Comparison of Pelagic and Nepheloid Layer Marine Snow: Implications for Carbon Cycling

ELSEVIER Marine Geology 150 (1998) 39±50

Comparison of pelagic and nepheloid layer marine snow: implications for carbon cycling

Barbara Ransom a,Ł,KevinF.Sheab, Patti Jo Burkett b, Richard H. Bennett c, Roy Baerwald d a Scripps Institution of Oceanography, University of California at San Diego, La Jolla, CA 92093-0220, USA b Naval Research Laboratory, Stennis Space Center, MS 39529, USA c SEAPROBE Inc., 01 Pine Street, Picayune, MS 39566, USA d Department of Biology, University of New Orleans, New Orleans, LO 70148, USA Received 2 June 1997; accepted 8 May 1998

Abstract

Marine snow from upper and mid-water (i.e., pelagic) depths on the California margin is texturally and compositionally different from that traveling in the nepheloid layer. Transmission electron microscopy shows that pelagic marine snow consists primarily of bioclasts (e.g., diatom frustules, foram tests), organic matter, and microbes. These components are entrained as discrete particles or small aggregates .Ä 10 µm in diameter) in a loose network of exocellular, muco-polysaccharide material. Clays are infrequent but, when present, are constituents of comparatively compact organic-rich microaggregates. Microbes are abundant and appear to decrease in number with increasing water depth. In contrast, marine snow aggregates collected from just above the sea ¯oor in the nepheloid layer are assemblages of clay particles, clay ¯ocs, and relatively dense clay±organic-rich microaggregates in an exocellular organic matrix. Bioclasts and microorganisms occur only rarely. The prevalence of clay±organic-rich aggregates in the nepheloid layer suggests that, prior to ®nal deposition and burial, marine snow from the pelagic zone is subject to disaggregation and recombination with terrigenous detrital material near or at the sea ¯oor. Results have signi®cant implications for the accumulation and burial rates of organic carbon on continental margins and the aging and bioavailability of sedimentary organic matter. Samples examined were collected offshore of northern and central California.  1998 Elsevier Science B.V. All rights reserved.

Keywords: nepheloid layer; marine snow; organic carbon; continental margin; TEM

1. Introduction has focused on surface to mid-water column aggregate production and the rates and variability of Marine snow, the population of marine aggre- organic carbon transport to the sea ¯oor (see review gates with diameters greater than 5 mm, is the in Alldredge and Silver, 1988). Recent research now primary means by which organic matter is trans- suggests that marine snow aggregates may also play ported down through the water column (Fowler and an important role in the development of sediment Knauer, 1986). In large part, marine snow research geotechnical properties (Syvitski, 1991) and the ultimate distribution of organic matter in continental Ł Corresponding author. Fax: C1 (619) 534-0784; E-mail: margin sediments (Ransom et al., 1997). Below we [email protected] present the results of a comparative study of the com-

0025-3227/98/$19.00  1998 Elsevier Science B.V. All rights reserved. PII S0025-3227(98)00052-8 40 B. Ransom et al. / Marine Geology 150 (1998) 39±50 ponents of marine snow from various water depths fabric and sediment chemical and physical properties offshore California. Our objectives were to identify across the California continental slope (Kim, 1996; the components and textures of aggregates traveling Ransom et al., 1997, 1998; Sawyer et al., 1997). in the nepheloid layer, improve our understanding Mendocino samples were retrieved from an anchored of how processes and conditions in the nepheloid array deployed for 64 days in ¾950 m of water with layer affect marine aggregates that originate in the collectors at water depths of 250, 750, and 930 m upper water column (pelagic zone), and examine the at 40ë1.990N, 124ë31.60W. San Luis Obispo samples possible impact these interactions may have on the were recovered from a ¯oating array deployed for 24 distribution of organic matter in the sediment. h at 35ë27.50N and 121ë31.00W. The ¯oating array Marine snow is a complex assemblage of or- had ®ve evenly spaced collectors at water depths ganisms, bioclasts (e.g., tests and frustules), organic from 50 to 250 m. matter, minerals, and fecal material entrained as dis- Trap depths were selected on the basis of trans- crete particles or small aggregates in a loose network missometer data which indicated that the water col- of exocellular muco-polysaccharide slime and ®brils umn in the sampling areas could be divided into (e.g., Alldredge, 1989; Decho, 1990; Heissenberger three zones. The surface layer, shown in representa- et al., 1996). The composition of marine snow is tive transmissometer traces in Fig. 2, coincides with a function of local particle abundance, particle en- the turbid photic zone which is characterized by high counter rates, and the ease with which different com- particle concentrations resulting from biological pro- ponents can be incorporated into larger aggregates. ductivity and sea-surface aggregation processes. The Through aggregation and disaggregation, as well as composition, size, and number of aggregates formed grazing by bacterial and planktonic heterotrophs, the in this layer are episodically affected by suspended architecture, particle composition and organic matter material from rivers during periods of heavy conti- in marine snow evolves through time (see review in nental runoff. The pelagic layer, a zone of reduced Wakehan and Lee, 1993). particle concentration, through which marine snow The direct observation of marine snow con- formed in the surface layer descends, occurs at the stituents and their textural relations requires use base of the surface layer and extends to the top of of a transmission electron microscope (TEM). To the nepheloid layer. The nepheloid layer stretches date, such studies have concentrated on fecal pel- from the sediment±water interface up to the base lets and=or aggregates generated in surface waters of the pelagic zone. This layer directly overlies the and pelagic regions of the water column (e.g., Silver sea ¯oor and experiences episodic turbidity from the and Bruland, 1981; Gowing and Silver, 1983, 1985; lateral transport and resuspension of detrital clastic Leppard et al., 1996). Only one study contains im- material as a result of oceanographic, sedimentolog- ages of nepheloid aggregates (Ransom et al., 1997). ical, and biological processes (e.g., bottom currents, The potential of this class of aggregates to in¯uence storms, sediment gravity ¯ows, slope failure, organ- sea ¯oor sediment microfabric and the distribution ism burrowing, etc.). To some extent, the boundaries and bioavailability of organic matter warrants further between the three layers shown in Fig. 2 are tem- investigation. porally and spatially variable, being dependent on water depth and prevailing biological, physical, and oceanographic conditions. 2. Sampling locations

