Planktic foraminiferal production stimulated by chlorophyll redistribution and entrainment of...

*Corresponding author. Present address: Geological Institute, ETHZ, Sonnegg Strasse 5, 8092 Zurich, Switzerland.Tel.: 41-1-6323676; fax: 41-1-6321080.

E-mail address: [email protected] (R. Schiebel).

Deep-Sea Research I 48 (2001) 721}740

Planktic foraminiferal production stimulated by chlorophyllredistribution and entrainment of nutrients

Ralf Schiebel!,*, Joanna Waniek", Matthias Bork!, Christoph Hemleben!

!Institute and Museum of Geology and Paleontology, Tu( bingen University, Sigwartstrasse 10, 72076 Tu( bingen, Germany"Institute of Marine Science, Kiel University, Du( sternbrooker Weg 20, 24105 Kiel, Germany

Received 1 February 2000; received in revised form 31 May 2000; accepted 6 June 2000

Abstract

During September and October 1996 planktic foraminifers and pteropods were sampled from the upper2500 m of the water column in the BIOTRANS area (473N, 203W), eastern North Atlantic, as part of theJGOFS program. Hydrography, chlorophyll #uorescence, and nutrient content were recorded at high spatialand temporal resolution providing detailed information about the transition time between summer and fall.At the beginning of the cruise a shallow pycnocline was present and oligotrophic conditions prevailed. Overthe course of the cruise, the mixed layer depth increased and surface water temperature decreased by 1.53C.Both chlorophyll-a dispersed in the upper 50 m by vertical mixing and chlorophyll-a concentrations at thesea surface increased. The nitracline shoaled and nutrient enriched waters were entrained into the mixedlayer. Planktic foraminifers and pteropods closely re#ected the changes in the hydrography by increasedgrowth rates and changes in species composition. Three main groups of planktic foraminiferal species wererecognized: (1) a temperate and low-productivity group dominated by Neogloboquadrina incompta character-ized the shallow mixed layer depths. (2) A temperate and high-productivity group dominated by Globigerinabulloides characterized the period with wind-induced dispersal of chlorophyll-a and entrainment of nutrient-enriched waters. (3) A warm water group containing Globigerinoides sacculifer, Orbulina universa,Globigerinoides ruber (white), and Globigerinella siphonifera was most common during the "rst days ofsampling. Synchronous with the hydrographic change from summer to fall, planktic foraminiferal andpteropod growth was stimulated by redistribution of chlorophyll-a and entrainment of nutrient-enrichedwaters into the mixed layer. In addition, the seasonal change in the eastern North Atlantic resulted ina transition of the epipelagic faunal composition and an increased calcareous particle #ux, which could beused to trace seasonality in fossil assemblages and allow for better paleoceanographic interpretation of theboreal Atlantic. ( 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Planktic foraminifers; Turbulent entrainment; Chlorophylls; Micropaleontology; Seasons; Populationdynamics; Eastern North Atlantic; BIOTRANS; JGOFS 473N, 203W

0967-0637/01/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved.PII: S 0 9 6 7 - 0 6 3 7 ( 0 0 ) 0 0 0 6 5 - 0

1. Introduction

Plankton blooms during fall have been described for coastal zones and shelf areas (Drebes, 1974;Le Fevre, 1986), but few records exist of pelagic plankton blooms. Seasonal plankton blooms occuronly in regions with well-de"ned seasonality, and hence records of seasonal peaks in the vertical#ux of sinking matter are restricted to mid and high latitudes, e.g., the eastern North Paci"c(Thunell and Honjo, 1987). Honjo and Manganini (1993) describe a slight increase in the #ux ofplanktic foraminifers at 483N, 213W (BIOTRANS area), between the 248th and the 262nd day in1989, which corresponds to the second and third week of September. Data from Ocean WeatherShip India (OWS I, at 593N, 193W, 1971}1975) show high chlorophyll-a values between 0.5 and1 mg m~3 during September and October, which are similar to the concentrations observed inspring (1990) at BIOTRANS (Barlow et al., 1993). However, high chlorophyll-a values in spring areoften accompanied by relatively low primary production (Williams and Robinson, 1971; Williamsand Wallace, 1975). Additionally, continuous plankton monitoring in the northeast Atlantic from1948 to 1995 indicates higher phytoplankton abundances during fall than spring for several years(Reid et al., 1998). Analysis of coastal zone color scanner data (CZCS) indicates a fall bloom in theNorth Atlantic between September and the beginning of November (Obata et al., 1996). Inaddition, observations from sediment traps have shown an enhanced #ux during fall (Newton et al.,1994).

The aim of this study is to describe the population development of planktic foraminifers andpteropods, and to investigate the reason for increased plankton productivity during fall. Extensivemeasurement of the hydrographic parameters by CTD and XBT was carried out, and nutrientconcentrations (NO

2, NO

3, PO

4, SiO

2) were recorded. Planktic foraminifers in particular re#ect

the upper ocean ecology on a "ne scale, and are main contributors to the pelagic calcareous particle#ux. To interpret vertical test #ux, it is crucial to understand planktic foraminiferal populationdynamics. Because of their wide distribution in the world oceans since the Cretaceous and the highfossilization potential of their tests, planktic foraminifers allow for reconstruction and interpreta-tion of marine paleoecology.

