Plankton metabolic balance at two North Atlantic seamounts

ARTICLE IN PRESS Deep-Sea Research II 56 (2009) 2646–2655

Contents lists available at ScienceDirect

Deep-Sea Research II journal homepage: www.elsevier.com/locate/dsr2

Plankton metabolic balance at two North Atlantic seamounts J. Arı´stegui a,, A. Mendonc- a b, J.C. Vilas a, M. Espino a, I. Polo a, M.F. Montero a, A. Martins b a b

Facultad de Ciencias del Mar, Universidad de Las Palmas de Gran Canaria, Spain Centro do IMAR da Universidade dos Acores, Departamento de Oceanografı´a e Pescas, Horta, Portugal

a r t i c l e in f o

a b s t r a c t

Available online 30 December 2008

We have studied the epipelagic (0–100 m) metabolic balance between gross and net community production (Pg, Pn) and community respiration (Rd) around two seamounts (Seine: 341N, 141W; Sedlo: 401N, 271W) located in the subtropical northeast Atlantic. We looked for local effects causing seamounts to increase community production and/or community respiration with respect to the surrounding open ocean. Comparatively, Seine presented similar average living plankton biomass—chlorophyll a (Chl) and particulate proteins (Pt)—but higher Pg, due to higher Rd, presumably the result of organic matter loading from the NW Africa upwelling system, as supported by field results and satellite imagery. Nevertheless, the large temporal and spatial variability at each seamount make the average differences non-significant. Temporal variability in P, Rd and Chl was evident around the two seamounts. Sedlo showed higher Rd, Chl and Pt during winter, but higher Pn in summer. Seine presented higher Pt, Chl and Pn during spring, but higher Rd in summer. On average, the metabolic balance was heterotrophic (Rd4Pg) during all the sampling periods and at most stations of the two seamounts. Both Sedlo and Seine, showed higher Rd with respect to average values reported for the global ocean. A clear seamount effect on phytoplankton was only observed in Seine during spring, when Chl and Pt were enhanced at the summit of the seamount. Our results suggest that, rather than increasing primary production significantly, the two seamounts could act as trapping mechanisms for organic matter, favoured by the development of Taylor Columns on the top of the seamounts. Nevertheless these effects seem to be of a lower magnitude than changes caused by temporal or regional variability, questioning the role of these seamounts as hot-spots of productivity in the oceans. & 2008 Elsevier Ltd. All rights reserved.

Keywords: Seamounts Metabolic balance Organic matter North Atlantic Sedlo Seine (34–411N 14–271W)

1. Introduction The balance between gross production (Pg) and dark-community respiration (Rd) in the upper ocean, the Pg/Rd ratio, sets the potential of surface-water ecosystems to pump organic carbon below the euphotic zone, hence contributing to carbon storage in the deep ocean. Both, Pg and Rd exhibit significant space and time variability (Robinson and Williams, 2005), which arises from the characteristics of different environmental regimes, and from the superposition of many mesoscale processes with a wide range of space and time-scales. Among these, the perturbation of the main flow induced by topographic features, like islands or seamounts, may represent a major source of mesoscale variability in the oceans, with important biological consequences (e.g. Rogers, 1994; Arı´stegui and Montero, 2005). Seamounts are widespread along all ocean basins. They are known to interact with ocean currents, generating large spatial and temporal variability in the physical and biological fields

 Corresponding author. Tel.: +34 928 452906; fax: +34 928 454490.

E-mail address: [email protected] (J. Arı´stegui). 0967-0645/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.dsr2.2008.12.025

(Royer, 1978; Boehlert and Genin, 1987). Most of the studies published in the last two decades regarding seamounts have dealt with the physical environment. Local effects include isopycnal doming above the seamount (Owens and Hogg, 1980), enhancement of vertical mixing (Navatov and Ozmidov, 1988; Lueck and Mudge, 1997; Eriksen, 1998), tidal amplification and rectification (Genin et al., 1989; Eriksen, 1991; Brink, 1990; Kunze and Toole, 1997; Noble and Mullineaux, 1989), or generation of local secondary circulation cells, like bottom-intensified Taylor columns (Roden, 1987; Freeland, 1994). All these hydrographic features can potentially enhance surface productivity by increasing nutrient pumping into the euphotic zone. Seamounts are therefore potentially important in oligotrophic oceans, where they could constitute isles of productivity, benefiting the development of highly diverse and productive pelagic and benthic communities (Rogers, 1994). Indeed, several studies have reported increases in phytoplankton biomass and productivity over seamounts, resulting from the combined effect of nutrient inputs into the euphotic zone and surface-water stratification, associated with isopycnal shallowing (Genin and Boehlert, 1985; Dower et al., 1992; Comeau ˜ o et al., 2001). et al., 1995; Odate and Furuya, 1998; Mourin Moreover, some studies have related the presence of highly

