Geert-Jan A. Brummer

Vigorous exchange between the Indian and Atlantic oceans at the end of the past five glacial periods

The magnitude of heat and salt transfer between the Indian and Atlantic oceans through ‘Agulhas leakage’ is considered important for balancing the global thermohaline circulation. Increases or reductions of this leakage lead to strengthening or weakening of the Atlantic meridional overturning and associated variation of North Atlantic Deep Water formation. Here we show that modern Agulhas waters, which migrate into the south Atlantic Ocean in the form of an Agulhas ring, contain a characteristic assemblage of planktic foraminifera. We use this assemblage as a modern analogue to investigate the Agulhas leakage history over the past 550,000 years from a sediment record in the Cape basin. Our reconstruction indicates that Indian–Atlantic water exchange was highly variable: enhanced during present and past interglacials and largely reduced during glacial intervals. Coherent variability of Agulhas leakage with northern summer insolation suggests a teleconnection to the monsoon system. The onset of increased Agulhas leakage during late glacial conditions took place when glacial ice volume was maximal, suggesting a crucial role for Agulhas leakage in glacial terminations, timing of interhemispheric climate change and the resulting resumption of the Atlantic meridional overturning circulation.

The effect of upwelling on the distribution and stable isotope composition of Globigerina bulloides and Globigrinoides ruber (planktic foraminifera) in modern surface waters of the NW Arabian Sea.

Abstract Hydrographic changes in the NW Arabian Sea are mainly controlled by the monsoon system. This results in a strong seasonal and vertical gradient in surface water properties, such as temperature, nutrients, carbonate chemistry and the isotopic composition of dissolved inorganic carbon ( δ 13 C DIC ). Living specimens of the planktic foraminifer species Globigerina bulloides and Globigerinoides ruber , were collected using depth stratified plankton tows during the SW monsoon upwelling period in August 1992 and the NE monsoon non-upwelling period in March 1993. We compare their distribution and the stable isotope composition to the seawater properties of the two contrasting monsoon seasons. The oxygen isotope composition of the shells ( δ 18 O shell ) and vertical shell concentration profiles indicate that the depth habitat for both species is shallower during upwelling (SW monsoon period) than during non-upwelling (NE monsoon period). The calcification temperatures suggest that most of the calcite is precipitated at a depth level just below the deep chlorophyll maximum (DCM), however above the main thermocline. Consequently, the average calcification temperature of G. ruber and G. bulloides is lower than the sea surface temperature by 1.7±0.8 and 1.3±0.9 °C, respectively. The carbon isotope composition of the shells ( δ 13 C shell ) of both species differs from the in situ δ 13 C DIC found at the calcification depths of the specimens. The observed offset between the δ 13 C shell and the ambient δ 13 C DIC results from (1) metabolic/ontogenetic effects, (2) the carbonate chemistry of the seawater and, for symbiotic G. ruber , (3) the possible effect of symbionts or symbiont activity. Ontogenetic effects produce size trends in Δ δ 13 C shell–DIC and Δ δ 18 O shell–w : large shells of G. bulloides (250–355μm) are 0.33‰ ( δ 13 C) and 0.23‰ ( δ 18 O) higher compared to smaller ones (150–250 μm). For G. ruber , this is 0.39‰ ( δ 13 C) and 0.17‰ ( δ 18 O). Our field study shows that the δ 13 C shell decreases as a result of lower δ 13 C DIC values in upwelled waters, while the effects of the carbonate system and/or temperature act in an opposite direction and increase the δ 13 C shell as a result lower [CO 3 2− ] (or pH) values and/or lower temperature. The Δ δ 13 C shell–DIC [CO 3 2− ] slopes from our field data are close to those reported literature from laboratory culture experiments. Since seawater carbonate chemistry affects the δ 13 C shell in an opposite sense, and often with a larger magnitude, than the change related to productivity (i.e. δ 13 C DIC ), higher δ 13 C shell values may be expected during periods of upwelling.

Coccolithophore export production in response to monsoonal upwelling off Somalia (northwestern Indian Ocean)