2. Materials and methods

During RV `Meteora cruise 36/5 in September and October 1996, we investigated the hydrogra-phy, chemistry, and biology of the eastern North Atlantic Ocean around 473N, 203W (Fig. 1).Hydrographic measurements were carried out in two parts, a pre-study and a main study. Thepre-study started 100 nautical miles southeast of 473N, 203W with expendable bathythermo-graphs (XBT 1}4) every 30 nautical miles, sampling with a surface pump system every 15 nauticalmiles, and CTD casts. After CTD recording and sampling at BIOTRANS, we sailed northwestto Station 300, turned south to Station 303, and then to the northeast (Station 307). Thepre-study ended with a north-to-south transect east of BIOTRANS "nishing at Station 310.The surface-pump system (SchuK ssler and Kremling, 1993) delivered underway samples from7 m water depth for "ltration (total CO

2, Chlorophyll-a, POC/PN, PSi, particulate CaCO

3,

HPLC, plankton samples), for determination of nutrients and salinity, and for investigation of thephytoplankton biomass distribution. Surface sampling, XBT drops, and CTD casts were carried

722 R. Schiebel et al. / Deep-Sea Research I 48 (2001) 721}740

Fig. 1. Location of the area of investigation and positions of the employed gear. The central BIOTRANS area wassampled at 47311@N, 19334@E, sites 1, 2, and 11. For detailed information about the sites see Table 1.

out to characterize the vertical structure of the hydrographic and biological "elds of the upperwater column.

Based on the pre-study, a grid of about 180]180 nautical miles around BIOTRANS wassampled between 17326@W to 22323@W and 47305@N to 48336@N (Fig. 1). The survey was a combina-tion of CTD casts, XBT drops, and multinet hauls, supported by continuous recording ofchlorophyll #uorescence. The mixed layer depth is calculated from CTD data (Fig. 2a). The netheat #ux at the ocean-atmosphere boundary is calculated from continuous ship based meteorologi-cal observations (air temperature, humidity, wind speed, water temperature) using standardformulae (Gill, 1982). In the absence of high quality measurements of incoming radiation duringthis cruise, the solar radiation received by the ocean was calculated from the formula of Hender-son-Sellers (1986) considering the cloud cover correction. Daily cloud cover values from theNational Center for Environmental Prediction (NCEP) re-analysis data (Kalnay et al., 1996) wereused for the calculation of longwave radiation.

Water samples were taken with a 24-bottle-rosette system equipped with Niskin bottles andmounted on a Neil Brown CTD system with an additional oxygen and #uorescence sensor. In total78 CTD casts and 77 XBT drops were made. Subsamples for nutrient analyses were taken from thebottles immediately after recovery of the rosette and processed within a few hours after samplingaccording to standard procedures (Grassho! et al., 1999). Chemical analyses of chlorophyll-a wereperformed according to Herbland et al. (1985). Continuous pro"les of chlorophyll-a #uorescenceand the continuous records at the sea surface were calibrated with discrete chlorophyll-a samples.The planktic foraminiferal and gastropod fauna of 172 multinet samples (100 lm mesh size) were

R. Schiebel et al. / Deep-Sea Research I 48 (2001) 721}740 723

Fig. 2. Spatial distribution (a) of the mixed layer depths calculated with *T"0.53C. CTD sites are marked (#). Mixedlayer depth was increasing (b) during the course of the cruise.

investigated (Table 1). Vertical multinet sampling depth intervals ranged from the surface down to100 m depth (0}20}40}60}80}100 m), down to 700 m (100}200}300}500}700 m), and down to2500 m (700}1000}1500}2000}2500 m). The samples were "xed aboard ship in 4% hexamethyltet-ramine bu!ered formaldehyde solution (pH 8.2) and processed in TuK bingen. Planktic foraminifersand gastropods were picked, dried, and separated into size classes (100 lm, 100}125}150}200}250}315 lm, and '315 lm, followed by splitting into aliquots when necessary. Testsfrom size fractions '125 lm were counted at the species level, and the average number of tests perm3 was calculated. Live foraminifers (cytoplasm-bearing) were distinguished from dead specimens(empty tests).

3. Results

The "eld data presented in this paper are available from the German JGOFS Project datamanagement at IfM Kiel (http://www.ifm.uni-kiel.de/pl/dataman/dmpag1.html).


Table 1Time, location, and water depth intervals of the RV Meteor 36/5 Multinet samples

d Date Lunarday

Site Latitude Longitude Multinet Water depth Timed N W d (m) (0}24 h)

1 10.09.96 13 298 47311.062@ 19334.032@ 1121 0}100 7:4647311.139@ 19334.123@ 1122 0}700 8:2947311.139@ 19334.123@ 1123 0}2500 10:31

2 13.09.96 16 305 47311.006@ 19333.947@ 1127 0}100 5:5247310.960@ 19333.957@ 1128 0}700 6:3447311.150@ 19333.463@ 1129 0}2500 10:52

3 14.09.96 17 307 48335.901@ 17327.486@ 1130 0}100 11:194 15.09.96 18 312 46305.792@ 18309.520@ 1132 0}100 16:525 16.09.96 19 314 47305.652@ 18309.905@ 1134 0}100 16:456 17.09.96 20 316 48305.761@ 18309.570@ 1136 0}100 3:18

48305.734@ 18309.608@ 1137 0}700 3:5448305.843@ 18309.687@ 1138 0}2500 6:10

7 18.09.96 21 320 47335.814@ 18352.136@ 1139 0}100 8:208 19.09.96 22 323 46305.687@ 18351.601@ 1143 0}100 13:099 21.09.96 24 326 48305.997@ 19334.059@ 1144 0}100 8:18

48305.951@ 19334.153@ 1145 0}700 8:5648305.994@ 19334.570@ 1146 0}2500 10:51

10 22.09.96 25 330 47335.903@ 20316.166@ 1147 0}100 6:1111 22.09.96 25 331 47310.951@ 19333.979@ 1149 0}100 18:05

