Document JKK5DmMDZLB16JKMVDKy2kGe

F - Pc p thick, and oriented parallel to (010) wan pre pared from the olivine crystal (composition de termined by microprobe analysis). Examination of the sample by optical microscopy before the experiment revealed no inclusions or hetero geneities other than subboundaries, in agree ment with the observations of S. H. Kirby and M. W. Wegner [Trans. Am. Geophys. Union 54. 452 (1973)1 and M. W. Wegner and i. M. Chris tie [Contrih, Minctol. Petrol. 43, 195 (|97{)) on olivine from the same locality. The disk was en cased in a stainless steel container which includ ed 'momentum trap" plates [D. G. Doran and R. K. Unde. Solid State Phys. 19. 262 (1966)]. and the container was impacted with a 16-mm. tungsten flyer-plnle launched from a high-per formance propellant gun. 9. We calculated the peak pressure, using standard impedance match techniques for multiple shock reverberations |J. M. Walsh and K. H. Chris tian. Phvs. Rev. 97, 1544 (1955): L. V. Al tshuler. Sow Phys. Usp. 8. 59 (1965): R. V. Gibbons and T. J. Ahrens, J. Geophys. Res. 76, 5489(1971)} based on the Hugoniot equation-ofstate data (//) for tungsten and stainless steel, and it is accurate to within about 5 percent. The lime at peak pressure is based on a calculation of Ihc type given by G. R. Fowles (J. Appl. Phys. 31.65$ (1960)| and by AITshuler (see above). 10. Carried out with the l-Mev electron microscope at the U.S. .Steel Research Laboratory, Monroe- vilie, Pa., and at 200 kv at the U.S. Geological Survey. Rcston. Vn. 11. R. (). McQueen, S. P. Marsh. J. W. Taylor, J. N. Fritz, W. J, Carter, in High Velocity Impact Phenomena, R. Kinslow. Ed. (Academic Press, New York. 1970), p. 294. 12. We made the simplifying assumption that the temperature in the sample which has attained peak pressure by multiple reverberations is ap proximately the same as the temperature of the container achieved by a single shuck to peak pressure. 13. T. J. Ahrens, F. D. Tsay, D. H. Live, Geochim. Cosntochim. Ada (Stippl. 7) (1976), p. 1143. This is a maximum estimate since it is for the attainment of peak picssure by a single shock [see L. V. ATtshnlcr (9)|. 14. T. J. Ahrens and J. D. O'Keefe. Moon 4, 214 (1973). 15. 1). R. Waldbmim, Nature (l.ondon) 232, 545 (1971). 16. N. 1.. Bowen and J. F. Schairer, Am. J. Sci. 29, Ifil (1935). 17. E, C\ T. Chao, in Researches in Geochemistry, P. AbcJson, Ed. (Wiley. New York. 1967). vol. 2. p. 204. Note that thctomorphic and diaplcctic arc equivalent terms. 18. T. J. Ahrens. J. H. Lower, P. L. Lagos. J. Geonhys. Res. 76, 518 (1971); T. I. Ahrens and C. V. Petersen, in The Application of Modern Physics to the Earth and Planetary Interiors. S. K. Runcorn. Ed. (Wiley-Inierscience, New York. 1969). p. 449. The Hugoniot data Rre con sistent with results from static, high-pressure experiments (L. Liu, Nature (h>ndon) 262, 770 (1976); Geophys. Res. Lett. 2, 417 (1975); Na ture 258, 510 (1975)J which indicate m^jor, reconstructive transformations at these pressures in magnesian olivine, for example, to (Mg.Fc)SK)3 pcrovskiie + (Mg.Fe)O rock salt phases jsee R. Jcanloz and T. J. Ahrens, in High Pressure Research: Applications to Geophysics, M. Mnnghnani and S. Akimoto. Eds. (Academic Press, New York, 1977), p. 439]. Direct evi- dcncc for similar major, reconstructive transfor mations and reactions in ncsositicates under shock can be found in L. Liu [Earth Planet. Sci. Lett. 26, 425 (1975)). T. J. Ahrens and E. K. Oraham Uhid. 14. 87 (1972)], and H. Schneider and IJ. Hornemann [Naturwissenschaflen 62, 296 (1975)1. 19. See B. ). skinner and J. J. Fahey, J. Geophys. Res. 68, 5595 (1963). and references therein. 20. L. Liu and A. B. Ringwood, Earth Planet. Sci. Lett. 28, 209(1975); A. E. Ringwood and A. Ma jor. ibid, 12, 411 (1971). In these experiments, CuSiOj perovskite formed (and identified) at high pressure could only be quenched to a glass. 21. The free energy of a crystal is increased by an amount of the order of 1/2 uh3 per atomic length of dislocation (where u. is (he shear modulus and h is the average length of the Burgers vectors of the dislocations). Taking the dislocation density to be about 10'* cm 1 or more in the tangles, h -- 5 A, p. 80 x 10* pascals, and assuming thnt a typical difference in free energy between silicate crystals and their glasses is about 5 kilo joules per mole, it is possible for the glass to have a lower free energy than the dislocated crystal. Sec also R. M. J. Cotterill, E. J. Jensen. W. Damgaard Kristensen, and R. Paetsch [7 Phys. (Pans} 36, C2 (1975)]. 