Marine snow samples were collected with Martin- 3. Methods Knauer multi-collector (eight tubes per water depth) sediment traps (Knauer et al., 1979) from transects Samples were collected and prepared to ensure that near Mendocino and San Luis Obispo on the Califor- aggregates represented the in-situ textures and parti- nia continental margin (Fig. 1) from June 13 to Au- cles present in the water column. Two of the eight col- gust 16, 1993. These samples form a companion data lectors at each water depth were used to obtain sam- set to a high-resolution study of the sediment micro- ples for TEM study; the six others were used for an B. Ransom et al. / Marine Geology 150 (1998) 39±50 41

Fig. 1. Location map of the Mendocino (northern) and San Luis Obispo (southern) locations sampled off the coast of California (see inset). Bathymetric contours are in meters. Sediment trap locations are indicated by crosses. investigation of organic carbon ¯ux and spiked with Prior to processing the carbon ¯ux material, formalin to kill planktonic scavengers. The TEM-des- planktonic scavengers were removed from the ®lter ignated collectors were ®lled with pH-buffered sea- papers so as not to affect the mass of organic carbon water containing 2 wt% glutaraldehyde. Sucrose was corresponding to that of the water column particles. added to these to increase solution density to ¾1.07 Filter papers were then dried and weighed to obtain gcm3 to minimize washout during trap recovery. the mass of marine aggregate material collected. The Upon retrieval, liquid in the upper three quarters of dried material was analyzed for total carbon using the TEM collectors was slowly siphoned off. All but a Perkin Elmer CHN analyzer (precision Dš0:3%, 50 ml of the remaining liquid was then slowly drained based on repeated measurements of a cystine stan- through the bottom of the collector. The resulting sus- dard) and inorganic carbon using acid-dissolution pension, which contained marine snow, was gently coulometry (precision Dš0:5%, based on repeated decanted into a jar spiked with glutaraldehyde and analysis of a carbonate standard). Organic carbon osmium tetroxide and refrigerated. In the carbon ¯ux was determined by difference. For further discus- collectors, ¯uid in each column was drained through sion of carbon analytical procedures and errors see the bottom and all entrained particulate matter was Ransom et al. (1998). deposited on a tared ®lter paper. In the laboratory, marine snow which was sus- 42 B. Ransom et al. / Marine Geology 150 (1998) 39±50