47310.956@ 19333.950@ 1150 0}700 18:4347311.020@ 19333.892@ 1151 0}2500 20:54

12 24.09.96 27 337 46305.877@ 20358.298@ 1153 0}100 8:3046305.786@ 20358.152@ 1154 0}700 9:0846305.891@ 20358.306@ 1155 0}2500 9:30

13 25.09.96 28 339 47305.987@ 20358.401@ 1156 0}100 3:4947305.843@ 20358.309@ 1157 0}700 4:2747305.827@ 20357.998@ 1158 0}2500 6:31

14 27.09.96 1 342 48305.927@ 21340.207@ 1160 0}100 8:4615 27.09.96 1 343 47335.501@ 21340.587@ 1162 0}100 9:2416 27.09.96 1 344 47304.784@ 21340.588@ 1163 0}100 20:1217 28.09.96 2 345 46335.939@ 21340.555@ 1165 0}100 2:1518 29.09.96 3 349 46305.922@ 22322.789@ 1168 0}100 9:4419 30.09.96 4 354 48335.714@ 22323.050@ 1172 0}100 17:12

48335.752@ 22323.230@ 1173 0}700 17:5048335.819@ 22322.721@ 1174 0}2500 19:51

20 03.10.96 7 358 47305.584@ 17326.768@ 1179 0}100 9:0347305.494@ 17326.781@ 1180 0}700 9:4147305.944@ 17327.342@ 1181 0}2500 11:54

3.1. Temporal development of physical and biological properties

The mixed layer hydrography obtained at the beginning of the cruise indicated a well-mixed,homogeneous upper water layer and a stable strati"ed horizon with a strong temperature gradient


Fig. 3. Daily averaged net heat #ux QNET

(W m~2) during September and beginning of October, 1996, was calculatedfrom ship based meteorological observations using standard formulae (Gill, 1982). With the increasing mixed layer depth,a negative heat #ux was recorded, indicating decreasing surface water temperature.

Fig. 4. During the course of sampling nitrate concentration at the sea surface increased. Samples for nutrient analysiswere taken from a surface-pump system at 7 m depth.

below (seasonal signal). Mixed layer depths, calculated with the classic criterion of *T"0.53C(sensitive to strong gradients), ranged between 30 and 55 m (Fig. 2a). During the course of thecruise, intense winds were recorded along with changes in the net heat #ux and a strong mixing ofsurface waters (Fig. 3). The mixed layer increased successively in depth, while the stability of thestrati"cation of the upper water column decreased (Fig. 2b). During the "rst-half of the cruise themixed layer depths ranged between 20 and 45 m (cast 1}43). After storm events the mixed layerdepth increased to 35}65 m (casts 44}80).

Nutrients were depleted in the shallow surface layer at the beginning of the cruise, and asubsurface chlorophyll-a maximum was observed mainly at the depth of the nitracline. The nitra-cline is de"ned by nitrate concentrations higher than 0.1 lM. Average nitracline depths rangedbetween 20 and 50 m. Maximum nitracline depths, below 50}80 m, occurred in the northern part ofthe investigated area. Two storms caused shoaling of the nitracline to 10}20 m depth. Although thenitracline showed strong temporal variability in its depth, a general shoaling of the nitracline wasobserved throughout the course of the cruise (Zeitzschel et al., 1998). The NO

3concentration at the

sea surface increased (Fig. 4) while the SiO4

and PO4

values did not signi"cantly change over thesampling campaign, with the exception of short-term peaks following storms.


Fig. 6. Temperature pro"les (a) and chlorophyll-a distribution (b) in the upper 150 m on September 13 (47311.0@N,19333.8@W) and September 22 (47311.0@N, 19334.0@W), before and after a storm (17}18.9.1996; 9}11 Bft). Mixing increasedthe thermocline depth from about 30 to 50 m and dispersed chlorophyll from the deep chlorophyll maximum (DCM)over the entire upper water column.

Fig. 5. The chlorophyll-a concentration (mg m~3) at the sea surface along the cruise track (M36/5) increased stepwise.Continuous measurements of chlorophyll-a #uorescence at 7 m depth (solid line), and chlorophyll-a concentrationmeasured from the pump system (#) were sampled from the same depth.

Sea surface chlorophyll-a values tripled during the cruise from 0.2 mg m~3 to about 0.6 mg m~3

(Fig. 5). In contrast, the integrated chlorophyll-a content in the upper 90 m of the water column didnot change signi"cantly during the cruise, ranging between 10 and 50 mg m~2 (on average37 mg m~2). However, the integrated chlorophyll-a content in the upper 15 m compared to the50}80 m horizon clearly shows redistribution of phytoplankton in the upper water column. Thechlorophyll-a content increased between the sea surface and 15 m depth, and decreased between 50and 80 m during the course of sampling when chlorophyll-a was spread from a narrow subsurfacemaximum to a thick and more uniform horizon down to about 100 m depth (Figs. 6 and 7).


Fig. 7. Integrated chlorophyll-a (mg m~2) content (a) over the upper 90 m and (b) between the surface and 15 m depth(bold line) and over 50}80 m depth (thin line). An increase in chlorophyll-a concentration at the sea surface andsynchronous decrease between 50 and 80 m depth occurred over the course of the cruise.

The chlorophyll-a and nitrate concentrations at the sea surface were negatively correlated to seasurface temperature (Fig. 8). Concerning the correlation between chlorophyll-a and nitrate concen-tration, two groups of samples were distinguished. One group had chlorophyll-a values of0.2}0.8 mg m~3 and low nitrate concentrations ((0.2 lM), and the second group had less variablechlorophyll-a concentrations of 0.4}0.6 mg m~3 and higher nitrate concentrations ('0.3 lM).

