Defecation by Salpa thompsoni and its contribution to vertical flux in the Southern Ocean

Measurements of the defecation rate of Salpa thompsoni were made at several stations during two cruises west of the Antarctic Peninsula in 2004 and 2006. Rates were quantified in terms of number of pellets, pigment, carbon and nitrogen for a wide size range of both aggregate and solitary salps. Measured defecation rates were constant over several hours when salps were held at near-surface conditions from which they had been collected. The defecation rate per salp increased with both salp size and the ambient level of particulate organic matter (POM) in the upper water column. The weight-specific defecation rate ranged between 0.5 and 6% day⁻¹ of salp body carbon, depending on the concentration of available particulate matter in the water. Carbon defecation rates were applied to biomass estimates of S. thompsoni to calculate daily carbon defecation rates for the populations sampled during the two cruises. Dense salp populations of over 400 mg C m⁻² were calculated to produce about 20 mg C m⁻² day⁻¹, comparable to other major sources of vertical flux of organic material in the Southern Ocean. Measured sinking rates for salp fecal pellets indicated that the majority of this organic material could reach deep sediments within a few days, providing a fast and direct pathway for carbon to the deep ocean.     

Sinking rates of fecal pellets from gelatinous zooplankton (Salps,Pteropods, Doliolids)

Sinking rates were determined for fecal pellets produced by gelatinous zooplankton (salps, Salpa fusiformis and Pegea socia; pteropods, Corolla spectabilis; and doliolids, Dolioletta gegenbaurii) feeding in surface waters of the California Current. Pellets from the salps and pteropods sank at rates up to 2 700 and 1 800 m d^-1, respectively; such speeds exceed any yet recorded for zooplankton fecal pellets. Fecal pellets of salps were rich in organic material, with C:N ratios from 5.4 to 6.2, close to values for living plankton. The relation between volume and sinking rate indicates that salp and pteropod pellets are slightly less dense than those of pelagic Crustacea; moreover, pellet density varied between different collection dates, probably because of differences in composition. In contrast, doliolid pellets sank at rates up to 208 m d^-1, a rate much lower than would be expected from pellet size. Thus, density and sinking rates of pellets are much more variable in zooplankton than would be expected from studies of crustaceans alone. Moreover, the extraordinarily high sinking rates of fecal pellets of salps indicates that these tunicates may be disproportionately important in the flux of biogenic materials during periods when they form dense population blooms.     

Sinking rates of natural copepod fecal pellets

Many pure samples of natural fecal pellets have been collected from mixed small copepods and from the pontellid copepod Anomalocera patersoni in the Ligurian Sea, using a specially designed pellet collection device. Sinking rates of fresh pellets and pellets aged up to 33 days have been determined at 14°C, the mean temperature of the essentially isothermal water column in the Ligurian Sea. Sinking rates of pellets collected during calm sea states increased with increasing pellet volume, but sinking rates of pellets collected during rough sea (Beaufort scale 6) showed little correlation with pellet size. Much of the variability in the sinking rate-pellet size relationships was the result of different pellet composition and compaction, but not pellet age. Pellets produced from laboratory diets of phytoplankton and phytoplankton-sediment mixes showed the expected wide variability in sinking rates, with sediment-ballasted pellets sinking much faster than pellets produced from pure algal diets; thus determination of vertical material fluxes in the sea using laboratory-derived fecal pellet sinking rates is unwarranted. Natural pellet sinking data for small copepods and A. patersoni have been combined with similar data for euphausiids, to yield sinking rates of roughly two orders of magnitude over three orders of magnitude in pellet volume. Pellets from small copepods sank at speeds too slow to be of much consequence to rapid material flux to the deep sea, but they undoubtedly help determine upper water distribution of materials. Recalculation of fecal pellet mass flux estimates from the literature, using our sinking rate data for natural small copepod pellets, yielded estimates about half those of previously published values. Earlier studies had concluded that small fecal pellets were of lesser significance to total material flux than fecal matter; our recalculation strengthens that conclusion. Pellets from large copepods and euphausiids, however, have the capability to transport materials to great depths, and probably do not substantially recycle materials near the surface. The fact that the majority of pellets which had previously been collected in deep traps by other workers were of a size comparable to pellets from our large copepods supports the contention that these larger pellets are the main ones involved in vertical flux.     

