Vol49_2_2006 705 ANNALS OF GEOPHYSICS, VOL. 49, N. 2/3, April/June 2006 Key words Benthic Boundary Layer – deep-sea- floor – ocean temperature – helium – ocean radioac- tivity 1. Introduction The Benthic Boundary Layer (BBL) is the dynamic interface between the lithosphere (sea- floor) and the ocean (seawater) where many physical, geochemical and biological processes play an important role in environmental global changes (Boudreau and Jorgensen, 2000). Geo- hazards, carbon cycle, heat flow, life generation, climatic oceanography, are only some examples of local or global processes whose understand- ing is today limited due to the lack of data relat- ed to the deep ocean floors (Thiel et al., 1994). Lithospheric processes at the BBL impact the marine environment at different temporal and spatial scales. Earthquakes produce short-term effects (landslides and tsunamis) that threaten the lives and economy of coastal communities, while the emission of greenhouse gases and mineral-rich fluids impact long-term global cli- mates and the formation of economically impor- tant mineral resources. The BBL has an impor- tant role on carbon cycle being either a potential sinking or transition zone for carbon coming from shallower zones of from the lithosphere. The Benthic Boundary Layer: geochemical and oceanographic data from the GEOSTAR-2 Observatory Giuseppe Etiope (1), Paolo Favali (1), Jean-Luc Fuda (2), Francesco Italiano (3), Matthias Laubenstein (4), Claude Millot (5) and Wolfango Plastino (1) (6) (1) Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Roma 2, Roma, Italy (2) Centre d’Océanologie de Marseille (COM), CNRS, Marseille, France (3) Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Palermo, Italy (4) Istituto Nazionale di Fisica Nucleare (INFN), Laboratori Nazionali del Gran Sasso, Assergi (AQ), Italy (5) Laboratoire de Oceanographie et de Geochimie (LOB), La Seyne-sur-Mer, France (6) Dipartimento di Fisica, Università degli Studi Roma Tre, Roma, Italy Abstract Geochemical and oceanographic data, acquired throughout 6 months by the GEOSTAR-2 benthic observatory in Southern Tyrrhenian Sea, disclosed ocean-lithosphere interactions in the 1900-m deep Benthic Boundary Lay- er (BBL), distinguishing two water masses with different origin and, possibly, benthic residence time. Gas con- centration, helium isotopic ratios, radioactivity, temperature, salinity and vertical component of the current con- verged towards the indication of a BBL characterised by a colder and fresher Western Water (WW), episodical- ly displaced by the cascading of the warmer and saltier Eastern Overflow Water (EOW). The benthic WW has a higher concentration of geochemical tracers diffusing from the seafloor sediments. The data set shows the po- tential of long-term, continuous and multiparametric monitoring in providing unique information which cannot be acquired by traditional, short-term or single-sensor investigations. Mailing address: Dr. Giuseppe Etiope, Istituto Nazionale di Geofisica e Vulcanologia, Sezione di Roma 2, Via di Vigna Murata 605, 00143 Roma, Italy; e-mail: etiope@ingv.it 706 Giuseppe Etiope, Paolo Favali, Jean-Luc Fuda, Francesco Italiano, Matthias Laubenstein, Claude Millot and Wolfango Plastino Oil platforms are sited on the BBL, and cables and pipelines are lain across it. Delicate marine ecosystems have evolved in this extreme envi- ronment, yet it may be significantly affected by man’s activities. The BBL dynamics however is not completely clear: it is not known how it is perturbed by energy and mass release from the seabed (seismicity, heat flow, gas diffusion, bio- logical productivity, diagenetic reactions), what distances the particles and bottom waves can be transported, what is the evolution and fate of bottom currents and storms. The deep-sea observation and monitoring systems developed in the last decade, in the framework of technological and scientific proj- ects in Japan, US, and