3.2. Temporal and spatial development of the fauna

During RV Meteor cruise 36/5, the habitat of most live planktic foraminifers expanded from theupper 40 m to the upper 100 m. For the "rst days of sampling, from September 10 to 16, theplanktic foraminiferal standing stock was low and no signi"cant strati"cation was recorded withrespect to number of specimens (Figs. 9 and 10). On September 17, the number of specimens in theupper 40 m increased while they remained low between 40 and 60 m. Until September 25, the maindepth habitat of the fauna increased to 40}60 m depth. During this "rst phase of faunal develop-ment a stepwise increase in numbers from 10 up to 40 specimens m~3 was recorded (Fig. 9). Themost distinct changes in depth distribution were recorded from September 27 through 30, whenhigh numbers of planktic foraminifers occurred throughout the upper 100 m of the water column,and the highest total number of specimens was recorded towards the end of the cruise (Fig. 10).Below 100 m down to 2500 m, live specimens, in general, were rare ((1 spec. m~3).

The temporal development of the pteropod standing stock was similar to that of the plankticforaminifers, with relatively low numbers during the "rst-half of the sampling campaign. Twopeaks in abundance, September 22 and 28, preceded the peaks in the planktic foraminifers by a fewdays and were less pronounced (Fig. 9). Other gastropods did not show pronounced maxima, andaverage abundance was higher during the second half of the sampling campaign. Smallest numberswere recorded on September 14 and 15, and October 3, similar to planktic foraminifers andpteropods. On both September 22 and 27, large di!erences in planktic foraminiferal and gastropodnumbers (Fig. 9) occurred at di!erent locations (Table 1, Fig. 1).


Fig. 8. Correlation between (a) temperature (3C) and chlorophyll-a concentration (mg m~3), (b) temperature (3C) andnitrate (lM), and (c) chlorophyll-a (mg m~3) and nitrate (lM), at the sea surface. While temperature decreased,chlorophyll-a and nitrate concentrations increased.

Empty planktic foraminiferal tests were rare but showed a very distinct depth distribution(Fig. 10). During the "rst-half of the sampling campaign, until September 22, most empty testsoccurred below 2000 m depth. After September 27, the maximum number of tests was recordedbetween 200 and 1500 m depth.

3.3. The species succession

Neogloboquadrina incompta and Globigerina bulloides were frequent throughout the samplingcampaign. These species alternated in dominance from station to station, and each constituted upto '60% of the fauna (Fig. 11). Large numbers of N. incompta were recorded throughout theupper 100 m without showing signi"cant depth preference. In contrast, G. bulloides was mostfrequent in the upper 60 m during the "rst half of the cruise. During the last 6 days of sampling(September 27 to October 3), G. bulloides was most frequent between 60 and 100 m. Along with G.bulloides, Turborotalita quinqueloba and Globigerinita glutinata were frequent between 80 and100 m from the end of September to early October (Fig. 11). Globigerinoides ruber (white),Globigerinoides sacculifer, Globigerinella siphonifera, and Orbulina universa were abundant untilSeptember 25 and 28, and October 3 (Fig. 12), and Globorotalia inyata and Globorotalia scitula werescarce.


Fig. 9. Temporal development of the (a) planktic foraminiferal, (b) pteropod, and (c) other shelled gastropod standingstocks in the upper 100 m of the water column (see Table 1). A stepwise increase in planktic foraminiferal numbers onSeptember 17, 22, 27, and 29 was preceded by the pteropod faunal development. Other gastropods showed less distinctchanges in numbers. September 22 and 27 were sampled 2 and 3 times, respectively, at di!erent locations (see Table 1 andFig. 1). New moon (NM) and full moon (FM) are indicated on top of the upper panel. Although reproduction of di!erentplanktic foraminiferal species takes place around the new moon and full moon, a signi"cant increase in numbers ofspecimens occurred only after the full moon. Therefore, we suspect that enhanced abundance ('100 lm) results fromenhanced individual growth rates.

The test-size spectrum of the total planktic foraminiferal fauna changed frequently in the upperwater column above 500 m depth. Large specimens of all the dominant species, N. incompta, G.bulloides, T. quinqueloba, Globigerinoides ruber (white), Globigerinoides sacculifer, Globigerinellasiphonifera, and Orbulina universa, occurred around September 15, 21}22, and 27}29. Throughoutthe deeper water column below 500 m depth, the test size composition of the planktic foraminiferalfauna was dominated by small tests ('90% of tests (250 lm) from the beginning of the cruiseuntil September 27. After September 27, an increasing number of tests '250 lm occurred until theend of the cruise on October 3. In particular, a sharp increase in the number of large tests of N.incompta, G. bulloides, and T. quinqueloba occurred on September 27. In the deeper water column,a maximum of large tests occurred on October 3 at around 2000 m depth.


Fig. 10. Temporal development of live and dead (empty tests) planktic foraminiferal specimens versus water depth. Alongwith increasing depth of the thermocline (diamonds), the habitat of planktic foraminifers expanded from 0}40 to 0}100 mdepth, and the numbers of specimens increased. Maximum numbers of empty tests below 1000 m during the "rst half ofthe sampling period may result from previous production. High numbers of empty tests between 200 and 1500 m duringthe second-half of the campaign result from enhanced foraminiferal growth in surface waters. September 22 and 27 weresampled 2 and 3 times, respectively, at di!erent locations (see Table 1). New moon (NM) and full moon (FM) areindicated on top of the graph (see Fig. 9).