Fig. 2. Transmissometer plots from the Mendocino and San Luis Obispo sediment trap transects. The distinct gradient changes at the base of the surface and pelagic layers were used to distinguish boundaries between layers and shaded zones and indicate the relative extent of each layer. pended in the TEM-designated collectors was gently clastic particles (pieces of diatom frustules), clay separated with forceps into pieces 1±3 mm in di- minerals, organic matter, and microbes. Identi®ca- ameter in preparation for embedding. Aggregates tion of speci®c components carries a number of were then quickly ethanol-exchanged (¾10 min per caveats and the reliability of such determinations is ¯uid) and embedded in Spurr's resin (Hayat, 1970). a function of the operators' expertise and experi- From these stubs, ultrathin-sections were cut and ence with the materials and their familiarity with mounted on TEM grids and selectively post-block- differences in appearance due to vagaries in parti- stained with uranyl acetate and lead citrate (Hayat, cle orientation in the TEM section. For this study, 1970; Baerwald et al., 1991). Samples were imaged our photomicrographs were examined by operators with a Hitachi H600 TEM at 100 kV. Approximately familiar with organic, biological, and mineralogical 80 grids were scanned with the TEM and hundreds components of marine sediments using TEM. In gen- of ®elds of view containing thousands of particles eral, bioclasts were identi®ed on the basis of their and hundreds of microaggregates were examined. Of geometric structures (e.g., smooth curves, polygonal these, photomicrographs of more than 300 represen- shapes), repetitive structural appearance (e.g., evenly tative ®elds of view were taken. Those that represent spaced protrusions or holes) and comparatively high the best examples of the images obtained are shown electron density. Clay minerals, when oriented per- in the plates below. pendicular to the section, were identi®ed by their For the most part, aggregates examined by TEM generally high electron density, commonly elongated contained four primary components: siliceous bio- or ®brous shapes with smooth upper and lower B. Ransom et al. / Marine Geology 150 (1998) 39±50 43 boundaries and feathery lateral edges. When ori- density, web-like material. TEP and its effect on ented parallel to the section, they were identi®ed by marine snow textures have been described in the lit- their relatively low electron density equant shapes, erature (e.g., Passow and Alldredge, 1994; Dam and generally with well de®ned boundaries. Amorphous Drapeau, 1995; Heissenberger et al., 1996) and the organic matter was generally identi®ed by its rela- features noted therein are consistent with those we tively low electron density, irregular size and shape, observed. and patchy appearance. Cellular material was identi®ed by its grainy appearance and the presence of 4.1. Pelagic aggregates organelles or vacuoles. Microbes were identi®ed by their circular shape, size (0.2 to 1 µmindiame- Aggregates from three depths in the pelagic layer ter), heavy take up of stain, and presence of internal over the California continental slope were examined. cellular structures. Representative images of samples from the three water depths (150 m in San Luis Obispo; 250 and 750 m in Mendocino) are shown in Fig. 3. Aggre- 4. Results and discussion gates from both locations and all three depths are strikingly similar in their relatively open architecture There is a large difference in size between ma- and overall composition. It is well known that phys- rine snow .> 5 mm in diameter) and its constituent ical conditions and water column biology control, in particles (¾10 to 0.1 µm in size). In the present large part, the types and compositions of aggregates study, we focused on the textures and compositions that form. The similarity in composition and struc- of the building blocks of marine snow, i.e., microag- ture of aggregates from the two deployment sites can gregates, those aggregates Ä10 µm in diameter, non- be accorded to their similar oceanographic charac- organic particles, and organic matter. Representative teristics (Huyer, 1983; Huyer et al., 1991; Strub et photomicrographs of marine snow components from al., 1991) and surface water biological communities pelagic and nepheloid layer sediment traps are shown (Rau et al., 1982; Sancetta et al., 1992). in Figs. 3 and 4. It should be noted that each piece Pelagic aggregates from both sampling sites and of marine snow embedded for our study acted as a the water depths sampled are dominated by siliceous loose, yet coherent mass when suspended in solu- bioclasts (pieces of diatom frustules) and amorphous tion. This coherency indicated that the constituent organic matter, with the textural relations of com- particles were all interlinked in some manner. This ponents of typical microaggregates being empha- interlinking is provided by exocellular organic mate- sized in Fig. 3. Within the larger aggregated masses, rial, similar in character to the polysaccharide ®brils microaggregates .Ä 10 µm) are common (e.g., and slime imaged by Leppard et al. (1996). In ma- Fig. 3A). These tend to be comprised of particles rine studies such material is commonly referred to as that are more closely associated with one another transparent exopolymer particles or TEP (Alldredge than with those in the surrounding aggregate. Not all et al., 1993; Schuster and Herndl, 1995). These ex- particles in the pelagic aggregates are found in mi- tensive slime and ®bril networks originate from a croaggregates. Individual bioclasts, caught up in the variety of sources, including the linking together of TEP network of the larger aggregates, are also preva- bacterial mucus secretions (Decho, 1990), and aggre- lent (e.g., Fig. 3A,D). Organic matter on the whole gation of colloids and mucus secretions by diatoms appears to be amorphous in character although, on (e.g., Logan et al., 1995). In general, we did not occasion, large pieces of disrupted cellular material directly observe TEP in our samples because the are evident (Fig. 3C). Microbes in pelagic aggre- non-speci®c organic staining protocol we adopted gates occur as both individuals (Fig. 3A) as well was best suited for our overall objectives, while as components of microaggregates (Fig. 3B); their optimal resolution of TEP requires a polysaccharide- abundance appears to decrease with water depth speci®c stain such as Ruthenium Red (Leppard et al., (compare Fig. 3A,B with Fig. 3E,F). 1996). Nevertheless, on occasion, TEP was detected Overall, clay minerals are a minor component and, when present, occurred as a thin, low electron of the pelagic aggregates caught in our traps. 44 B. Ransom et al. / Marine Geology 150 (1998) 39±50 B. Ransom et al. / Marine Geology 150 (1998) 39±50 45