Differential feeding and fecal pellet composition of salps and pteropods,and the possible origin of the deep-water flora and olive-green “Cells”

Salps (mainly Salpa fusiformis and, to a lesser extent, Pegea socia) and a web-building pteropod (Corolla spectabilis) were studied in epipelagic waters of the central California Current. Although both kinds of gelatinous zooplankton trap phytoplankton in a mucus net, a fecal pellet analysis indicated that their diet differs significantly when they feed together, probably because of differences both in the pore sizes of their nets and in their feeding methods. Salps have a finemesh filter, on which they can retain even the smallest phytoplankton; thus, when small coccolithophores are abundant, as they were in our study, salp feces contain such cells and the coccoliths derived from them. In contrast, pteropods feeding in the same area produce fecal pellets consisting chiefly of larger phytoplankton, especially diatoms. Since fecal pellets transport most biogenic material to the deep sea, changes in herbivore species composition at a given geographic location can change the chemistry of materials entering deep water; at our study site, the more salps, the greater the calcite flux, and, the more pteropods, the greater the silica flux. In addition, fecal pellets of both salps and pteropods include partially digested residues of phytoplankton that appear as olive-green spheres, having an ultrastructure identical with that of the socalled olive-green cells. Presumably, fecal pellets, after sinking into deep water, ultimately disintegrate. releasing both the viable phytoplankton and the olive-green spheres into aphotic waters. Thus the feces of epipelagic herbivores are likely sources of much of the flora of the deep ocean.     

Trophodynamic function of copepods,appendicularians and protozooplankton in the late summer zooplankton community in the Skagerrak

The study was carried out in the Skagerrak during late summer when population development in the pelagic cycle culminated in the yearly maximum in zooplankton biomass. The cyclonic circulation of surface water masses created the characteristic dome-shaped pycnocline across the Skagerrak. The large dinoflagellate Ceratium furca dominated the phytoplankton biomass. Ciliates and heterotrophic dinoflagellates were the major grazers and, potentially, consumed 43–166% of daily primary production. The grazing impact of copepods was estimated from specific egg production rates and grazing experiments. The degree of herbivory differed between species (14–85%), but coprophagy (e.g. feeding on fecal pellets) and ingestion of microzooplankton were also important. The appendicularian Oikopleura dioica was present in lower numbers than copepods, but cleared a large volume of water. The grazing impact of copepods and O. dioica was estimated to 57±24% and 12±12% of daily primary production, respectively. Sedimentation of organic material (30 m) varied between 169 and 708 mg C m^–2 day^–1, and the contribution from the mesozooplankton (copepod fecal pellets and mucus houses with attached phytodetritus of O. dioica) was 5–33% of this sedimentation. Recycling of fecal pellets and mucus houses in the euphotic zone was 59% and 36%, respectively. However, there was a high respiration of organic material by microorganisms in the mid-water column, and 34% of the sedimenting material actually reached the benthic community in the deep, central part of the Skagerrak.     

Sinking rates and dissolution of midwater fish fecal matter

The sinking rates of fecal matter from 7 southern California midwater fish species were investigated. Feces were obtained from 162 specimens of Stenobrachius leucopsarus, Triphoturus mexicanus, Leuroglossus stilbius, Lampanyctus ritteri, Argyropelecus affinis and Parvilux ingens, which were collected in the Santa Barbara and San Clemente Basins between 1977 and 1979. In addition, feces obtained from 6 laboratory-maintained specimens of the midwater zoarcid Melanostigma pammelas were used for repeated sinking-rate measurements. The mean of the measured sinking rates for all species was 1.19 cm s^-1 (1 028 m d^-1), which is much higher than the known descent rates of euphausiid and copepod fecal pellets and of most other particulate organic detritus. Dissolution characteristics were also investigated for fecal matter from 4 species collected by the same series of net hauls: S. leucopsarus, T. mexicanus, A. affinis, and Sternoptyx obscura. The release of dissolved organic compounds from this material is low and does not represent a significant output during the relatively short time required to sink through the water column. These findings suggest that midwater fish fecal matter may represent a major source of organic transfer between the pelagic community and the benthos.     