Europe (Beranzoli et al., 2002, and references therein) can contribute to cover such a scientific gap. GEOSTAR-2 (GEophysical and Oceano- graphic STation for Abyssal Research) is the first European deep-sea observatory for geophysical and environmental monitoring at the seabed be- coming operative in 2000. It was deployed in September 2000 from the Italian R/V Urania, in the Southern Tyrrhenian Sea, between the is- lands of Sicily and Ustica, at about 1900 m b.s.l. (fig. 1). This area was chosen as a key site for Tyrrhenian seismicity and oceanographic studies (Beranzoli et al., 1998). The sensors used for this mission (2 magnetometers, a gravitymeter, an hydrophone, a Doppler currentmeter, a single- point currentmeter, an automatic water sampler for laboratory geochemical analysis, a CTD and a transmissometer) were continuously controlled and managed by a data acquisition and control system able to transmit the data via surface buoy and radio or satellite link to on-shore operators. After 206 days the observatory was recovered in April 2001. More than 4100 h of data were recorded continuously. This mission represented the longest experiment using a complex module, with an intelligent unit, deployed at great depth. This paper presents geochemical and oceano- graphic data obtained from the automatic water sampler, the CTD and the Acoustic Doppler Cur- rent Profiler (ADCP) to show how multidiscipli- nary and integrated monitoring at deep BBLs may provide various original and unique infor- mation, which cannot be acquired by traditional, short-term or single-sensor recordings. Specifi- cally, temperature, salinity, light transmission, currents, radionuclides and gases, including heli- um isotopes, have been examined to trace envi- ronmental changes in the benthic boundary sea- water. 2. Monitoring and data acquisition Geochemical data refer to laboratory analy- ses carried out on 23 water samples collected by the GEOSTAR automatic sampler (RAS-500 McLane) from 26 September 2000 to 9 April 2001. The sampler, using aluminium bags and a pressure compensation system, was programmed for sampling every 10 days over 240 days. Four samples were duplicated and the sequence is composed of 19 time-series data (one sample was lost due to bag rupture). The sampling port was positioned at ∼1 m from the seabed. The seawater samples have been analysed for the following parameters: – gas concentration: CO2, CH4, He, Ne; – helium isotopic composition: R/Ra (3He/ 4He sample/3He/4He atm);Fig. 1. Location of the GEOSTAR-2 deployment site. 707 The Benthic Boundary Layer: geochemical and oceanographic data from the GEOSTAR-2 Observatory – radioactivity (226Ra, 235U, 234mPa, 228Ra, 228Th, 137Cs and 60Co). CO2 and CH4 were analyzed by gas chro- matography with micro-TCD (thermal conduc- tivity detector; Etiope, 1997). Helium analysis was carried out with a Perkin Elmer 8500 gas- chromatograph (Flame Ionization Detector and Hot Wire Detector; 5 ppmv as detection limit; analytical errors of ± 5%). The 3He/4He isotopic ratio, was determined by a static mass spectrom- eter (VG5400TFT, VG Isotopes; typical uncer- tainties are about ± 1% for 3He/4He ratios in the range of atmospheric values; below ± 0.1% for high-3He samples and below ± 3% for low-3He radiogenic samples). The analysis of radionuclides was carried out by gamma spectrometry with coaxial High Puri- ty Germanium (HPGe) detectors having volumes ranging from 200 to 500 cm3 and a total back- ground rate in the energy range [(60÷2700) keV] varying from (221±2) to (980 ±10) counts/days depending on the detector (Arpesella, 1996). Each seawater sample was measured for about ten days using a polystyrene box of 70 mm di- ameter and 30 mm height (Plastino et al., 2003). Temperature and salinity were recorded by a high accuracy and stability CTD (SBE-16 SEACAT, Sea-Bird Electronics, Inc.; nominal accuracy-stability/month of 5×10−3−2×10−4°C/ 5×10−4−3×10−4 S/m, equipped