When present, clays occur primarily as compo- atively more dense and composed dominantly of clay nents of clay±organic-rich microaggregates or ¯ocs minerals and organic matter. The large bioclasts that (Fig. 3D). Only infrequently do dispersed clay par- dominate pelagic macroaggregates are infrequent, ticles occur inside the bioclast-rich pelagic aggre- and when present commonly occur as single entities gates (Fig. 3C,F). It should be noted that bioclasts or as isolated particles (e.g., Fig. 4A,C). On oc- are rarely present inside clay-dominated microaggre- casion, a bioclast-dominated microaggregate, which gates. Because the composition of marine aggregates displays textures and compositions identical to those is a function of particle abundance, it is probable that in the overlying column, appears to be included, with the clay-rich microaggregates formed prior to enter- little or no modi®cation, as part of a larger nepheloid ing an area of high marine productivity. Thus, these aggregate. Such behavior is predicted by studies of aggregates probably formed near shore as the result aggregate strength (Alldredge et al., 1990). In such of the ¯occulation of clay minerals suspended in studies, resistance to disaggregation is closely re- river plumes during the mixing of salt and fresh wa- lated to aggregate size, with resistance increasing as ter. After formation, these clay-rich microaggregates aggregate diameter decreases. must have been advected out over the continental Organic matter in our nepheloid aggregates is pri- slope into the area of high biological productivity marily amorphous in character. For the most part it where they could be incorporated into bioclast-dom- is closely associated with the clays (e.g., Fig. 4E,F). inated marine snow aggregates. The relative paucity In some instances, degrading organic matter forms of clay in our pelagic layer aggregates indicates that a nucleus around which clays accumulate (Fig. 4D). during the time our traps were deployed, the impact Overall, however, the proportion of organic matter of cross-shelf processes on aggregate development in nepheloid aggregates appears to be signi®cantly was minimal. less than that in pelagic aggregates. This observation is supported by the organic carbon content of 4.2. Nepheloid aggregates pelagic and nepheloid layer marine snow caught in our traps (Table 1). These data show a more than Representative TEM images of aggregates col- fourfold decrease in the wt% organic carbon in ma- lected about 20 m above the sea ¯oor inside the terial collected in the nepheloid layer, compared to nepheloid layer in the Mendocino area are shown in that retrieved from the upper part of the water col- Fig. 4. As for our pelagic study, the textural relations umn. Although a qualitative assessment, nepheloid and components of microaggregates representative organic matter appears, overall, to be slightly more of those comprising the larger aggregates are em- electron dense under the beam, suggesting that it phasized. Comparison of images in Figs. 3 and 4 may be more highly polymerized (i.e., degraded) shows that nepheloid aggregates differ signi®cantly, than that in pelagic aggregates. both texturally and compositionally, from those in Another notable feature of our nepheloid aggre- the overlying water column. Unlike the pelagic ag- gates is the apparent near absence of microbes. gregates, those from the nepheloid layer are compar- While this could be an artifact of sampling or di-

Fig. 3. Representative TEM photomicrographs of pelagic layer marine snow particles and aggregates from the water column above the California continental slope offshore of San Luis Obispo (SLO) and Mendocino (M). Scale bars D 2 µm. (A) Partial view of a typical pelagic macroaggregate (SLO, 150 m water depth) containing bacteria (b), bioclasts (bc), and amorphous organic matter (o). Note the lack of clay and the presence of discrete microaggregates (ma) within the larger aggregate that ®lls the entire ®eld of view; matrix displays a signi®cant lack of clay grains or domains. (B) Typical pelagic microaggregate dominated by bioclasts, bacteria, and organic matter, within a larger macroaggregate (SLO, 150 m water depth). (C) Microaggregate containing bioclasts, localized zones of degraded amorphous organic matter, and a clay particle (c) (SLO, 150 m water depth). A mass of degraded cellular material (dcm) caught in the marine snow is located just above the microaggregate. (D) Unusual pelagic, clay±organic-matter-dominated, microaggregate. Clays tend to be arranged as domains (i.e., parallel stacks) and chains (M, 250 m water depth). (E) Bioclast-dominated microaggregate with amorphous organic matter (M, 750 m water depth). (F) Example of a rare mixed microaggregate containing bioclasts, clay, and organic matter (M, 750 m water depth). 46 B. Ransom et al. / Marine Geology 150 (1998) 39±50 B. Ransom et al. / Marine Geology 150 (1998) 39±50 47