Interactions of marine plankton with transuranic elements

In a series of laboratory experiments, the biokinetics of ²⁴¹Am, an important transuranium element, was studied in Meganyctiphanes norvegica, a euphausiid common in the northwestern Mediterranean. The euphausiids accumulated Am from water by passive adsorption onto exoskeletons, achieving wet weight concentration factors on the order of 10² after 1 wk exposure; concentration factors varied inversely with the size of the euphausiids and linearly with their surface area:wet weight ratios. Essentially all (96±10%) of the Am taken up from water was associated with the exoskeleton, so that negligible Am was retained by the euphausiids after molting. The retention half-time of Am in molts was 2.9 d. Euphausiids could also concentrate Am from feeding suspensions by ingesting Am-labelled diatom cells, although there was negligible Am assimilation (3±2% after 4 d feeding); after passage through the gut, virtually all (99%) of the ingested Am was defecated within 1 wk. The retention half-time of Am in fecal pellets was 41 and 51 d at 13° and 5°C, respectively. In oceanic waters, where the preponderance of ²⁴¹Am is in the dissolved phase, uptake of Am from water by euphausiids would be the dominant route of bioaccumulation. The results underscore the importance of sinking biogenic debris from zooplankters in mediating the vertical transport of Am in the sea. Given their retention half-times for ²⁴¹Am and their rapid sinking rates, fecal pellets and discarded molts have the potential to deliver most of their Am to the sediments.     

Origins and microenvironments of bacteria mediating fecal pellet decomposition in the sea

The fecal pellets of zooplankton are thought to be major carriers of organic matter from surface to deeper waters of the oceans. As the pellets descend, they release soluble components, partially due to breakdown by associated microorganisms. Previous laboratory work of other investigators has suggested that the surface of a fecal pellet rapidly acquires bacteria, which increase in abundance until they and their protozoan consumers disrupt the pellet membrane, spilling contents into the water. In contrast, our field collections of fecal pellets from free-floating particle interceptor traps (from the Vertex project off Central California in 1980 and off Mexico in 1981), suggest that microbial decomposition probably is initiated in the sea from inside the fecal pellets. Transmission and scanning electron microscopy indicate that bacterial populations are most abundant in the interior of fecal pellets obtained from the sea, but that the same pellets will acquire the surface bacterial lawn typically observed in laboratory studies if maintained aboard ship. If the fecal pellets are decomposed from the inside, then the principal agents are enteric bacteria or ingested, digestion-resistant bacteria, or both. Such bacteria may differ metabolically from those that colonize the fecal pellet surfaces. Further-more, the abundance of healthy-appearing bacteria inside the pellets suggests that their metabolic activities may produce microhabitats of reduced oxygen tension that could differ considerably from that of the pellet exteriors. Decomposition in these semi-enclosed microenvironments may proceed in a manner not yet predicted by models that attribute decomposition to well-aerated, surface-dwelling bacterial populations on fecal pellets in the sea.     