ARTICLE IN PRESS J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

diverse and abundant benthic communities with phytoplankton enrichments around seamounts (Rogers, 1994). For instance, Uchida and Tagami (1984) postulated that the cause of high abundances of groundfish populations in the North Pacific Ocean was related with the presence of numerous seamounts in the region. Other authors have argued, however, that increases in biomass and production over seamounts may be due to the retention and accumulation of living organisms and organic matter, favoured by local anticyclonic circulation cells generated around the topographic features (Dower et al., 1992; Mullineaux and Mills, 1997; Beckmann and Mohn, 2002; Genin, 2004). Overall, however, there is a high degree of uncertainty in the biological responses to flow perturbation by seamounts. Many directed studies looking for cause–effect relationships yielded inconclusive evidence for an enhancement in productivity or biomass accumulation due to physical forcing (Genin and Boehlert, 1985; Dower et al., 1992; Comeau et al., 1995; Odate ˜ o et al., 2001). Weak or no biological and Furuya, 1998; Mourin enhancements in plankton communities have been reported from many seamounts studies, in spite of perturbation of the local ˜ o et al., physical environment (e.g., Dower et al., 1992; Mourin 2001). Whether these inconsistencies in the biological response to variability in the impinging flow are the rule or the consequence of an inadequate temporal and spatial sampling resolution is still uncertain. Within the framework of the European project OASIS (OceAnic Seamounts, an Integrated Study), we carried out an interdisciplinary study around two North Atlantic seamounts: Sedlo and Seine. The main goal of our research was to look for biological responses in microplankton (o200 mm) biomass and metabolism to seamounts effects, and compare them in the context of temporal and regional variability. Up to 6 stations were sampled around Seine during March 2004 (R.V. Poseidon 309) and July 2004 (R.R.S. Discovery 282), while up to nine stations were sampled at Sedlo during November 2003 (R.V. Meteor M60/1) and July 2004 (R.R.S. Discovery 282). We first analysed the differences in the average chlorophyll a, microplankton proteins, and community production and community respiration, between the two seamounts. Then, we evaluated the temporal variability at each station, and finally we looked at the spatial variability within each seamount, compared with the far field.

2647

‘‘time-zero’’, dark and light 125-ml-BOD bottles, and incubated for 24 h. Dissolved oxygen was measured by the micro-Winkler technique, following the recommendations of Carrit and Carpenter (1966), Bryan et al. (1976) and Grasshoff et al. (1983). The entire contents of the bottles were titrated during o3 min by means of an automated, precise titration system, with colorimetric end-point detection (Williams and Jenkinson 1982). The precision achieved in replicates was %CV o0.05. Rd was estimated from the difference in oxygen concentration between the time-zero and dark bottles. Pn in a daily basis was estimated as the difference between the light and time-zero bottles, by assuming that respiration in dark and light were equal. Pg was calculated as the sum of Pn and Rd. A Q10 of 2 for P and R was used to correct for temperature differences with in situ values (Robinson and Williams, 1993). The depth of the 1% surface-light ranged from 85 to 120 m; hence, we integrated our metabolic rates down to 100 m, in order to look for the metabolic balance in the euphotic zone. 2.2. Phytoplankton pigments Chlorophyll a (Chl) and phaeo-pigments (Pha) were estimated fluorometrically according to Parsons et al. (1984). Seawater samples (1 L) were filtered through Whatman GF/F fiber-glass filters. The filters were stored in liquid nitrogen until assayed. Pigments were extracted in cold acetone (90% v/v) for 24 h. Fluorescence before and after acidification was measured by means of a Turner Designs bench fluorometer, previously calibrated with pure chlorophyll a (Sigma Co.). 2.3. Microplankton proteins