Table 1 been described (Gowing and Silver, 1985). There is, Mass and composition of particulate matter of pelagic and neph- however, a substantial degree of uncertainty associ- eloid samples from the Mendocino moored sediment trap array ated with positive identi®cation of these aggregates a a Water depth Mass Corg Cinorg as mini-fecal pellets (M. Silver, pers. commun.). The (m) (g) (wt%) (wt%) possibility that some nepheloid microaggregates may 250 (pelagic) 0.161 13.33 1.28 have originated as mini-fecal pellets cannot be over- 750 (pelagic) 0.124 9.13 1.66 looked. Nevertheless, in the case of understanding 930 (nepheloid) 0.920 3.77 0.85 aggregate deposition and origin, the process of ag- Note. Collection surface area for samples from each water depth gregation is less important than its location of origin. was 48.7 cm2. With the exception of omnivorous protozoans that a Corg and Cinorg denote organic and inorganic carbon, respec- might preferentially feed on clay particles in the gen- tively. erally clay-poor pelagic layer aggregates in the upper water column, the origin of the nepheloid pellet-like lution due to increased amounts of inorganic detritus microaggregates we have observed must be either in suspension in the nepheloid layer (Table 1), this within the clay-rich nepheloid layer or in underlying appears to represent the continuation of the trend surface sediments that have been resuspended. observed in the pelagic aggregates, i.e., decreasing bacterial colonization of aggregates with increasing 4.3. Aggregate evolution and implications for water depth. organic carbon distribution and burial The textural arrangement of the clays in the nepheloid aggregates collected in our study was also The change in composition between pelagic and examined. These arrangements commonly consisted nepheloid layer marine snow we have observed can of random domains and chains of clay particles be explained in terms of aggregation theory. For a (e.g., Fig. 4C,D,F). Upon closer inspection, sub- particle to be incorporated into an aggregate, it must units in the chains and domains consist of parallel come in contact with the aggregate and stick upon and sub-parallel arrangements of clay minerals (see contact. These two conditions are referred to as con- Fig. 4A,C). Such sub-units may represent small mi- tact rate and contact ef®ciency (Hill, 1992; Alldredge croaggregates that were later incorporated into larger and Jackson, 1995). Contact rate depends on particle microaggregates. A class of microaggregates com- and aggregate size, concentration, sinking rate, and monly observed was one with clays arranged tan- on the amount of energy in the system. Contact ef®- gentially along the boundary (e.g., Fig. 4E; also see ciency is a measure of how many contacts result in Ransom et al., 1997). These latter aggregates re- the incorporation of a particle and is a function of the semble fecal pellets in morphology, but their size, `stickiness' of the colliding particles=aggregates. The compactness, and the absence of a discernible pel- abundance of a given type of particle in a population licle leaves their origin in doubt. Similar marine of aggregates is an indication of its concentration in aggregates, labeled mini-pellets and attributed to fe- the water column and the energy of the system at the cal material of protozoa and planktonic larva have time of aggregation. As can be seen in Figs. 3 and

Fig. 4. Representative TEM photomicrographs of nepheloid layer particles and aggregates about 20 m above the sea ¯oor in 950 m of water on the California continental slope near Mendocino. Scale bars D 2 µm. (A) Partial view of typical nepheloid macroaggregate containing clay minerals (c), amorphous organic matter (o), and occasional dispersed large bioclasts (bc). Dark lines indicated by AR are artifacts resulting from wrinkling of the section and h represents a hole in the section. (B) Typical nepheloid microaggregate dominated by clays and amorphous organic matter. (C) Clay-dominated microaggregate containing clay and rare large bioclastic pieces. A single large bioclast caught in the marine snow is located just below the microaggregate. (D) Clay±organic-matter-dominated microaggregate with a core of organic matter around which clays have accumulated. (E) Clay±organic-rich microaggregate showing parallel alignment and chaining of clays internally and around the exterior of the aggregate. Note the presence of stacked clay domains within some of the clay particles. (F) Clay±organic-matter-dominated microaggregate with a few large bioclasts associated on its boundary and nearby. Note arrangement of stacked clay domains. 48 B. Ransom et al. / Marine Geology 150 (1998) 39±50