From the species distribution, three faunal groups can be distinguished: (1) N. incompta and N.dutertrei, (2) G. bulloides, G. glutinata, and T. quinqueloba, and (3) G. sacculifer, G. ruber (white), O.universa, and G. siphonifera. These faunal groups are also recognized by multivariate factoranalysis (SYSTAT 7.0 for Windows, principal component analysis (PCA), Varimax rotated)(Fig. 13). Neogloboquadrina incompta and G. bulloides dominate the "rst and second groups withhigh factor scores, respectively. Neogloboquadrina dutertrei, G. glutinata, and T. quinqueloba are lessabundant and are not prominent in the faunal groups. In contrast to groups (1) and (2), the thirdgroup is not dominated by only one species but is characterized by all four species.

The N. incompta fauna (1) and G. bulloides fauna (2) frequently alternated over time, and highfactor loadings ('0.8) were recorded throughout the course of sampling. High factor loadings(0.6}0.8) for the G. sacculifer fauna (3) occurred only between September 13 and 17, after whichspecies of the G. sacculifer fauna were rare.

The N. incompta fauna (1) was found throughout the upper 100 m of the water column, showingno systematic temporal changes in water depth. In contrast, the G. bulloides fauna was at "rst mostfrequent in the upper 60 m, descending to 20}80 m, and to 20}100 m over the course of sampling


Fig. 11. Temporal and water depth related distribution of dominant and associated planktic foraminiferal species.Maximum portions of N. incompta and G. bulloides alternate temporally. Multiple samples on the same day are fromdi!erent locations (Fig. 1 and 14). Turborotalita quinqueloba occurred mainly below 60 m depth. Globigerinita glutinatawas most frequent at the end of September and in early October. Although G. bulloides reproduces around new moon(NM) and T. quinqueloba around full moon (FM), both species show similar temporal distribution.

(Fig. 13). The geographic distribution of faunal groups (1) and (2) shows a distinct interlockedpattern (Fig. 14). The N. incompta fauna (1) dominated at the central BIOTRANS area and at thesouthwestern edge of the investigated area (stations 1, 2, 7, 9, 10, 11, 18). The G. bulloides fauna (2)was found mainly in the northeast (stations 3, 5, 6) and at stations 14 and 18. Both faunal groups (1)and (2) were found in vertical succession in the west and east of the BIOTRANS area, marked bythe shaded regions in Fig. 14. The G. sacculifer fauna (3) occurred in the upper 60 m (Fig. 13) anddisplayed no distinct horizontal distribution pattern.


Fig. 12. Temporal and spatial distribution of subtropical species. In general, the highest portion of subtropical speciesoccurred between September 10 and 17. Subsequently, subtropical species occurred only sporadically. NM"new moon,FM"full moon.

4. Discussion

4.1. Physical properties and trophic conditions

The surface ocean in the BIOTRANS region during September is in a transient state as suggestedby various data sets (e.g. NCEP re-analysis data). September is the "rst month following borealsummer, with a negative though small net heat #ux and slightly decreasing sea surface temperature(Fig. 3). At about 5}15 m s~1, wind velocities are half-way from the comparatively calm summer tothe windy winter conditions. Therefore, in September the mean potential mixed layer begins to


Fig. 13. Temporal development of planktic foraminiferal groups according to multivariate factor analysis (3 factorsselected). For factor loadings ((0 to '0.8) see lower panel. Dominant species and respective factor scores are given atthe right. Factor 1 is dominated by N. incompta. High factor loadings are given for the upper 100 m. High loadings ofFactor 1 alternate with that of Factor 2. Factor 2 is dominated by G. bulloides and has the highest loadings in surfacewaters during the "rst-half of the cruise, and below 20 m depth during the second-half. Factor 3 loadings '0.6 arecalculated for surface waters (0}60 m) between September 13 and 17. Di!erent faunal groups do not correspond toreproduction of di!erent species (NM"new moon, FM"full moon).

increase in depth due to wind-stress and convection, and may shift below the mean critical depth.At the same time, the critical depth decreases due to seasonal changes of the light climate in theupper water column (decrease of insolation during interim periods between photic and aphoticconditions). Thus, the potential for net phytoplankton growth decreases during September. On theother hand, nutrients are entrained in the euphotic mixed layer by wind driven and convectiveprocesses (Zeitzschel et al., 1998). Intensive phytoplankton growth may be initiated during thisperiod until the establishment of aphotic conditions in October (Obata et al., 1996). During M36/5,entrainment of nitrate may have triggered phytoplankton growth, as can be deduced from thephytoplankton succession. Dino#agellates dominated the phytoplankton community throughoutmost of the cruise, except for the last stations, where the phytoplankton composition changed anddiatoms equaled dino#agellates (Sellmer et al., 1998). In addition, redistribution of chlorophyll-aled to the enhanced availability of phytoplankton at the sea surface (Fig. 7). As can be deduced from


Fig. 14. Suspected location of the frontal zone (shaded). Station numbers (1 to 20) are given in chronological order.Globigerina bulloides characterized the fauna at the stations north of the front (Factor 3). South of the front N. incomptafauna prevailed (Factor 1). Within the frontal zone both faunas were interlocked on a depth related scale. For detailedinformation about sites please see Table 1.

the high chlorophyll and low nitrate concentrations (Fig. 8), chlorophyll redistribution wasindependent of nitrate entrainment. On the other hand, high nitrate and high chlorophyll concen-trations prove growth following entrainment. However, a fall bloom of phytoplankton did not takeplace in September 1996 at BIOTRANS.