Flux of 141Ce through a euphausiid crustacean

The role of the Mediterranean euphausiid Meganyctiphanes norvegica in the cycling of radiocerium (¹⁴¹Ce) was examined. When uptake of ¹⁴¹Ce occurs directly from the water, a dynamic population equilibrium is reached at a concentration factor of about 250. Molting was responsible for up to 99% loss of total body burden at first molt, and about 45% of the remaining activity at second molt, thus denying true longterm equilibrium to individual animals. Fecal pellets did not contain measureable ¹⁴¹Ce activity when the euphausiids accumulated the isotope from water, thus proving that surface adsorption was the key accumulating process from water. When radiocerium was taken in through ingestion of labelled Artemia, about 99% of the body burden was voided as fecal pellets. Excretion by this route was accelerated when euphausiids were fed non-radioactive Artemia during loss phase. Radioactive counts of the pellets confirmed that all ingested ¹⁴¹Ce was lost through defecation. When ¹⁴¹Ce was ingested as labelled phytoplankton, a substantial fraction of the total body burden occurred in the molts, which indicated that the phytoplankton lost ¹⁴¹Ce to the water and the radioactivity was subsequently adsorbed to outer surfaces of the euphausiids. Molts, fecal pellets, and freshly-killed euphausiids lost ¹⁴¹Ce to the water exponentially, the rates being similar to the exponential portions of the loss curves for live, non-molting individuals. It is suggested that M. norvegica, and probably other pelagic zooplankters, can greatly accelerate radiocerium transport to the ocean floor by packaging the isotope as fecal pellets. In coastal areas subject to low-level radioactive waste disposal, ¹⁴¹Ce might be ionic (or at least soluble) to a great extent, in which case euphausiids could take up the isotope rapidly and accelerate its vertical transport via molting.     

Mass occurrence of Salpa fusiformis in the spring of 1984 off Ireland: implications for sedimentation processes

Zooplankton species composition and biomass were investigated during the spring of 1984 in three areas west of Ireland. In general, biomass of the gelatinous zooplankters [Salpa fusiformis (Cuvier) forma gregata and solitaria, Cymbulia sp., Euclio sp.; max. 360 mg Cm^-3] exceeded that of other zooplankton namely copepods (max. 70 mg C m^-3). Feeding by salps in the upper layers of all areas during the observed diatom spring bloom resulted in sedimentation of diatom-rich salp fecal pellets. This process ended the diatom spring bloom prior to nutrient depletion in surface waters and, thus, prior to mass sedimentation of algal cells.Publication No. 17 of the SFB 313 at Kiel University     

Ecological significance of salp fecal pellets collected by sediment traps in the eastern North Pacific

Five sediment traps were moored at depths of 740, 940, 1 440, 3 440 and 4 240 m for 7 d in December 1982 at Station 5 in the eastern North Pacific about 400 km from San Francisco. Dark green sinking particles enclosed in tough membrane consisted of mostly coccolithophores with some diatoms, dinoflagellates and chrysophytes. The average size of the particles was 10x5x2 mm. These characteristics indicate that the particles were fecal pellets of salp inhabiting the surface waters. Vertical fluxes of the organic carbon and nitrogen through sinking of the salp fecal pellets ranged from 6.7 to 23 mgC m^-2 d^-1 and from 0.88 to 3.2 mgN m^-2 d^-1, respectively. These values were several times higher than those determined in other oceanic areas by sediment trap experiments. Hydrocarbons consisting of short-chain n-alkanes (n-C₁₅-C₂₀) with n-C₁₇, the most predominant component, heneicosa-hexaene (n-C_21:6), br-C₂₅ alkenes and long-chain n-alkanes (n-C₂₁-C₃₀), without any odd or even carbon number predominance, were found from five depths. The presence of short-chain n-alkanes and n-C_21:6 indicated that phytoplankton in the surface waters was a primary source of organic matter in the sinking particles. Two isomers of br-C_25:3 and br-C_25:4 alkenes found in these particles also indicated that br-C₂₅ alkenes were the important biological marker of fecal pellet of zooplankton. The distribution pattern of the long-chain n-alkanes suggested that the sinking particles may be affected by bacteria to some extent. Fatty acids of the sinking particles were separated into free, triglyceride and wax ester fractions consisting of mono- and poly-unsaturated, and saturated fatty acids, with a range from C₁₄ to C₃₀. Concentrations of mono- and poly-unsaturated fatty acids decreased more rapidly toward the deep than those of saturated fatty acids, which cause low ratios of mono- and poly-unsaturated fatty acids/saturated fatty acids. This indicates that unsaturated fatty acids were more rapidly decayed by marine microbes than saturated fatty acids in the deep water, despite the fact that a significant amount of unsaturated fatty acids still remained in the sinking particles collected from the deep waters. Our results revealed that the salp fecal pellet plays an important role in supplying foods to organisms in intermediate and deep seas.     