Abstract Coccolithophores, collected in a sediment trap at 1032 m depth off Somalia from June 1992 to February 1993, responded rapidly to the SW and NE Monsoons with changes in productivity, species composition, coccolith:coccosphere ratio, and CaCO3 production. Statistically identified species clusters and sample clusters consistently matched hydrographic periods distinguished by sea-surface temperature, wind velocities and nutrient concentrations, as well as δ15N values in the sediment-trap samples. The deep-photic zone species Florisphaera profunda and Gladiolithus flabellatus were dominant in the early SW Monsoon when coastally upwelled water carried by a large gyre, passed over the station. The opportunistic species Emiliania huxleyi and Gephyrocapsa oceanica dominated during the height of the SW Monsoon and exhibited maximum fluxes during the beginning of upwelling relaxation in September. Most other coccolithophore species showed maximum relative abundances in a later stage of the upwelling relaxation when coccolithophore fluxes started to decrease. In the inter-Monsoon period stratified, nutrient-depleted, surface waters coincided with low coccolithophore fluxes. Several less abundant, oligotrophic species exhibited maximum percentages in this period. During the NE Monsoon both deep and shallow living species displayed a rapid increase in fluxes in response to a moderate wind-induced entrainment of nutrients. During this period maximum fluxes were recorded, with 1.1×10 9 coccolithsm −2 d −1 and 4.8×10 6 coccospheresm −2 d −1 . Average, time-weighted, total fluxes were 4.7×108 coccolithsm −2 d −1 and 12.2×105 coccospheresm −2 d −1 , corresponding to a (computed) coccolithophore–CaCO3 flux of 8.1 mgm −2 d −1 (=2.9 gm −2 yr −1 ) , which is 13.2% (13.5% including Thoracosphaera spp.) of the yearly total CaCO3 flux. The average CaCO3 flux in the fine fraction ( μm ) was 21.4 mg / m −2 / day −1 (=7.8 gm −2 yr −1 ) , contributing 34.3% to the total CaCO3 flux on a yearly basis.

(Palaeo-) ecological significance, transport and preservation of organic-walled dinoflagellate cysts in the Somali Basin, NW Arabian Sea

To date, relatively little information is available about factors influencing organic-walled cyst production of tropical dinoflagellates and processes influencing the final burial of cysts in bottom sediments, such as transport and preservation. To extend this information, cyst fluxes were documented for three sediment traps from June 1992 to February 1993 at two sites in the Somali Basin (northwestern Arabian Sea) as well as the cyst association of underlying sediments. By comparing cyst associations of contemporaneously collected trap samples at different depths at one site, information about transport and processes of decay in the water column was obtained. Neither transport nor decay appears to have any detectable influence on cyst association during cyst settlement through the water column. Comparing the trap associations with the underlying sediments indicates that downslope transport appears to have influenced the cyst association on a local scale only. Species-selective decay, probably related to the presence of oxygen in bottom sediments, has influenced the cyst association most pronouncedly at the most offshore site. Relating variations in the trap associations with environmental conditions of the overlying surface waters indicates that highest production of both filled and empty cysts occurs during the SW Monsoon upwelling. Based on this correlation three groups of species can be distinguished: Species with highest fluxes during (1) the first-half of the SW Monsoon (June-August); Bitectatodinium spongium, Echinidinium granulatum, Echindinium transparantum, Echinidnium spp., cysts of Protoperidinium compressum and cysts of Protoperidinium subinerme, (2) the transition between the SW-Monsoon and inter-Monsoon; Spiniferites mirabilis and Spiniferites spp., (3) no particular season; all other species. Cyst associations of all trap samples are dominated by cyst of Protoperidinium species. Cysts with highest fluxes during the SW-Monsoon form about a third of the associations.

Seasonal stratification, shell flux, and oxygen isotope dynamics of left-coiling N. pachyderma and T. quinqueloba in the western subpolar North Atlantic

blooms. This annual population overturning is marked by a 0.7‰ increase in d 18 O in both species. The shell flux of N. pachyderma peaks during the spring bloom and in late summer, when stratification is close to its minimum and maximum, respectively. Both export periods contribute about equally and account for >95% of the total annual flux. Shell fluxes of T. quinqueloba show only a single broad pulse in summer, thus following the seasonal stratification cycle. The d 18 Oo fN. pachyderma reflects temperatures just below the base of the seasonal SML without offset from isotopic equilibrium. The d 18 Op attern ofT. quinqueloba shows a nearly identical amplitude and correlates highly with the d 18 Oo fN. pachyderma. Therefore T. quinqueloba also reflects temperature near the base of the SML but with a positive offset from isotopic equilibrium. These offsets contrast with observations elsewhere and suggest a variable offset from equilibrium calcification for both species. In the Irminger Sea the species consistently show a contrast in their flux timings. Their flux‐weighted Dd 18 O will thus dominantly be determined by seasonal temperature differences at the base of the SML rather than by differences in their depth habitat. Consequently, their sedimentary Dd 18 O may be used to infer the seasonal contrast in temperature at the base of the SML. Citation: Jonkers, L., G.‐J. A. Brummer, F. J. C. Peeters, H. M. van Aken, and M. F. De Jong (2010), Seasonal stratification, shell flux, and oxygen isotope dynamics of left‐coiling N. pachyderma and T. quinqueloba in the western subpolar North Atlantic, Paleoceanography, 25, PA2204, doi:10.1029/2009PA001849.

Quantifying carbonate dissolution and calibration of foraminiferal dissolution indices in the Somali Basin.