4.2. Planktic foraminiferal distribution

Enhanced production of planktic foraminifers was synchronous with increasing wind speed.Along with a negative net heat #ux and a deeper convective mixing the Ekman depths increasedfrom 30 m (beginning of September) to 90 m (end of September). The phytoplankton biomassbelow the thermocline (deep chlorophyll maximum, DCM) that accumulated during summer wasredistributed and uniformly spread over the upper water column (Figs. 6 and 7). At the same time,the N. incompta fauna was replaced by the G. bulloides fauna. Numbers of G. bulloides increased inthe upper 40}60 m and T. quinqueloba between 80 and 100 m depth. Highest numbers of T.quinqueloba from below the thermocline (cf. Ortiz et al., 1995), in contrast to shallow dwelling G.bulloides, may point towards di!erent trophic strategies.

Following a shift in the mixed layer depth to below the mean critical depth, at which timenutrients were entrained into the euphotic mixed layer, a second increase in G. bulloides and T.quinqueloba numbers was observed (Fig. 11). Entrainment caused an input of nitrate into the mixedlayer and is suspected to have initiated phytoplankton growth. The phytoplankton compositionchanged and, in addition to dino#agellates, high numbers of diatoms (Rhizosolenia sp.) wereobserved (Sellmer et al., 1998). Although a remarkable increase in the abundance of heterotrophicprotozooplankton (tintinnids, radiolarians) was observed (Sellmer et al., 1998), the chlorophyllconcentration remained constant during this study, except for short-term changes (Fig. 7a).


Therefore, new production of phytoplankton was necessary to compensate for grazing. In additionto tintinnids, radiolarians, and other zooplankton, most planktic foraminifers also feed on phytop-lankton (e.g., Anderson et al., 1979). Maximum numbers of the diatom grazing species G. glutinata(Hemleben et al., 1989) occurred as a reaction to enhanced nutrient concentrations and highnumbers of diatoms (Figs. 4 and 11). Similar successions have been observed during spring bloomin the eastern North Atlantic and often coincide with an increase in the mixed layer depth (Schiebeland Hemleben, 2000). Concerning the distribution of G. bulloides, entrainment of nutrient-richwaters from below the mixed layer may be compared to upwelling. Globigerina bulloides reachesa maximum frequency when surface water temperatures are decreased and nutrient levels are high,as reported from various upwelling areas of the world ocean (e.g., BeH and Tolderlund, 1971; Thiede,1975; Hemleben et al., 1989; Marchant et al., 1998).

Increased numbers of planktic foraminifers are unlikely to have resulted from reproductionduring this investigation, although the maximum increase in numbers of individuals (September 27)occurs one day after the full moon (September 26). Following full moon, an enhanced number oflarge specimens ('250 lm) was observed relative to the smaller individuals ((250 lm), showingthat enhanced availability of food stimulated individual growth. In contrast, enhanced numbers ofjuvenile specimens ((250 lm) would have documented reproduction. Moreover, various speciesreproduce during di!erent phases of the lunar cycle. For example, G. bulloides is shown toreproduce around the new moon (Schiebel et al., 1997), and T. quinqueloba reproduces during thefull moon (Volkmann, in press). However, during September 1996 highest abundance and max-imum numbers of large specimens of both G. bulloides and T. quinqueloba occurred at the sametime. This again demonstrates that the availability of food does not initiate reproduction buttriggers individual growth.

At several stations the N. incompta fauna was recorded either from shallower waters (stations 4,12, 13, 20), or from deeper waters (stations 15, 16, 19) than the G. bulloides fauna, or interlockedwith the G. bulloides fauna (station 17). At these stations, water masses may have been laterallyadvected and subducted, including their biological stock. We suspect these stations to be locatedwithin an oceanic frontal zone (Fig. 14). As the frontal zone is displayed in Fig. 14, it marks theboundary between the N. incompta and G. bulloides fauna and separates low productivity waters tothe south from high productivity waters to the north of the BIOTRANS area. However, oceanicfrontal zones are dynamic features (e.g., Gould, 1985) that have no "xed position. This implies thatdi!erent faunal groups can occur at one location in a temporal succession, as the front passesthrough.

The planktic foraminiferal fauna observed at BIOTRANS in other seasons is di!erent from thefauna discussed here. In January 1994, G. scitula was dominant and total planktic foraminiferalnumbers were much lower than in September and October 1996. Maximum total plankticforaminiferal numbers occur during spring (cf. Schiebel et al., 1995; Schiebel and Hemleben,in press). The spring fauna is characterized by high numbers of G. glutinata related to highnumbers of diatoms. Throughout summer, the numbers are medium to low and N. incomptaand T. quinqueloba are most frequent. Therefore, we suspect N. incompta, recorded during fall,to be a remnant of the low productive summer fauna. In fall, the low-productivity summer fauna isreplaced by a high-productivity fauna, which is characterized by G. bulloides. The fall faunaprogrades from the north towards the south, along with changing atmospheric and hydrographicconditions (Fig. 15).


Fig. 15. Schematic development of hydrographic, trophic, and faunal conditions between September 10 and October 3,1996. The planktic foraminiferal assemblage "rst changed as a result of mixing and chlorophyll redistribution. A secondchange was due to increased mixing depth and entrainment of NO

3from below the nitracline followed by new

phytoplankton production. As a result of chlorophyll redistribution, mainly G. bulloides numbers increased. Subsequentto entrainment of NO

3, numbers of G. glutinata increased (front panel). On a geographic scale the scheme crosses

a frontal zone (top panel). On a water depth versus latitudinal scale, both N. incompta and G. bulloides faunas areinterlocked within the frontal area (side panel).