210Po and 210Pb in zooplankton fecal pellets

²¹⁰Po and ²¹⁰Pb concentrations in fecal pellets from the zooplanktonic euphausiid Meganyctiphanes norvegica are reported. The ²¹⁰Po:²¹⁰Pb activity ratio is 2.2±0.3, a value in good agreement with that found in suspended particulate matter in surface seawater. Estimates of ²¹⁰Po and ²¹⁰Pb removal times from the mixed layer by fecal pellets alone yield values which are of the same order of magnitude as the removal times for these nuclides by all routes. It is suggested that there is a high probability that zooplanktonic fecal pellets play a significant role in the removal of both these nuclides from the surface layers of the ocean.     

Rates of seasonal sediment reworking in Axiothella rubrocincta (Polychaeta: Maldanidae)

Sediment reworking rates of Axiothella rubrocincta (Johnson, 1901) (Polychaeta: Maldanidae) were measured in situ in Tomales Bay, California (USA), from August, 1969 through July, 1970. On the average, each adult worm (approximate fresh weight=1 g) reworks about 5 g dry sediment d^-1 at a mean temperature of 13.4°C and a mean salinity of 31.8.Reworking rates are positively correlated with temperature and salinity, and negatively with sediment organic carbon, sedimentation rates and grain size. An inverse correlation exists between sediment reworking rates (g dry sediment g^-1 wet wt of worm d^-1) and g wet weight of worm. Sediment parameters deseribing unworked sediments are not significantly different from those for fecal sediments. Although these data suggest this species to be a non-selective deposit-feeder, it is more likely that it is faculatively selective.     

Contribution of salps to carbon flux of marginal ice zone of the Lazarev Sea, southern ocean

In order to estimate the in situ grazing rates of Salpa thompsoni and their implications for the development of phytoplankton blooms and for the sequestration of biogenic carbon in the high Antarctic, a repeat-grid survey and drogue study were carried out in the Lazarev Sea during austral summer of 1994/1995 (December/January). Exceptionally high grazing rates were measured for S. thompsoni at the onset of a phytoplankton bloom (0.2 to 0.8 μg chlorophyll a l⁻¹) in December 1994, with up to ≃160 μg of plant pigments consumed by an individual salp of 7 to 10 cm length per day. Dense salp swarms extended throughout the marginal ice zone, consuming up to 108% of daily phytoplankton production and 21% of the total chlorophyll a stock. Due to the much faster sinking rates and higher carbon content of salp faecal pellets, the efficiency of downward carbon flux through salps is much higher than through the other major grazers, krill and copepods. S. thompsoni can thus export large amounts of biogenic carbon from the euphotic zone to the deep ocean. With the observed ingestion rates during December 1994, this flux could have attained levels of up to 88 mg C m⁻² d⁻¹, accounting for the bulk of the vertical transport of carbon in the Lazarev Sea. However, in January 1995, when phytoplankton concentrations exceeded a threshold level of 1.0 to 1.5 μg chlorophyll a l⁻¹, salps experienced a drastic reduction in their feeding efficiency, possibly as a result of clogging of their filtering apparatus. This triggered a dramatic reversal in the relationship, during which a dense phytoplankton bloom developed in conjunction with the collapse of the salp population. Increases in the biomass and geographic range of the tunicate S. thompsoni have occurred in several areas of the southern ocean, often in parallel with a rise in sea-surface temperature during sub-decadal periods of warming anomalies. Received: 10 August 1997 / Accepted: 21 October 1997     