22. D. F. Grady, in High-Pressure Research: Appli cations to Geophysics, M. Manghnani and S. Akimoto, Eds. (Academic Press, New York, 1977). p. 389;______ W. J. Mum, P. S. DeCarli, J. Geophys. Res. 80. 4857 (1975); D. E. Grady and W. J. Mum, Geophys. Res. Lett. 3, 472 (1976). 23. R. M. J. Cotterill. Phys. Lett. 60A, 61 (1977). 24. We are grateful Tor discussions wi(h I. Jackson and L. Liu, and for review, from A. Albee and D. Grady, all of which were very helpful. This research was supported by NASA giant NGL 05-002-10} and contract NAS9-I4749(U.S. Steel Research Laboratory). Contribution No. 2889, Division of Geological and Planetary Sciences, California Institute of Technology. 24 March 1977; revised 25 April 1977 Polychlorinated Biphenyls: Penetration into the Deep Ocean by Zooplankton Fecal Pellet Transport Abstract. High concentrations ofpolychlorinated biphenyls (PC8's) were found in fecal pellets from natural populations of euphattsiids collected in the Ligurian Sea. Since biogenic particulate products, especiallyfecal pellets, are known to sink rapid ly and intact to the ocean bottom, the transport of PCD's by such sinking particles could be an important mechanism which contributes to the penetration of PCB's into , the deep sea. The rate of vertical mixing in the ocean (!) is too slow to account for the quantities of polychlorinated biphenyls (PCB's) found in Atlantic and Mediterra nean abyssal sediments (2) if one as sumes that they penetrate into the deep ocean only in the dissolved state. An al ternative explanation is that PCB's arc carried to the sediments by rapidly sink ing particles. Several investigators have hypothesized that sinking biogenic mate rial such as zooplankton fecal pellets and molts may accelerate the downward ver tical transport of certain metals and ra dionuclides (.?). Analytical work has shown that both euphausiid fecal pellets and molts contain significant quantities of trace elements (4, 5) and radionuclides (6); therefore, the sinking of these partic ulate products, especially fecal pellets, is strongly implicated in the removal of many such elements from surface wa ters. Experimentally determined sinking rates of several hundred meters per day (7-9) for zooplankton fecal pellets and molts are compatible with the hypothesis that these products could act as effective conveyors of surface-introduced pollu tants into the deep sea. Recent field stud ies (9, 10) have shown conclusively that zooplankton fecal pellets not only pene trate to great depth (2000 to 4000 m) in tact but also in large numbers--650 pel lets per square meter per day. We have found that freshly released euphausiid fecal pellets collected from natural popu lations contain relatively high concentra tions of PCB's. and we propose that such biogenic particles make a significant con tribution to the vertical transport of PCB's in the ocean. The euphausiid Meganyctiphanes norvegica was examined since it is an abun dant member of the zooplankton com munity in the western Mediterranean and North Atlantic and sufficient biologi cal data on the production rate of partic ulate products have been compiled for this species (4, II). During 1974-1975 M. nonrgica and the microplankton which serve as its food were collected at a station 5 km off Villefranche-sur-Mer. France. We fished for euphausiids and microplankton at night with an IsaacsKidd mid-water trawl and a I -m plankton net (76-jxm mesh aperture), respectively, by making several short oblique tows be tween 100 m and the surface. All micro plankton samples were carefully exam ined and found to be free from adhering tar, oil, and paint chips. Glass and metal implements cleaned in advance were used to sort euphausiids from other species in order to avoid con taminating the samples. Euphausiid fe cal pellets and molts were collected on board and later in the laboratory by methods described elsewhere (it. 12) with the exception that all containers were either glass or metal. Although ex treme care was taken to avoid unneces sary contact between the samples and plastic materials, some contact inevita bly occurred between the organisms