Two sediment traps moored off Somalia in 1992–1993 collected similar settling fluxes of carbonate and siliceous shells formed by various plankton groups. Planktic foraminifera showed large seasonal variations, with more than 74% of the total planktic foraminifera flux collected during the SW monsoon (summer upwelling), when Globigerina bulloides was dominating along with Globigerinita glutinata and Neogloboquadrina dutertrei. The intermonsoon and NE monsoon assemblages were dominated by Globigerinoides ruber. We used the trap records as our ‘no dissolution’ reference for comparison with three boxcore recorders in order to quantify the carbonate dissolution along a depth transect. Dissolution increases downslope, from Station 905 to 907 and 915 at depths of 1567 m, 2807 m and 4035 m, respectively. The carbonate fraction of the sediment at Station 915, which is located near the CCD, is the most affected by dissolution, with more than 97% of the planktic foraminifera dissolved. Here, the planktic foraminifera assemblage is strongly modified, with thick walled species such as N. dutertrei, Globorotalia tumida and Pulleniatina obliquiloculata as the most resistant. It is not representative of the settling assemblage. The planktic foraminiferal assemblages of the sediment surface at Stations 905 and 907 remain similar to the trap assemblages and the foraminifera are well preserved, although only 25% of foraminifera are apparently preserved at Station 905 and 8% at Station 907. Those numbers are surprisingly low and infer that only a small fraction of the foraminiferal carbonate production is buried and removed from the carbon cycle. This discrepancy between the export and buried flux is partly be due to (bio)mechanical destruction by benthic processes and to supralysoclinal dissolution, due to metabolic CO2 generated by the benthic organisms. Another important factor is the interannual variability of the productivity, well known in the Arabian Sea. The calibration of commonly used foraminiferal dissolution indices (percentage of foraminiferal fragments, percentage of resistant species, foraminiferal dissolution index (FDX)) to our data only shows reliable results for high dissolution levels (>97%). Off Somalia the most accurate of the proxies is the percentage of foraminiferal fragments compared to the other methods tested, i.e., FDX, planktic foraminiferal loss (L), percentage of radiolarians and diatoms, percentage of benthic foraminifera. The species assemblage appears to be not significantly modified by dissolution unless the estimated shell loss is high, >92% of the arriving shells in our samples. This level is expressed in the percentage of fragmentation, the percentage of radiolarians and diatoms, the percentage of resistant foraminifera species, and FDX as 80%, 35%, 25% and 1.8, respectively. The relative abundance of Globigerina bulloides is a valid SW monsoon/upwelling proxy only when dissolution is moderate (<92%). Globigerina bulloides and Globigerinoides ruber have similar burial efficiencies and susceptibilities to dissolution. Thus, the ratio G. bulloides/G. ruber is a valid proxy for past changes in the intensity of the SW monsoon even in the strongly dissolved samples. In our sediment record the ratio G. bulloides/G. ruber indicates that the SW monsoon was stronger in the recent past than in 1992–93.

Settling, dissolution and burial of biogenic silica in the sediments off Somalia (northwestern Indian Ocean)

Abstract Particle fluxes of biogenic silica through the water column, silica burial fluxes into the sediments, and the flux of dissolved silica across the sediment-water interface estimated from pore water profiles are used to assess the behaviour of biogenic silica at two stations 80 and 270 km offshore along a transect off the Somali coast in the northwestern Indian Ocean. Particulate biogenic silica fluxes varied from 0.3 mmol m−2 day−1 in the non-upwelling season to 6 mmol m−2 day−1 during upwelling on the Somali slope. Fluxes were lower in the Somali Basin, from 0.2 to 2.3 mmol m−2 day−1. Evaluation of the dissolution curves derived by wet chemical leaching in sediment trap and sediment samples shows that the Km values, the apparent reactivity rates in alkaline medium, are higher for the shallow sediment traps than for deep trap and boxcore sediments. Modelling of pore water profiles shows that in the sediment most dissolution occurs in the top halfcentimetre, and pore water effluxes are in close agreement with those from in situ benthic incubations. Our results show that less than 10% of the biogenic silica arriving on the Somali Margin is buried in the sediments, giving a burial efficiency lower than the approximately 20% reported from the open Arabian Sea.