with a pump and a quartz pressure sensor, resolution 0.01 ppm, repeatability 0.005% FS) positioned at ∼1 m from the seabed. Calibrations of temperature and conductivity sensors performed by the manufacturer before and after the mission indi- cate no significant drifts and guarantee the val- ues accuracy. Seawater turbidity was monitored by a 0.25-m-pathlength transmissometer, Al- phatracka Mk II (Chelsea Instruments Ltd.). The sampling rate was one sample per hour. Current magnitude and direction were mon- itored by a 300 kHz Acoustic Doppler Current Profiler. It provides ∼ 120-m 3D current pro- files, representative of water layers (cells) ∼ 3- m thick located at increasing vertical distances from the ADCP. The ADCP was configured for providing hourly profiles resulting from the av- erage of 100 pings evenly transmitted during 10 min. The first cell was ∼ 7 m above the sea- floor. 3. Results 3.1. Pressure, temperature, salinity and light transmission The pressure sensor was confirmed to be a highly performing tide gauge, being able to sense sea level variations of a few mm at ∼1900 m. In addition to the expected semi-diurnal and diurnal tidal signals, long term variations, as well as what can probably be interpreted as high frequency signals non-correctly resolved (the 1-h sampling was too large in this respect), can be evidenced (fig. 2a). The most remarkable characteristics regard- ing the temperature and salinity records (fig. 2b,c) resides in the almost regular occurrence, roughly every 2-3 weeks, of sharp peaks deviating from the background (T: 13.05°C, S: 38.51 psu) with values up to 13.45 °C and 38.63 psu, respectively. On the basis of T-S peak height, duration and den- sity variations (fig. 2d) we recognise seven major events occurring on 24 October 2000, 14 and 30 November 2000, 9 and 25 January 2001, 10 and 26 February 2001. These events appear perfectly coherent with the stratification reported in the re- gion 10 years ago (Sparnocchia et al., 1999) and characterized, below 1500 m, by T-S gradients of the order of 0.1°C/100 m and 0.02 psu/100 m. These T-S peaks suggest a rapid (hours/days) low- ering of the interface separating the relatively warm and saline waters of eastern origin (the so- called Eastern Overflow Water, EOW), cascading from the Channel of Sicily, from the underlying waters of western origin (Fuda et al., 2002), sam- pled most of the time by GEOSTAR. Sporadic light transmission drops represented by a unique hourly data point were recorded, with a large part of spike-like events; they are expect- ed to be due to isolated large-size particles (fig. 2e). A dramatic few-hour drop down to 3.7 V (i.e. a relative light transmission loss of ∼15 %) occurred on October 24, 2000, following several weaker events during the preceding days and co- inciding with the first major T-S peak. Other events coincide mainly with T-S peaks of 9 and 25 January and 10 February. However, while most of the T-S peaks can be associated with transmissometer peaks, many transmissometer peaks are not associated with significant T-S ones. 708 Giuseppe Etiope, Paolo Favali, Jean-Luc Fuda, Francesco Italiano, Matthias Laubenstein, Claude Millot and Wolfango Plastino Fig. 2a-e. CTD data: a) pressure; b) potential water temperature; c) salinity; d) density; e) light transmission. The major 7 events are numbered over the salinity curve. a b c d e 3.2. Currents The ADCP provided different information on the deep-water circulation and on its evolu- tion throughout 6 months. The horizontal speed (not shown) displays peaks that are very similar to the T-S ones. For the purpose of this paper, we mention only that significant vertical com- ponents of current were recognised associated with the T-S peaks. A noticeable anomaly is the occurrence of permanent downward velocities of the order of a few mm/s at all depths. Even though relatively weak, this background resid- ual is a priori abnormal as vertical