4, the inorganic particulate species that predominate exact mechanism by which the transfer occurred. in pelagic and nepheloid systems are different. From Conclusions drawn on the basis of our observations the bioclast-rich, clay-poor character of the pelagic would provide such a mechanism. aggregates we imaged, it its apparent that primary As can be deduced from composition, the pelagic productivity in the surface layer controls aggregate aggregates examined in our study were formed near composition in much of the water column in our the sea surface. Once greater than about 5 mm in study area (i.e., the surface and pelagic regions). The diameter, such aggregates would have attained suf- dramatic change in aggregate character that occurs ®cient size and mass to sink through the water col- near the sea ¯oor must therefore be attributed to a umn. In general, as marine snow passes through the signi®cant change in both particle composition and less turbid (i.e., less aggregate-rich) pelagic layer, water column energetics. remineralization of incorporated organic matter oc- The nepheloid layer is a dynamic and ubiquitous curs (e.g., Alldredge et al., 1986). Consequently, phenomenon that occurs on continental slopes. Its aggregates reaching the nepheloid layer contain less presence, however, can be temporally and spatially organic matter, as well as that which is less labile variable (see review in Gorsline, 1984). Suspended than that of its counterparts in the overlying water particle concentrations in this layer are generated column. Indeed, this has been documented by or- and maintained by a variety of processes that bring ganic geochemical studies of sediment trap material in new detritus and=or resuspend and laterally advect from the Santa Monica Basin just a few tens of previously deposited clastic material (Biscaye et al., kilometers south of our San Luis Obispo sampling 1988; Chasse®ere, 1990; Sherwood et al., 1994). The site (Venkatesan and Kaplan, 1992). It is therefore change in character with increasing water depth in likely that the clay-dominated nepheloid aggregates the marine snow we examined indicates that when we examined contain organic matter that is more re- pelagic marine snow intersects the nepheloid layer, fractory than that in their upper water column pelagic it encounters a zone of comparatively high shear counterparts. and terrigenous particle concentration. At times of The release of insecurely bound, possibly aged, high pelagic particle ¯ux, such as plankton blooms, pelagic organic matter into the nepheloid layer and pelagic macroaggregates falling to the sea ¯oor may it role in nepheloid aggregation provides a mecha- scavenge ®ne detritus moving laterally in the neph- nism by which clay particles and organic matter can eloid layer and arrive at the sea ¯oor virtually in- become closely associated to produce the nepheloid tact (Hill and Nowell, 1990). However at the ¯ux aggregates observed in our study. Similar textural re- rates that more commonly prevail, the higher-energy lations between clays and organic matter have been regime near the sea ¯oor and the subsequent changes observed in sediments deposited on the sea ¯oor in density distribution that pelagic macroaggregates (Lavoie et al., 1996; Ransom et al., 1997). This sug- in our study area must experience, as they are ex- gests that the association of organic matter with ter- posed to shear, seem to destabilize their structure and rigenous clastic material in continental margin sedi- cause disaggregation. This releases pelagic aggregate ments and the distribution of a potentially signi®cant components (i.e., microaggregates, TEP, cellular ma- percentage of the organic matter in such sediments terial, etc.) into the siliclastic-rich nepheloid system happens prior to its permanent deposition and burial. where they are assimilated by nepheloid aggregates. The redistribution of pelagic organic matter among It would, therefore, appear that nepheloid processes terrigenous clays in the nepheloid layer provides a redistribute organic matter descending from the over- mechanism that explains the recently observed cor- lying water column and disperse this material down relation between sediment organic carbon content, and across the California continental slope. Similar speci®c surface area, and clay content (Mayer, 1994; conclusions regarding the offshore transport of or- Keil et al., 1994; Ransom et al., 1998) and may ganic carbon at other sites have been drawn on the shed light on the controversy surrounding the con- basis of changes in sediment bulk properties (e.g., trolling mechanisms of organic carbon preservation Walsh et al., 1981; Grant et al., 1987; Walsh, 1989). on continental margins. A further possible conse- These studies, however, were unable to specify the quence of the periodic resuspension and deposition B. Ransom et al. / Marine Geology 150 (1998) 39±50 49 that can occur as a result of nepheloid layer±surface Acknowledgements sediment interactions is a considerable increase in organic matter residence time at, or near the sea ¯oor This research was carried out with funding from in oxic or suboxic aerobic conditions. It would seem the Of®ce of Naval Research, Program Element that such an increase in exposure could substantially 0601153N and ONR grants N00014: 92-J-1225 and increase the degradation of organic matter prior to its 98-10183. The authors would like to thank Nancy ®nal incorporation and burial in the sediment, affect- Carnaggio for sample preparation and photographic ing signi®cantly its bioavailability to local benthic work. We would also like to thank Vernon Asper, fauna. Miriam Kastner, and Dennis Lavoie and for many stimulating conversations on this and related topics.