and the nylon plankton nets. Cross-con tamination of PCB's between plankton and nylon nets can occur (/.?); however, extraction of the nets used in our study indicated that PCB contamination from this material was negligible. Samples were analyzed by standard procedures (14) with modifications to adapt to small sample quantities. All samples were frozen, lyophilized, and pulverized in preparation for extraction. The relatively abundant samples, such as microplankton (0.1 (o 1.5 g dry) and eu phausiids (1 to 8 g dry), were subjected , 29 JULY 1977 - 459 OSW 032144 STLCOPCB4016106 lo extract!on with hexane in a Soxhlet extractor (8 hours at four cycles per hour). We batch-extracted the less abun dant samples, fecal pellets (15 to 90 mg dry) and molts (50 to 150 mg dry), by shaking with hexane in settled, conicalbottomed, centrifuge tubes. We removed the extracts by first centrifuging to settle the sample residue and then transferring the supernatant solvent with the aid of a Pasteur pipette, liach sample was ex tracted ten times, Interfering substances were eliminated from the hexane ex tracts by chromatography on Florisil and by treatment w'ilh concentrated H2S04. It was not necessary to remove fatty sub stances by acctonitrilc-hcxanc partition ing. Hexitne extracts were then further concentrated to it convenient volume (usually 0.2 to 0.5 ml) in a Kuderna-Danish concentrator and analyzed by gas chromatography. Precautions were em ployed to ensure that contamination from chemicals, apparatus, and glass ware was negligible (15). During the No vember 1974 collection, a surface water sample was also collected and analyzed for PCU's according to the method of Harvey and Slcinhauer (16). The PCB concentration (DP-5) was 2.5 ng/liter, a value that is considered representative for these waters (17). Concentrations of PCB's in euphausiid bodies, molts (18), and fecal pellets as well its in the microplankton upon which they feed are given in Table I. Fecal pel lets contained the highest concentrations of PCB's in all samples examined. The PCB concentration in feces (wet weight) from the November 1974 collection was 1.5 x I0'1 greater than that present in the surrounding water. Furthermore, the PCB concentrations in feces (dry weight) ranged from 3.5 to 21 times higher than those in the food organisms which formed the fcccs. A similar concen trating process has been found to occur for several trace elements and radio nuclides in the same species (4-6). Recently, estimates have been made of PCB deposition rates into sediments in the northwestern Mediterranean re gion (19). These estimates are based on sediment concentrations in a scries of cores taken at a depth of 400 m off the coast of Nice, France. Computed values range from approximately 80 to 125 /xgof PCB (DP-5) per square meter per year over a period of 15 to 20 years. It is conceivable that PCB's could be transported to sediments by way of fecal pellets, molts, eggs, and carcasses re leased from an overlying euphausiid pop ulation. Molting, egg laying, and death (carcass production) in cuphausiids are 460 Table I. Polychlorinated biphenyls (IS) in cu phausiids (Meganyctiphanes norvcgica), their particulate products, and microplankton which serve as the euphausiids1 food. Water sampled during the November 1974 collection contained 2.5 ng of PCB per liter. Ratio of wet DP-5 Sample weight to (Mg/kg dry weight (dry weight)] -- November 1974 Wlude animal 4.7 620 Molts 4.6 1,400 Fecal pellets 4.4 16,000 Microplankton 10.7 4,500 January 1975 Whole animalt 260, 290 Molts Fecal pellets 170 4.800 Mareh 1975 Whole animal 38 Molts Not detectable Fecal pelletst Microplankton + 11,000; 38,000 1.800 `Principally copepods, phytoplankton, .and detri tus. tTwo separate samples. discontinuous processes compared to defecation. Hence, the production rates for these products are fnuch lower than that for fecal pellets (4, 8). This lower production rate and the fact that PCB concentrations in molts and whole eu phausiids appear to be much lower than corresponding concentrations in feces indicate that defecation should be far more effective in removing PCB's from surface waters than the discontinuous processes. Accordingly, we have at tempted to calculate roughly how much of the total vertical flux of PCB's into sediments in this area could be attributed to