velocities (either upward or downward) are not expected near the seafloor, except during specific transi- 709 The Benthic Boundary Layer: geochemical and oceanographic data from the GEOSTAR-2 Observatory Fig. 3. CO2 and CH4 variation. Vertical bars are the seven main T/S events marked in fig. 2a-e. ASW is the Air Saturated Water (equilibrium with the atmosphere calculated at seafloor conditions). Checks are the samples where no preservative was used and therefore all CH4 was consumed by biodegradation. tory processes, and except if there is a marked slope of the bottom, or up to a few meters above the seafloor where the observatory was demon- strated to deflect the flow and to induce upward motions (Fuda et al., unpublished data). The observed vertical velocities depend on both the current direction and the current speed. For a given current direction, the higher the hor- izontal speed, the higher the vertical velocity. 3.3. Gases and radionuclides Figure 3 shows the pattern of dissolved CH4 and CO2 throughout the mission. Their concen- trations are coherently close to the theoretical atmospheric equilibrium level (ASW, Air Satu- rated Water) at the seafloor conditions, suggest- ing that no significant external gas sources ex- ist. All bottles, except two (as control samples), were initially conditioned by adding 0.1% Hg- Cl2 to prevent bacterial consumption of CH4. As expected, the two non-conditioned samples dis- played no CH4 concentration (or it was below the detection limit). Figure 4 shows the varia- tion of helium concentration and its isotopic ra- tio (3He/4He). Although He isotopes were measured only in 11 samples, the 3He/4He ratio coupled with He/Ne ratio suggested clearly the distinction of two different waters, as shown in fig. 5. Five water samples showed a significant enrichment of radiogenic He (low 3He content) and these samples have also higher He mass concentration (mean of 3.2 ppmv) respect to 710 Giuseppe Etiope, Paolo Favali, Jean-Luc Fuda, Francesco Italiano, Matthias Laubenstein, Claude Millot and Wolfango Plastino the others (2 ppmv). The lower 3He content is also indicative of lower tritiugenic 3He, pro- duced by tritium decay within the water col- umn. Fig. 5. Helium isotope ratio versus He/Ne. The plot indicates two different waters probably linked to the different water masses whose interface oscillates pro- ducing T/S peaks. The isotope ratio of 1.4×10−6 is that of the atmosphere. Samples closer to this ratio indicate major atmospheric signal (shallower water). Lower ra- tios are along the mixing line with crustal (radiogenic) sources, and this can be indicative of bottom water, closer to seafloor, with higher residence time. Fig. 4. Helium variation. 3He/4He ratio (×10−6) is reported for each data point. Underlined ratios are those of the bottom-water group identified in fig. 5. The radioactivity data are coherent with this pattern (figs. 6 and 7). The same group of sam- ples having lower gas content and less radi- ogenic helium display drops of radioactivity for all radionuclides with typical values of standard seawater (Inn et al., 2001). 4. Discussion Physico-chemical and oceanographic data (one sample per hour) show a fair relationship between temperature/salinity, turbidity peaks and vertical movement and horizontal speed of water masses. The ADCP confirms the interpre- tation of CTD data regarding the existence of two water masses as it sensed the interface mo- tions, with remarkable dynamic features found coincident with the T-S peaks. As also support- ed by previous local CTD stratification profiles, it is most likely that this correlation is consis- tent with rapid lowering of the interface sepa- rating the warm, saline and more turbid waters of eastern origin (the so-called Eastern Over- flow Water, EOW), cascading from 400 m in the Channel of Sicily, from the underlying cold- er waters of western origin (WW), that are also «quiet» and hence not turbid, sampled most of the time by GEOSTAR. 