5. Conclusions References Marine aggregates in the upper water column are signi®cantly different in texture and composition Alldredge, A.L., 1989. The signi®cance of suspended detrital from those that occur in the nepheloid layer. Because aggregates of marine snow as microhabitats in the pelagic zone of the ocean. In: Hattori, T., Ishida, Y., Muruyama, Y., marine snow is the dominant means by which or- Morita, R., Uchida, A. (Eds.), Recent Advances in Microbial ganic matter is transported to the sea ¯oor, our study Ecology, Proceedings of the 5th International Symposium on provides insights into processes that ultimately con- Microbial Ecology, Japan. Scienti®c Societies Press, pp. 108± trol the organic carbon contents of surface sediments 112. and the preservation and bioavailability of sedimen- Alldredge, A.L., Jackson, G.A., 1995. Aggregation in marine systems. Deep-Sea Res. II 42, 1±7. tary organic matter. Our investigation suggests that Alldredge, A.L., Silver, M.W., 1988. Characteristics, dynamics pelagic marine snow aggregates undergo repeated and signi®cance of marine snow. Prog. Oceanogr. 20, 41±82. episodes of disaggregation, aggregation, and mixing Alldredge, A.L., Cole, J.J., Caron, D.A., 1986. Productivity of with terrigenous detritus in the nepheloid layer near heterotrophic bacteria inhabiting macroscopic organic aggre- and at the sea ¯oor, such that the components, tex- gates (marine snow) from surface waters. Limnol. Oceanogr. 31, 68±78. tural relations, and distribution of organic matter in Alldredge, A.L., Granata, T.C., Gotshalk, C.C., Dickey, T.D., the resulting macroaggregates bear little resemblance 1990. The physical strength of marine snow and its im- to those in the overlying water column. It is these plications for particle disaggregation in the ocean. Limnol. strongly modi®ed bottom-water aggregates and their Oceanogr. 35, 1415±1428. included organic matter that provide the baseline for Alldredge, A.L., Passow, U., Logan, B.E., 1993. The abundance and signi®cance of a class of large, transparent organic parti- understanding the preservation of organic carbon in cles in the ocean. Deep-Sea Res. I 40, 1131±1140. continental margin sediments and the evolution and Baerwald, R.J., Burkett, P.J., Bennett, R.H., 1991. Techniques in-situ distribution of sedimentary organic matter. for the preparation of submarine sediments for electron mi- Our study indicates that for oxic and suboxic conti- croscopy. In: Bennett, R.H., Bryant, W.R., Hulbert, M.H. (Eds.), Microstructure of Fine-Grained Sediments. Springer, nental margins, models of organic carbon deposition Berlin, pp. 309±331. should include provisions for lateral transport of or- Biscaye, P.E., Anderson, R.F., Deck, B.L., 1988. Fluxes of par- ganic matter and its possible extended residence time ticles and constituents to the eastern United States continental in near-sea-¯oor, oxygenated, aerobic environments. slope and rise: SEEP-I. Cont. Shelf Res. 8, 855±904. During such intervals, organic matter would be re- Chasse®ere, B., 1990. Mass-physical properties of sur®cial sediments on the Rhone continental margin: implications for the peatedly cycled between surface sediments and the nepheloid benthic layer. Cont. Shelf Res. 10, 857±867. overlying bottom water, being moved down slope Dam, H.G., Drapeau, D.T., 1995. Coagulation ef®ciency, or- from its initial point of impact with the sea ¯oor to ganic-matter glues and the dynamics of particles during a its site of ultimate burial and removal from the car- phytoplankton bloom in a mesocosm study. Deep-Sea Res. II bon cycle. It seems likely that such a process would 42, 111±123. Decho, A.W., 1990. Microbial exopolymer secretions in ocean increase signi®cantly the refractory character of the environments: their role(s) in food webs and marine processes. organic matter, thus reducing its bioavailability to Oceanogr. Mar. Biol. Annu. Rev. 28, 73±153. benthic fauna. Fowler, S.W., Knauer, G.A., 1986. Role of large particles in 50 B. Ransom et al. / Marine Geology 150 (1998) 39±50

