along the Antarctic continent, causing a relative highly 129 I level in the APCC waters.
It has been proposed that there are a large number of different scales gyres between the ACC and the APCC (i.e. Antarctic Divergence), which are more complicated and contain numerous discontinuous and closed/ unclosed gyres 41 Table S3, Supplementary Fig. S2).
Relatively high 129 I concentration of about 2.8 × 10 6 atoms/L was measured at locations 40, 41 and 46, suggesting additional two branches of ACC that carried the high 129 I ACC water moving southward across these sampling sites. Relatively low ratios of 127 I − / 127 IO 3 − (0.15-0.68) and 129 I − / 129 IO 3 − (0.58-0.96) were observed in this area compared to those in locations 2 and 3 in the ACC and the APCC. This feature should result from the oxidation of iodide during its transport to this location before emerging with the APCC. 129 I concentration at location 55 (2.41 × 10 6 atoms/L) is nearly 2 times higher than that at locations 47-54 (1.34 × 10 6 atoms/L) and the iodine at this site is mainly iodide (4.85 for 127 I − / 127 IO 3 − and 1.00 for 129 I − / 129 IO 3

−
). This trend indicates that there is another branch of APCC moving northward through location 55 (167.01°W) towards the ACC. The southward moving ACC branch through location 46 forms a closed gyre (Ross Gyre) with the northward moving branch of the APCC through location 55 (167.01°W). The Ross Gyre was actually one of the first discovered subpolar cyclonic vorticity 43,44 .
Relatively high concentrations of 129 I ((2.57-2.61) × 10 6 atoms/L) were also observed at locations 64 and 65, indicating the APCC received the high 129 I seawater from the ACC branches reached to these locations in west bottom of the Ross Sea. This transport pathway is confirmed by the high 129 I − / 129 IO 3 − ratio (1.72) at location 65, which agrees with the observed 129 I species and the high biological activity in the APCC.
This investigation revealed that the eastward flowing ACC has 6 southward moving branches at the 67.96°W, 75.71°W, 113.19°W, 126.45°W, 143.1°W and 173.7°W, respectively. This non-zonal behavior might be mainly attributed to the complicated submarine topography 45 . Meanwhile, it also confirmed here that the ACC moves eastward along the Pacific-Antarctic ridge to the Drake Passage in the South Pacific and then split into three branches. One branch moves away from the Drake Passage northwards, the second moves eastwards and the third one moves toward the south along the Antarctic Peninsula coast driven by the polar easterlies, and then moves westwards along the Antarctic continent, forming the APCC. Four branches of the APCC move north toward to the ACC. The eastwards moving ACC and westwards moving APCC interact with each other through their branches in opposite directions, forming numerous discontinuous and closed and/or unclosed gyres in-between them.

Materials and Methods
Samples and chemicals. Sixty-four surface seawater samples were collected in the Drake Passage, Bellingshausen, Amundsen, Marie Byrd and Ross Seas during cruise in the Antarctic (Fig. 5, Supplementary  Table S3). The surface seawater was pumped through the built-in seawater sampler in the research vessel N.B. Palmer from 2 m under the sea surface directly into 2 L polyethylene bottles, which were sealed and shipped to Xi'an, China for analysis. Meanwhile, temperature, salinity and chlorophyll concentrations and partial pressure of CO 2 (pCO 2 ) were measured by on-line detecting system in the research vessel.
The 129 I standard solution (NIST-SRM-4949c, National Institute Standard and Technology Gaithersburg, MD) and 127 I carrier solution (Woodward iodine, Woodward Iodine Corporation, Oklahoma, U.S.A.) prepared by dissolution of iodine crystal into 0.40 mol/L NaOH-0.05 mol/L NaHSO 3 solution were used. All chemical reagents used were of analytical grade, and all solutions were prepared using deionized water (18.2 MΩ·cm).