4.3. The settling assemblage of empty tests

Empty test assemblages result from the faunal composition and from di!erential settling ofvarious test shapes and sizes (Schiebel and Hemleben, 2000), and were found in two distinctpatches of maximum test numbers during M36/5 (Fig. 10). The "rst maximum of empty tests wasfound in the water column below the thermocline between September 10 and 22, and consistedmainly of small tests ((150 lm) of T. quinqueloba. Due to its small and light tests, the verticalsinking velocity of T. quinqueloba is low (100}115 m day~1). Therefore, the settling time of T.quinqueloba tests, between the live habitat at about 100 m and the empty test maximum at2000}2500 m, is about 20 days. Hence, the "rst empty test maximum resulted from a period beforeour cruise. Tests with a higher settling velocity than T. quinqueloba had already passed the sampledwater depth interval and sunk below 2500 m depth. It is unlikely that the "rst deep empty testmaximum resulted from resuspension from the sea #oor (3717}4967 m water depth), because testsare well preserved and show no signs of dissolution, and because no sediment contamination hasbeen observed.

The second maximum of empty tests (Fig. 10) obviously resulted from enhanced test productionin late September. During times of enhanced test production, the number of empty tests and thevertical test #ux also increase (Schiebel et al., 1995). In addition, along with increasing test size, theempty shell production could have been increased as a consequence of gametogenesis around fullmoon (Bijma and Hemleben, 1994; Schiebel et al., 1997) on September 26. As a result of di!erential


settling velocity, the assemblage of empty tests during late September did not simply correspond tothe species composition of the live planktic foraminiferal fauna. While the upper part (200}300 m)of the test patch (Fig. 10) mainly consisted of small and slowly settling tests of T. quinqueloba, thelarger and heavier tests of N. incompta and G. bulloides settle faster and had already reached 1500 mwater depth, forming the lower part of the settling test cloud. As has already been reported, thevertical planktic foraminiferal test #ux does not occur as a steady rain but rather in distinct pulses,resulting from a changing environment (Schiebel et al., 1995) and from reproduction (Schiebel et al.,1997). A combination of both reproduction and enhanced growth led to a pulsed test #ux duringSeptember 1996.

5. Conclusions

Planktic foraminifers respond to the redistribution of chlorophyll and entrainment of nitrate byenhanced growth rates, increasing numbers of large individuals and changing faunal composition.The faunal portion of G. bulloides and T. quinqueloba increased shortly after chlorophyll redistribu-tion, while N. incompta decreased in its faunal portion, in September and early October 1996. Assurface water mixing shifted below the mean critical depth, nitrate was entrained into the mixedlayer, phytoplankton (diatom) production was initiated, and a second increase of foraminiferalnumbers occurred, characterized by maximum numbers of G. glutinata.

Two faunal groups of planktic foraminifers were distinguished in correlation to the transienthydrographic situation at BIOTRANS in fall 1996. A N. incompta fauna mirrors less-mixedsouthern water and displays hydrographic conditions typical for summer. A G. bulloides faunadominates northern surface water, displaying improved feeding conditions that are due to winddriven mixing and redistribution of chlorophyll. Both faunas are interlocked at a frontal zone(Fig. 15).

An increased concentration of empty tests occurred between 200 and 1500 m depth subsequentto increased planktic foraminiferal abundance in the surface mixed layer (Fig. 10). The upper partof the test patch (200}300 m) consisted mainly of small and slowly settling tests of T. quinqueloba,while larger and heavier tests of N. incompta and G. bulloides formed the lower part of the settlingtest cloud. These species reproduce at di!erent times of the lunar cycle and show a di!erentialreaction to feeding conditions. Therefore, we conclude that both enhanced growth andgametogenetic production of empty tests cause an increased #ux rate of tests and CaCO

3.

Acknowledgements

Master and crew of RV Meteor cruise 36/5 are gratefully acknowledged. We thank M. Bayer, P.Fritsche, K. Nachtigall, and C. Reineke for shipboard assistance and sample preparation. We aregrateful to C. Sellmer for providing data on the concentration of nutrients and chlorophyll. Wefurther thank J. Bijma and an anonymous reviewer for their valuable comments. Hilary Paulkindly improved the English. The BMBF "nancially supported our work as part of the GermanJoint Global Ocean Flux Study (JGOFS), grant No. 03F0160A (Ch. Hemleben) and 03F0202A and D (B. Zeitzschel).


References

Anderson, O.R., Spindler, M., BeH , A.W.H., Hemleben, Ch., 1979. Trophic activity of planktonic foraminifera. Journal ofthe Marine Biological Association of the United Kingdom 59, 791}799.

Barlow, R.G., Mantoura, R.F.C., Gough, M.A., Fileman, T.W., 1993. Pigment signatures of the phytoplankton composi-tion in the northeastern Atlantic during the 1990 spring bloom. Deep-Sea Research Part II 40 (1/2), 459}477.

BeH , A.W.H., Tolderlund, D.S., 1971. Distribution and ecology of living planktonic foraminifera in surface waters of theAtlantic and Indian Oceans. In: Funnel, B.M., Riedel, W.R. (Eds.), Micropaleontology of the Oceans. CambridgeUniversity Press, London, pp. 105}149.

Bijma, J., Hemleben, Ch., 1994. Population dynamics of the planktic foraminifer Globigerinoides sacculifer (Brady) fromthe central Red Sea. Deep-Sea Research Part I 41 (3), 485}510.

Drebes, G., 1974. Marines Phytoplankton. Eine Auswahl der HelgolaK nder Planktonalgen (Diatomeen. Peridineen).Georg Thieme, Stuttgart, p. 186.