sinking fecal pellets from an overlying population of M. norvegica. Production rates of fecal pellets for this euphausiid have been found to range between ap proximately 1.8 and 5.1 percent of the organism's dry body weight per day, de pending on the available food supply (4). From the data of Franqueville (20). we have computed that the annual average biomass of M. norvegica in the Ligurian Sea is approximately 0.17 mg (dry weight) per cubic meter. Coupling these data with the mean fecal pellet PCB con centration [= 17,000/xg/kg (dry weight)] for the three sampling periods shown in Table I leads to values for fecal-released PCB's ranging from 5.2 x 10-5 to 15 x I0~5 /xg m-J day-1. If these data are in tegrated over the photic zone (top 75 m) where the majority of the fecal pellets are released during the time the euphau siids are actively feeding, we obtain de livery rates of PCB's to sediments of 1.4 to 4.1 pg m~2 year-1, which are between one and two orders of magnitude lower than estimates based on deposition by all routes. However, measurements of zoo plankton biomass (20, 21) made in the same region indicate that M. norvegica comprises only l to 5 percent of the total zooplankton biomass. If PCB concentra tions in feces and fecal pellet production rates measured in M. norvegica are typi cal of those for other zooplankton spe cies (notably copepods, which probably form the bulk of the zooplankton bio mass), the delivery rates arising from zooplankton defecation will approach those based on PCB concentrations in the sediments. Our calculations are ad mittedly crude because of the limited data available and are subject to many assumptions that may not be entirely val id; nevertheless, the dose similarity be tween PCB dux rates derived from fecal pellet data and those based on sediment values suggests that sinking zooplankton fecal pellets contribute significantly to the downward vertical transport of sur face-introduced PCB compounds. D. L. Flour Sc o t i W. Fowli.a International Laboratory of Marine Radioactivity, Musee Oceanogruphique, Principaute dc Monaco References and Notes L W. S. Broecker, Chemical Oceanography (H;ir* court Brace Jovanovich, New York, 1974), n. 59. 2. 0. R. Harvey and W. G. Steinhuuer, in Environ mental Biogeochemistry, ). O. Nriagu, Ed. (Ann Arbor Science Publishers, Ann Arbor, Mich., 1976), vol. I, p. 203; D. L. Elder. 3. P. Villeneuve, P. Parsi. G. R, Harvey, in ArhV/nVj of the International Laboratory of Marine Radio activity, 1976 Report (International Atomic En ergy Agency, Vienna, 1976), p. 136. 3. C. Osterbero. A. G. Carey, H. Curl. Nature {London) 200, 1276 (1963); F. G. Lowman, T. R. Rice. F. A. Richards, in Radioactivity in the Marine Environment (National Academy of Sci ences. Washington, D.C., 197)), p. 161: ). H. Marlin, Limnol. Oceanogr. 15. 756 (1970); S. Krishnaswami and M. M. Sarin. Earth Planet. Set. Lett. 32, 430 (1976); S. Krishnaswami. D. Lai, B. L. Somayajulu, R. F. Weiss. H. Craig, ibidp. 420; M. P. Bacon. D. W. Spencer. P. G. Brewer, ibid., p. 277; Y. Nozaki, J. Thompson. K. K. Turckian, ibid., p. 304. 4. L, F. Small, S. W. Fowler, S. Keckcs, \n Radio active Contamination of the Marine Environ ment (International Atomic Energy Agency. Vienna. 1973). n. 437. 5. G. Benayoun, S. W. Fowler, B. Orcgioni, Mar. Biol. 27, 205 (1974); S. W. Fowler and B. Oregioni, in Activities of the International Labora tory of Marine Radioactivity. 1974 Report (In ternational Atomic Energy Agency, Vienna, 1974), p. 55; S. W. Fowler and G, Bermyoun, Mar. Sci. Commun. 2, 43 (1976). 6. R. D. Cherry. S. W. Fowler, T. M. Beasley, M. Heyraud. Mar. Chem. 3. 105 (1975); M. Heyraud, S. W. Fowler. T. M. Beasley, R. D. Cherry, Mar. Biol. 34. 127 (1976); J. J. W. Higgo, K. D. Cherry, M. Heyraud, S. W, Fowler. Nature {London) 266, 623 (1977). 7. T. J. Smayda. Mar. Geoi. 11. 105 (197!); Lim nol. Oceanogr. 14, 621 (1969); S. W. Fowler and L. F. Small, ibid. 17, 293 (1972). 8. L. F. Small and S. W. Fowler. Mar. Biol. 18. 284 (1973). 9. P. H. Wiebe. S. H. Boyd, C. Winget, J. Mar. Res. 34, 34| (1976). 10. H.-J. Schrader, Science 174, 55 < 1971). 11. S. W. Fowler, G. Benayoun. L. F, Small, Tha- lassia Jtigosi. 7, 35 (|97|); S. W. Fowler, L. F. Small, S. Keckes, Mar. Biol. II, 45 (1971). 12. J. La Rosa, Deep-Sea Res. 23, 995 (1976). SCIENCE, VOL. 197 DSW 032145 STLCOPCB4016107