the transport of elements and organic compounds through the rial colonization of transparent exopolymer particles (TEP) in oceanic water column. Prog. Oceanogr. 16, 147±194. the ocean. Mar. Ecol. Prog. Ser. 113, 185±198. Gorsline, D.S., 1984. A review of ®ne-grained sediment origins, Ransom, B., Bennett, R.H., Baerwald, R., 1997. TEM study of characteristics, transport and deposition. In: Stow, D.A.V., in situ organic matter on continental margins: occurrence and Piper, D.J.W. (Eds.), Fine-Grained Sediments: Deep Water the `monolayer' hypothesis. Mar. Geol. 138, 1±9. Processes and Facies. Blackwell, Oxford, pp. 17±33. Ransom, B., Kim, D., Kastner, M., Wainwright, S., 1998. Or- Gowing, M.M., Silver, M.W., 1983. Origins and microenviron- ganic matter preservation on continental slopes: importance of ments of bacteria mediating fecal pellet decomposition in the mineralogy and surface area. Geochim. Cosmochim. Acta 62, sea. Mar. Biol. 73, 7±16. 1329±1345. Gowing, M.M., Silver, M.W., 1985. Mini-pellets: A new and Rau, G.H., Sweeney, R.E., Kaplan, I.R., 1982. Plankton 13C=12C abundant size class of marine fecal pellet. J. Mar. Res. 43, ratio changes with latitude: differences between northern and 395±418. southern oceans. Deep-Sea Res. 29, 1035±1039. Grant, J., Volckaert, F., Roberts-Regan, D.L., 1987. Resuspend- Sancetta, C., Lyle, M., Heusse, L., Zahn, R., Bradbury, J.P., 1992. able organic matter in Nova Scotian shelf and slope sediments. Late Glacial to Holocene changes in winds, upwelling, sea- Cont. Shelf Res. 7, 1123±1138. sonal productivity of the northern California current system. Hayat, M.A., 1970. Principles and Techniques of Electron Mi- Quat. Res. 38, 359±370. croscopy, Vol. 7. Van Nostrand Reinhold, New York. Sawyer, W., Ransom, B., Bennett, R.H., 1997. Geotechnical Heissenberger, A., Leppard, G.G., Herndl, G.J., 1996. Ultrastruc- property variability of continental margin sediments: high-res- ture of marine snow, II. Microbiological considerations. Mar. olution lateral and vertical data from the northern California Ecol. Prog. Ser. 135, 299±308. slope. Mar. Georesour. Geotechnol. 15, 283±304. Hill, P.S., 1992. Reconciling aggregation theory with observed Schuster, S., Herndl, G.J., 1995. Formation and signi®cance of vertical ¯uxes following phytoplankton blooms. J. Geophys. transparent exopolymeric particles in the northern Adriatic Res. 97, 2295±2308. Sea. Mar. Ecol. Prog. Ser. 124, 227±236. Hill, P.S., Nowell, A.R.M., 1990. The potential role of large, Sherwood, C.R., Butman, B., Cacchione, D.A., Drake, D.E., fast-sinking particles in clearing nepheloid layers. Philos. Gross, T.F., Sternberg, R.W., Wiberg, P.L., Williams, A.J., Trans. R. Soc. London A 331, 103±117. 1994. Sediment-transport events on the northern California Huyer, A., 1983. Coastal upwelling in the California current. continental shelf during the 1990±1991 STRESS experiment. Prog. Oceanogr. 12, 259±284. Cont. Shelf Res. 14, 1063±1099. Huyer, A. et al., 1991. Currents and water masses of the coastal Silver, M.W., Bruland, K.W., 1981. Differential feeding and fecal transition zone off northern California, June to August, 1988. pellet composition of salps and pteropods, and the possible J. Geophys. Res. 96, 14743±14768. origin of deep-water ¯ora and olive-green `cells'. Mar. Biol. Keil, R.G., Tsamakis, E., Fuh, C.B., Giddings, C., Hedges, J.I., 40, 359±369. 1994. Mineralogical and textural controls on the organic com- Strub, P.T., Kosro, P.M., Huyer, A., 1991. The nature of cold position of coastal marine sediments: hydrodynamic separation ®laments in the California current system. J. Geophys. Res. using SPLITT-fractionation. Geochim. Cosmochim. Acta 58, 96, 14743±14768. 879±893. Syvitski, J.P.M., 1991. The changing microfabric of suspended Kim, D., 1996. Biochemical Cycling of Carbon, Phosphorous, particulate matter Ð the ¯uvial to marine transition: ¯oc- and Silica in California Continental Slope Sediments. Ph.D. culation, agglomeration, and pelletization. In: Bennett, R.H., dissertation, Univ. of California at San Diego, La Jolla, CA. Bryant, W.R., Hulbert, M.H. (Eds.), Microstructure of Fine- Knauer, G.A., Martin, J.H., Bruland, K.W., 1979. Fluxes of Grained Sediments. Springer, Berlin, pp. 131±137. particulate carbon, nitrogen, and phosphorus in the upper Venkatesan, M.I., Kaplan, I.R., 1992. Vertical and lateral trans- water column of the northeast Paci®c. Deep-Sea Res. 26A, port of organic carbon and the carbon budget in Santa Monica 97±108. Basin, California. Prog. Oceanogr. 30, 291±312. Lavoie, D.M., Lavoie, D.L., Pittenger, H.A., Bennett, R.A., 1996. Wakehan, S.G., Lee, C., 1993. Production, transport, and alter- Bulk sediment properties interpreted in light of qualitative and ation of particulate organic matter in the marine water column. quantitative microfabric analysis. Geo-Mar. Lett. 16, 226±231. In: Engel, M.H., Macko, S.A. (Eds.), Organic Geochemistry. Leppard, G.G., Heissenberger, A., Herndl, G.J., 1996. Ultra- Plenum Press, New York, pp. 145±169. structure of marine snow, I. Transmission election microscopy Walsh, J.J., 1989. How much shelf production reaches the deep methodology. Mar. Ecol. Prog. Ser. 135, 289±298. sea? In: Berger, W.H., Smetacek, V.S., Wefer, G. (Eds.), Pro- Logan et al., 1995. Deep-Sea Res. II, 42: 203±214. ductivity of the Ocean: Present and Past. Wiley, New York, pp. Mayer, L.M., 1994. Surface area control of organic carbon ac- 175±191. cumulation in continental shelf sediments. Geochim. Cos- Walsh, J.J., Rowe, G.T., Iverson, R.L., McRoy, C.P., 1981. Bi- mochim. Acta 58, 1271±1284. ological export of shelf carbon is a sink of the global CO2 Passow, U., Alldredge, A.L., 1994. Distribution, size and bacte- cycle. Nature 291, 196±201.