A size analysis of planktic foraminifera from the Arabian Sea

Abstract Planktic foraminiferal faunas from different environments in the Arabian Sea were size fractionated using 14 sieves with meshes between 100 and 710 μm, to assess the effect of the sieve mesh size cut off level on the faunal composition and to determine the size frequency distribution of individual species. Nine samples from a plankton pump and a towed net, a sediment trap, a box-core and a piston core were selected, to cover living and settling flux faunas as well as fossil faunas from the sediment. In living faunas, most species show an exponential size frequency distribution, with highest numbers in the finest interval of the size spectrum. In sediment trap and core samples, individual species size frequency distributions may consist of: (1) an exponential distribution of relatively small pre-adult specimens; (2) a Gaussian-shaped distribution of larger specimens, which may be classified as adult or terminal; or (3) a combination of both. The distributions are separated using a best fit technique. The composition of the total planktic foraminiferal fauna strongly changes along the size spectrum. Dominant taxa in >355 μm fractions are Orbulina universa, Globorotalia menardii, Globorotalia tumida, Globigerinella siphonifera and Globigerinoides sacculifer, in 125–355 μm fractions Globigerina bulloides, Globigerinoides ruber, Neogloboquadrina dutertrei and Globigerinita glutinata, and in 100 μm) is present in the 100–125 μm fraction and 1–6% is larger than 250 μm. In samples representing a settling flux (sediment trap and sediment samples) 29–57% of the fauna is present in the 100–125 μm fraction, while 6–23% is larger than 250 μm. Size frequency distributions of the dextral Neogloboquadrina complex (= Neogloboquadrina dutertrei and Neogloboquadrina pachyderma + P–D intergrades) show a bimodal pattern; a smaller peak reflecting dextral Neogloboquadrina pachyderma, and a larger peak of adult Neogloboquadrina dutertrei. By applying a best fit technique to the data, the two species may be separated from each other. In size fractions larger than 150 μm most species have reached the adult stage of ontogeny and we recommend this mesh size for standard faunal analysis. In addition, sieve mesh sizes of 125 and 250 μm have to be used to obtain a reliable estimate of the abundance of small and large species, respectively.

LIVING PLANKTONIC FORAMINIFERA IN THE EASTERN SOUTH ATLANTIC DURING SPRING: INDICATORS OF WATER MASSES, UPWELLING AND THE CONGO (ZAIRE) RIVER PLUME

Abstract Planktonic foraminifera were collected from pumped surface-water samples and net tows over the upper 150 m of the water column in the eastern South Atlantic Ocean during October and November 1989. Of the 24 recorded species, 12 occur abundantly and form 6 foraminiferal assemblages. The assemblages correlate with the physico-chemical properties of the 6 vapor surface-water masses. The correlation of species with low salinity is due to properties co-variant with salinity. At several locations, high concentrations of planktonic foraminifera were observed caused by frontal mixing or reproduction. The G. sacculifer-G. ruber white assemblage marks the South Equatorial Under Current. minor contribution of G. menardii and N. dutertrei to this assemblage reflects the subsurface Equatorial Under Current. The N. dutertrei-G. siphonifera assemblage reflects the South Equatorial Counter Current and Equatorial Under Current. The N. pachyderma-G. bulloides assemblage reflects the coastal upwelling over the shelf and the filaments of mixed oceanic and upwelled water in the coastal of the Benguela Current (BC). The Angola-Benguela Front, at the interface of the warm currents and the BC, is reflected by the G. inflata-G. bulloides assemblage. An unusual coincidence of G. sacculifer-G. bulloides in one assemblage results from the interaction of the subtropical gyre and the oceanic branch of the BC. The G. rubber pink-N. dutertrei assemblage is attributed to the warm low-saline plume in front of the Congo (Zaire) River.

Intense mid-slope resuspension of particulate matter in the Faeroe–Shetland Channel: short-term deployment of near-bottom sediment traps

An array of four moorings was deployed on a transect perpendicular to the south-eastern slope of the Faeroe– Shetland Channel to measure near-bottom fluxes during a 12-day period in spring 1999. Each mooring combined current meters and two sediment traps equipped with optical backscatter sensors (OBS) situated at 2 and 30 m above the bottom, at water depths of 470, 700, 800 and 1000 m. During the deployment, near-bottom current velocities at 470, 700 and 800 m water depth increased abruptly to values as high as 55 cm s � 1 (within a few hours). The sudden change in the crossslope component of the current at 470 m was immediately followed by a severe drop in temperature at this depth, and large fluxes were intercepted in the mid-slope traps (700 and 800 m) on the same day, probably associated with the strengthening of the along-slope component. Organic carbon and nitrogen content of the trap samples were, except for the shallowest mooring, much lower than in particulate matter suspended in the water column. During the high flux event, particles possessed lower organic carbon and nitrogen values than during lower flux periods. This indicates that sedimentary material entered the traps. A simple multi-component mixing model was applied to the trap samples to estimate the relative contribution to our total mass fluxes of material from different sources: primary settling from the water surface, rebound material, and the relatively aged sediment. It showed that fluff was the main contributor to the resuspended flux but that sediment proper was resuspended at mid-slope during the strong increase in the current velocities and represented up to 70% of the total mass flux at that time. This study shows that near-bottom resuspension on the slope may be intense and, although the scale is here short-term the phenomena described might be larger in time and space. r 2002 Elsevier Science Ltd. All rights reserved.