711 The Benthic Boundary Layer: geochemical and oceanographic data from the GEOSTAR-2 Observatory Fig. 6. Variation of activity of some radionuclides. Lower activities correspond to T/S peaks and are indicative of shallower water mass, as suggested by helium and its isotopic ratio. Fig. 7. 235U versus 226Ra. A group of lower activity is distinguished, referring to samples closer to T/S peaks, in perfect analogy with fig. 5. The smaller plot shows the clustering in relation to the temporal distance (in days) of water sampling after a T/S event. The main result of geochemical analyses is the sharp distinction of two geochemically dif- ferent water masses, as suggested by He iso- topes (fig. 5) and radionuclides (fig. 7). The main problem is the comparison be- tween geochemical and hydrographic data, due to the limited availability of water samples for laboratory analysis, whose collection followed 712 Giuseppe Etiope, Paolo Favali, Jean-Luc Fuda, Francesco Italiano, Matthias Laubenstein, Claude Millot and Wolfango Plastino a pre-determined schedule; the coincidence of the water sampling (i.e. the geochemical da- tum) with the T-S peaks is therefore only casu- al, with shifts variable of several hours. It is not possible therefore to make a point-to- point correlation but from a detailed examination of all the time-series data and as it appears in figs. 3, 4 and 6, it is possible to observe that: – Samples relatively far or prior to T-S peaks have higher gas concentration, more He radi- ogenic and higher radioactivity. – Samples taken the same day or just some tens of hours after T-S peaks have gases and He isotopic ratio closer to atmospheric equilibrium and lower radionuclide concentration. Only two helium data, 10 December and 8 February, do not respect this pattern displaying atmospheric signature far or before T-S peaks. Moreover all geochemical data seem to not be influenced by the T-S peak no. 5, which is, in- deed, the shortest event. It is possible to argue that during or close to the major events of T-S variation the gas and ra- dionuclide content decreases and the He iso- topic ratio increases. This fact would support the interpretation of the CTD peaks as due to the vertical oscillation of the interface of two different water masses, close to the seafloor (fig. 8). Events of lower He and radionuclides, and higher R/Ra (closer to the atmospheric ra- tio) would reflect the sinking of a less deep wa- ter (EOW). However we can think about this water mass either of having been less deep, al- though still not in contact with the atmosphere, or as a younger water mass (having been more recently at the surface in its zone of formation). The deeper water (WW), likely with higher age, has higher gas content and radioactivity and a lower He isotopic ratio, with more significant radiogenic component. The enrichment of ra- dionuclides such as 226Ra in deep-sea waters due to diffusion from sediments is a well- known phenomenon (e.g., Nozaki, 1986). Fig- ures 5 and 7 show clearly the two water mass- es. This distinction is less marked by CO2 and CH4, which may have sources and sinks within the seawater column (and therefore a lower or nil difference between «seabed» signature and oceanic background) and the data can suffer higher biases from sampling to analysis. 5. Conclusions The GEOSTAR 2 data-set represents the first long-term multidisciplinary monitoring of deep BBL including altogether seawater tem- perature, salinity, light transmission, 3D cur- rent, gas concentration, helium isotopes and ra- dioactivity. The various geochemical and oce- anographic data are basically coherent converg- ing towards the indication of a BBL mainly characterised by a colder and fresher western water which is episodically displaced by the cascading of a warmer and saltier eastern water. The colder fresher water has higher concen- trations of elements produced in the