Separation of total iodine and its species (iodide and iodate) from seawater. After removal of potential suspended particles in the seawater by filtration through a 0.45 μm membrane, 0.60 L and 1.20 L of seawater samples were taken for separation of total iodine and iodine species, respectively. A modified procedure from our previous method was used for separation of total inorganic iodine, iodide and iodate using AgI-AgCl coprecipitation from seawater 46,47 . In brief, 0.60 L seawater was transferred to a beaker. 0.5 kBq of 125 IO 3 − tracer was spiked, 0.20 mg of 127 I carrier and 0.50 ml of 2.0 mol/L NaHSO 3 solution were added into the beaker, and then 3 mol/L HNO 3 was added to adjust pH 1-2 to convert all iodine species to iodide. 30 mg Ag + (28 ml of 0.01 mol/L AgNO 3 solution) was dropwise added to the solution under stirring to form AgI-AgCl-Ag 2 SO 3 -AgBr coprecipitate. The precipitate was separated by centrifuge, and then sequentially washed with 3 mol/L HNO 3 , H 2 O, 30% and 20% NH 4 OH to remove Ag 2 SO 3 and most of AgCl and AgBr until 1-3 mg of precipitate was obtained. 1.20 L seawater was transferred to a beaker for separation of iodide in seawater. 0.5 kBq of 125 I − tracer and 0.2 mg of 127 I − carrier (KI, 129 I/ 127 I atomic ratio <2.0 ± 0.3 × 10 −13 carrier) were spiked, NaHSO 3 was added into the sample to a final concentration of 0.30 mmol/L, and then 0.5 mol/L HNO 3 was slowly added under stirring to adjust pH 4.2-5.5 (measured using a pH meter). 150 mg Ag + (45 ml of 0.03 mol/L AgNO 3 ) was dropwise added to the solution to form AgI-AgCl-Ag 2 SO 3 -AgBr coprecipitate. The precipitate was separated by centrifuge and the supernatant was used for separation of iodate. The separated precipitate was sequentially washed with HNO 3 , H 2 O and NH 4 OH until 1-3 mg of precipitate were obtained. To the supernatant, 0.5 kBq 125 IO 3 − tracer was spiked for separation of iodate, 0.2 mg of 127 I carrier, 0.5 ml of 2.0 mol/L NaHSO 3 solution were added, and then 3.0 mol/L HNO 3 was added to adjust pH 1-2 to convert all iodine species to iodide. Then follow the procedure for total inorganic iodine to separate iodate. 125 I in the precipitate was measured using a NaI gamma detector (Model FJ-2021, Xi'an Nuclear Instrument factory, Xi'an, China) for monitoring of the chemical yield of iodine in the procedure. The recovery of iodine and its species in the entire procedure is higher than 80%. The schematic diagram of the analytical procedure is shown in Supplementary Figure S3.
The procedure blanks were prepared using the same procedures as for separation of total iodine, iodide and iodate in seawater but no samples were added. 129 I/ 127 I standards containing 1.0 mg iodine in AgI−AgCl form were prepared for calibration of the AMS measurement. See the detailed method in the Supporting Information. Iodine in the commercial 125 I tracer exists as iodide (NaI). To synthesize 125 IO 3 − tracer, NaClO was added into 125 I − solution, then HCl was added to adjust pH 1-2 to oxidize iodide to iodate. The detailed method is presented in the Supporting Information.
Measurement of 129 I using AMS and 127 I using ICP-MS. The prepared AgI-AgCl coprecipitate was dried in an oven at 60-70 °C for 3-6 h, the dried precipitate was ground to fine powder and mixed with five times by mass of niobium powder (325 mesh, Alfa Aesar, Ward Hill, MA), which was finally pressed into a copper holder using a pneumatic press (Zhenjiang Aode Presser Instruments Ltd.). 129 I/ 127 I atomic ratios in the prepared targets were measured by AMS using 3MV Tandem AMS system (HVEE) in the Xi'an AMS center. All samples, blanks and standards were measured for 6 cycles and 5 minutes per sample in each cycle. A detailed description of AMS system and measurement of 129 I has been reported elsewhere 48 .
1.0 mL solution of the iodide fraction and the iodate fraction separated using anion exchange chromatography (See the detailed method in the Supporting Information.) and the original seawater were taken to a vial, and the samples were diluted for 10 times using 1% NH 4 OH solution, Cs+ solution was spiked to a concentration of 2 ng/mL. 127 I in the prepared samples was measured using ICP-MS (X-series II, Thermo Scientific, USA). The SCIENTIfIC REPORTS | 7: 7726 | DOI:10.1038/s41598-017-07765-w detection limit of 0.02 ng/mL for 127 I was obtained. Iodide concentration was corrected for chemical yield during chromatographic separation. Data availability statement. All data analyzed during this study are included in this published article and its Supplementary Information files.