Analysis of phytoplankton pigments by HPLC before,during and after mass occurrence of the microflagellate Corymbellus aureus during the spring bloom in the open northern North Sea in 1983

Daily observations in April and May 1983 near a subsurface drifter launched in the Fladen Grounds area of the North Sea revealed that a large crop of the microflagellate Corymbellus aureus Green, producing up to 3 g C m^-2 d^-1, succeeded the diatoms dominating during the earlier phase of the spring bloom. The concentration of fucoxanthin was highest during the first half of May, while pigment fingerprints of high-performance liquid chromatograms of suspended matter sampled during the second half of May were dominated by 19-hexanoyloxyfucoxanthin, the main carotenoid of C. aureus (0.8x10^-9 mg cell^-1). This colonial Prymnesiophycean was described for the first time less than 10 years ago and has nerver been reported to be present in the North Sea. After 16 May the C. aureus population rapidly lost its growth potential, but the peak of the bloom in terms of cell number was found several days later (up to 9x10⁶ cells per dm³ on 19 May). The population then started to decline; senescence was associated with decreasing pigment contents of cells. The low growth rate of copepods registered during the second half of May was probably related to the poor quality of C. aureus as copepod food: the concentration of a phaeophorbide a typically found in copepod fecal pellets was highest during the diatom phase of the spring bloom preceding the C. aureus bloom.     

Biological production of nitrite in seawater

The relative importance of 3 different sources for biological production of nitrite in seawater was studied. Decomposition of fecal pellets of the copepod Calanus helgolandicus (at a concentration of approximately 12 g-at N/l), in seawater medium, released small amounts of ammonia over a 6 week period. It nitrifying bacteria were added to the fecal pellets nitrite was barely detectable over the same period. Decomposition of phytoplankton (present at a concentration of about 8 g-at particulate plant N/l) with added heterotrophic bacteria, released moderate amounts of ammonia over a 12 week period. If the ammonia-oxidizing bacterium Nitrosocystis oceanus was added to the decomposing algae, nitrite was produced at a rate of 0.2 g-at N/l/week. Heterotrophic nitrification was not observed when 7 open-ocean bacteria were tested for their ability to oxidize ammonia. The diatom Skeletonema costatum, either non-starved or starved of nitrogen, produced nitrite when growing with 150 or 50 g-at NO ₂ ^- -N/l at a light intensity of about 0.01 ly/min. When nitrate in the medium was exhausted, S. costatum assimilated nitrite. If starved of vitamin B₁₂, both non-N-starved and N-starved cells of S. costatum produced nitrite in the medium with 150 g-at NO ₃ ^- -N/l. Nitrate was not exhausted and cell densities reached 2x10⁵/ml due to vitamin B₁₂ deficiency. If light intensity was reduced to 0.003 ly/min under otherwise similar conditions, cells did not grow due to insufficient light, and nitrite was not produced. In the sea, it appears that, in certain micro-environments, decomposition of particulate matter releases ammonia with its subsequent oxidation to nitrite. The amounts of these nutrients and the rate at which they are produced are dependent upon the nature of the materials undergoing decomposition and the associated bacteria. In certain other areas of the sea, where phytoplankton standing stock is high and nitrate is non-limiting, excretion by these organisms is a major source of nitrite.     

Diet of anguilloid larvae: leptocephali feed selectively on larvacean houses and fecal pellets

Gut contents of 234 leptocephali comprising eight species of eels were examined from five families (Congridae, Muraenidae, Muraenesocidae, Nettastomatidae, and Ophichthidae). The larvae belonged to developing leptocephalus (215 specimens) and the early metamorphic stage (19 specimens). Visible gut contents were recognized in 111 individuals among the eight species, regardless of developmental stage. Two kinds of visible objects, transparent (0.4 to 1.2 mm) and opaque (20 to 380 m), were found in the gut of leptocephali. From their morphological characteristics, the former and the latter were identified as larvacean houses and zooplankton fecal pellets, respectively. Furthermore, most fecal pellets in the gut were identified as oikopleurid larvacean's fecal pellets. No trace of the many other phytoplankton or zooplankton, which were found with leptocephali in the ambient waters and could be suitable-size food, was found in the gut of any leptocephalus. On the basis of the importance of larvecean houses in the diet of several species of leptocephalus larvae, it is proposed that the peculiar, large, fang-like teeth of leptocephali are used for feeding, and evolved to pierce and grasp the mucous houses of larvaceans.     