Gill, A.E., 1982. Atmosphere-Ocean Dynamics. International Geophysical Series 30, Academic Press, New York, pp.1}662.

Gould, W.J., 1985. Physical oceanography of the Azores Front. Progress in Oceanography 14, 167}190.Grassho!, K., Erhardt, M., Kremling, K., 1999. Methods of Seawater Analysis, Wiley-VCH, Weinheim, 1}600.Hemleben, Ch., Spindler, M., Anderson, O.R., 1989. Modern Planktonic Foraminifera, Springer, Berlin, 363pp.Henderson-Sellers, B., 1986. Calculating the surface energy balance for lake and reservoir modeling: a review. Review of

Geophysics 24, 625}649.Herbland, A., LeBouteller, A., Raimbault, P., 1985. Size structure of phytoplankton biomass in the equatorial Atlantic

Ocean. Deep-Sea Research 32, 819}836.Honjo, S., Manganini, S.J., 1993. Annual biogenic particle #uxes to the interior of the North Atlantic Ocean; studied at

343N 213W and 483N 213W. Deep-Sea Research Part II 40 (1/2), 587}607.Kalnay, E., Kanamitsu, M., Kistler, R., Collins, W., Deaven, D., Gandin, L., Iredell, M., Saha, S., White, G., Woollen, J.,

Zhu, Y., Chelliah, M., Ebisuzaki, W., Higgins, W., Janowiak, J., Mao, K.C., Ropolewski, C., Wang, J., Leetmaa, A.,Reynolds, R., Jenne, R., Joseph, D., 1996. The NCEP/NCAR 40-year Reanalysis Project. Bulletin of the AmericanMeteorological Society 3 (77), 437}471.

Le Fevre, J., 1986. Aspects of the Biology of Frontal Systems. Advances in Marine Biology 23, 163}299.Marchant, M., Hebbeln, D., Wefer, G., 1998. Seasonal #ux patterns of planktic foraminifera in the Peru}Chile Current.

Deep-Sea Research Part I 45, 1161}1185.Newton, P.P., Lampitt, R., Jickells, T.D., King, P., Boutle, C., 1994. Temporal and spatial variability of biogenic particle#uxes during the JGOFS northeast Atlantic process studies at 473N, 203W. Deep-Sea Research Part I 41 (11/12),1617}1642.

Obata, A., Ishizaka, J., Endoh, M., 1996. Global veri"cation of critical depth theory for phytoplankton bloom withclimatological in situ temperature and satellite ocean color data. Journal of Geophysical Research 101 (C9),20657}20667.

Ortiz, J.D., Mix, A.C., Collier, R.W., 1995. Environmental control of living symbiotic and asymbiotic foraminifera of theCalifornia Current. Paleoceanography 10 (6), 987}1009.

Reid, P.C., Edwards, M., Hunt, G.H., Warner, A.J., 1998. Phytoplankton change in the North Atlantic. Nature 391, 546.Schiebel, R., Hemleben, Ch., 2000. Interannual variability of planktic foraminiferal populations and test #ux in the

eastern North Atlantic Ocean (JGOFS). Deep-Sea Research Part II 47 (9}11), 1809}1852.Schiebel, R., Bijma, J., Hemleben, Ch., 1997. Population dynamics of the planktic foraminifer Globigerina bulloides from

the eastern North Atlantic. Deep-Sea Research Part I 44 (9), 1701}1713.Schiebel, R., Hiller, B., Hemleben, Ch., 1995. Impacts of storms on recent planktic foraminiferal test production and

CaCO3#ux in the North Atlantic at 473N, 203W (JGOFS). Marine Micropaleontology 26 (1/4), 115}129.

Sellmer, C., Fehner, U., Nachtigall, K., Reineke, C., Fritsche, P., Lisok, K., ObermuK ller, B., Adam, D., 1998. Planktologi-cal Studies. In: Mienert, J., Graf, G., Hemleben, Ch., Kremling, K., Pfannkuche, O., Schulz-Bull, D. (Eds.), Nordatlan-tik 1996. Meteor-Berichte 98-2, 197}200.

SchuK ssler, U., Kremling, K., 1993. A pumping system for underway sampling of dissolved and particulate trace elementsin near surface waters. Deep-Sea Research Part I 40 (2), 257}266.


Thiede, J., 1975. Distribution of foraminifera in surface waters of a coastal upwelling area. Nature 253, 712}714.Thunell, R.C., Honjo, S., 1987. Seasonal and interannual changes in planktonic foraminiferal production in the North

Paci"c. Nature 328, 335}337.Volkmann, R. Planktic foraminifers in the eastern Arctic Ocean * modern distribution and ecology. Journal of

Foraminiferal Research, in press.Williams, R., Robinson, G.A., 1971. Biological sampling at Ocean Weather Station INDIA 593N, 193W in 1971. Annales

Biologiques 28, 1}579.Williams, R., Wallace, M.A., 1975. Biological sampling at Ocean Weather Station INDIA 593N, 193W in 1975. Annales

Biologiques 32, 62}64.Zeitzschel, B., Waniek, J., BuK low, K., SchroK der, P., 1998. Hydrographical Studies, Leg M 36/5. In: Mienert, J., Graf, G.,

Hemleben, Ch., Kremling, K., Pfannkuche, O., Schulz-Bull, D. (Eds.), Nordatlantik 1996. Meteor-Berichte 98-2,181}185.


Planktic foraminiferal production stimulated by chlorophyll redistribution and entrainment of...

Documents

Transcript of Planktic foraminiferal production stimulated by chlorophyll redistribution and entrainment of...