lithosphere (helium, radionuclides and partially methane and carbon dioxide), likely due also to a higher benthic residence time. The warmer saltier wa- ter has geochemical features typical of ocean background. The geochemical data, although potentially affected by sampling and analytical biases, sug- gest that gas and radionuclides in BBL seawater, deriving from Earth degassing and diffusion from sediments, can be useful tracers to distin- Fig. 8. Sketch of the oscillating water mass inter- face in the BBL. EOW – Eastern Overflow Water; WW – Western Water; bnl – benthic nepheloyd lay- er; inl – intermediate nepheloyd layer. GEOSTAR detected WW for most of the time and events of EOW cascading, as suggested by T/S peaks, lower gas and radiogenic helium contents and lower ra- dionuclide activities. 713 The Benthic Boundary Layer: geochemical and oceanographic data from the GEOSTAR-2 Observatory guish water masses of different origin and benth- ic residence times. GEOSTAR-2 demonstrated the potential of long-term, continuous multipara- metric monitoring in providing unique informa- tion which cannot be acquired by traditional, short-term or single-sensor investigations. REFERENCES ARPESELLA, C. (1996): A low background counting facility at Laboratori Nazionali del Gran Sasso, Appl. Rad. Isot., 47, 991-996. BERANZOLI, L., A. DE SANTIS, G. ETIOPE, P. FAVALI, F. FRU- GONI, G. SMRIGLIO, F. GASPARONI and A. MARIGO (1998): GEOSTAR: a GEophysical and Oceanograph- ic STation for Abyssal Research, Phys. Earth Planet. Inter., 108, 175-183. BERANZOLI, L., T. BRAUN, M. CALCARA, D. CALORE, R. CAM- PACI, J.-M. COUDEVILLE, A. DE SANTIS, G. ETIOPE, P. FAVALI, F. FRUGONI, J.-L. FUDA, F. GAMBERI, F. GASPA- RONI, H.W. GERBER, M. MARANI, J. MARVALDI, C. MIL- LOT, P. PALANGIO, G. ROMEO and G. SMRIGLIO (2000): European seafloor observatory offers new possibilities for deep-sea study, Eos, Trans. Am. Geophys. Un., 81 (5), 45 and 48. BERANZOLI, L., P. FAVALI and G. SMRIGLIO (Editors) (2002): Science-technology synergy for research in marine en- vironment: challenges for the XXI century, Develop- ments in Marine Technology Series (Elsevier, Amster- dam), 12, pp. 268. BOUDREAU, B.P. and B.B. JORGENSEN (Editors) (2000): The Benthic Boundary Layer: Transport Processes and Biogeochemistry (Oxford University Press), pp. 440. ETIOPE, G. (1997): Evaluation of a micro-gas chromato- graphic technique for environmental analyses of CO2 and C1-C6 alkanes, J. Chromatogr. A, 775, 243-249. FUDA, J.-L., G. ETIOPE, C. MILLOT, P. FAVALI, M. CALCARA, G. SMRIGLIO and E. BOSCHI (2002): Warming, salting and origin of the Tyrrhenian deep water, Geophys. Res. Lett., 29, doi: 1029/2001GL014072. INN, K.G.W., ZHICHAO LIN, ZHONGYU WU, C. MCMAHON, J.J. FILLIBEN, P. KREY, M. FEINER, CHUNG-KING LIU, R. HOLLOWAY, J. HARVEY, L. LARSEN, T. BEASLEY, C.A. HUH, S. MORTON, D. MCCURDY, P. GERMAIN, J. HANDL, M. YAMAMOTO, B. WARREN, T.H. BATES, A. HOLMS, B.R. HARVEY, D.S. POPPLEWELL, M.J. WOODS, S. JERO- ME, K.J. ODELL, P. YOUNG and I. CROUDACE (2001): The NIST natural-matrix radionuclide standard reference material program for ocean studies, J. Radioanal. Nu- cl. Chem., 248 (1), 227-231. NOZAKI, Y. (1986): 226Ra-222Rn-210Pb systematics in seawater near the bottom of the ocean, Earth Planet. Sci. Lett., 80, 36-40. PLASTINO, W., M. LAUBENSTEIN, G. ETIOPE and P. FAVALI (2003): Environmental radioactivity analysis in sea- water samples collected by GEOSTAR deep-sea obser- vatory (Istituto Nazionale di Fisica Nucleare, Labora- tori Nazionali del Gran Sasso), Annual Rep. 2002, 201- 206. SPARNOCCHIA, S., G.P. GASPARINI, M. ASTRALDI, M. BOR- GHINI and P. PISTECK (1999): Dynamics and mixing of the Eastern Mediterranean outflow in the Tyrrhenian Basin, J. Mar. Syst., 20, 301-317. THIEL, H., K.O. KIRSTEIN, C. LUTH, U. LUTH, G. LUTHER, L.A. MEYER-REIL, O. PFANNKUCHE and M. WEYDERT (1994): Scientific requirements for an abyssal benthic laboratory, J. Mar Syst., 4, 421.