Photoadaption in marine phytoplankton: Response of the photosynthetic unit

Some species of phytoplankton adapt to low light intensities by increasing the size of the photosynthetic unit (PSU), which is the ratio of light-harvesting pigments to P700 (reaction-center chlorophyll of Photosystem I). PSU size was determined for 7 species of marine phytoplankton grown at 2 light intensities: high (300 E m^-2 s^-1) and low (4 E m^-2 s^-1); PSU size was also determined for 3 species grown at only high light intensity. PSU size varied among species grown at high light from 380 for Dunaliella euchlora to 915 for Chaetoceros danicus. For most species grown at low light intensity, PSU size increased, while the percentage increase varied among species from 13 to 130%. No change in PSU size was observed for D. euchlora. Photosynthetic efficiency per chlorophyll a (determined from the initial slope of a curve relating photosynthetic rate to light intensity) varied inversely with PSU size. In contrast, photosynthetic efficiency per P700 was enhanced at larger PSU sizes. Therefore, phytoplankton species with intrinsically large PSU sizes probably respond more readily to the rapid fluctuations in light intensity that such organisms experience in the mixed layer.Contribution No. 1180 from the Department of Oceanography, University of Washington, Seattle, Washington, USA     

Grazing of Acartia hudsonica (A. clausi) on Skeletonema costatum in Narragansett Bay (USA): Influence of food concentration and temperature

Grazing experiments were performed with temperatureacclimated Acartia hudsonica fed the diatom Skeletonema costatum in concentrations ranging from 50 to 3×10⁴ cell ml^-1 at 5°, 10° and 15°C. The ingestion data were best fit by an Ivlev equation. Feeding threshold values of 39 and 59 cells ml^-1 were not significantly different from zero; however, filtration rates were depressed at low food concentrations. Maximum filtration rates increased exponentially with temperature, reaching a maximum with copepods collected at 14°–15°C, and then declining. Both the increase in ingestion rate with increasing food concentration and the maximum ingestion rate were significantly greater as experimental temperature was increased. Maximum ingestion rates were reached at concentrations greater than 6×10³ cells ml^-1. Percent of body carbon ingested per day at 5 g C L^-1 increased from 1.5% at 5°C to 6.7% at 15°C. At 500 g C L^-1, the ingestion increased from 84% (5°C) to 660% (15°C). Percent of body nitrogen at 0.5 g N L^-1 increased from 0.6% per day at 5°C to 2.5% per day at 15°C. At 50 g N L^-1, the ingestion was 42% body nitrogen at 5°C and 250% at 15°C. The influence of grazing by A. hudsonica on phytoplankton in Narragansett Bay, USA was estimated for 1972–1977. The percent of standing stock removed by grazing rarely exceeded 5% per day except during the late spring when S. costatum growth becomes nutrient limited and higher temperatures favor the rapid population growth of A. hudsonica.     

Role of salps in the flux of organic matter to the bottom of the Ligurian Sea

Sediment-trap samples were collected during and after the occurrence of a salp (Salpa fusiformis) bloom in the Bay of Villefranche, Mediterranean Sea, in April/May 1985. Large amounts of organic aggregates and faecal pellets were collected during the bloom. The aggregates were rich in carbohydrates and mineral grains and had similar rates of sedimentation (900 to 2 100 m d^-1) to those of the faecal pellets (1 000 to 2 000 m d^-1). The results of mineralogical and organic chemical analyses indicate the potential effect of these mucus-rich aggregates on local biogeochemical fluxes.