2. Material and methods

Microplankton proteins (Pt) were determined according to the Peterson0 s modification (Peterson, 1983) of the Lowry et al. (1951) method; using a protein assay kit provided by Sigma Co. Water samples (4 L) were concentrated on GF/F filters. Pt were extracted by grinding the filters with Lowry reagent directly and diluting with water afterwards. Sodium dodecylsulfate, included in the Lowry reagent, facilitates the dissolution of relatively insoluble lipoproteins. Replicate assays were run for each sample. Bovine serum albumin (BSA) standards with Pt concentrations between 4 and 400 mg m3 were run at the same time to obtain a calibration curve. The average precision obtained in replicate samples was %CV o3.0.

2.1. Gross and net community production and respiration by oxygen changes

3. Results

Discrete samples for metabolic experiments were collected at each station with Niskin-type bottles from six depths ranging from surface to 150 m. Gross (Pg) and net (Pn) community production, and community respiration (Rd) were determined by the oxygen method after incubations inside borosilicate bottles. Samples were incubated in ‘‘on-deck’’ incubators, reproducing the in situ light conditions. The attenuation of photosynthetic active radiation (PAR) in the water column was measured with a profiling natural fluorescence PNF300 instrument (Biospherical Co.). PAR irradiances were reproduced in the incubator by screening with neutral-density meshes. The PAR inside each bottle was measured with a quantum scalar irradiance meter QSL100 (Biospherical Co.). The temperature control of the incubator was achieved by continuous seawater flow inside the incubation chambers. Differences in temperatures between in situ depths and the on-deck incubator ranged from 0 to 6 1C at 100 m. Water samples were allowed to reach the incubator temperature, before carefully siphoned using a silicone tube into 4–5 replicate

3.1. Hydrographic context Seine and Sedlo seamounts are located in the same biogeographical region—the North Atlantic Subtropical Gyral (Longhurst, 1998), under the influence of the North Atlantic Central Waters (NACW) in the upper 700 m (Bashmachnikov et al., 2009). The two seamounts are isolated features, but differ in their topography and summit depths (Fig. 1). Seine is a cone-shaped seamount with a single shallow summit (175 m), placed between Madeira and the NW African coast, whilst Sedlo is a chain seamount composed of three deep peaks (being the shallowest at 780 m depth), situated north of Azores. The upper 200-m layer at Sedlo is characterized by a weak and unstable meandering eastward flow, branching from the North Atlantic Current. Seine, however, is situated in the area of direct influence of the more constant eastward flow of the Azores current (AC). Mohn et al. (2009) and Bashmachnikov et al. (2009) described complex hydrographical patterns around both seamounts, but concluded

ARTICLE IN PRESS 2648

J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

Fig. 1. Geographical location with main surface currents, and maps of Sedlo and Seine seamounts with station positions. Stations F and X1 at Sedlo and stations H and I at Seine were considered ‘‘far field’’ stations. NC: North Atlantic Current; AC: Azores Current; CC: Canary Current.

from their observations that the two features presented typical anticylonic circulations around their summits, driven principally by the Taylor Column formation process, as tidal rectification was weak. In November 2003, the anticyclonic rotation at Sedlo was disrupted by the collision of a Mediterranean water eddy (Meddy), leading to a counter-pairing vortex at the upper layers (Mohn et al., 2009), with strong implications on the organic matter and plankton distribution (see Section 4 below). The analysis of satellite sea-surface temperature (SST; AVHRR) and ocean-colour (Chl; SeaWiFS and MODIS) data around Seine and Sedlo regions, during a 8-year period (1999–2006), allowed to distinguish characteristic seasonal differences between the two regions, due to latitudinal variability (A. Mendonc- a, unpublished results). Sedlo, is characterized by a mid-latitude seasonality with large spring and small autumn phytoplankton blooms, and higher average values in Chl relative to Seine seamount, which presented a lower seasonal variability in temperature and Chl (Fig. 2). Monthly distributions of SST and mean surface Chl at the two seamounts (data not shown) show that the highest Chl values in Seine are found in March–April, coinciding with the lowest SST. This reflects a typical subtropical situation with phytoplankton production triggering after the erosion of the surface thermochline due to winter cooling. In Sedlo, however, the phytoplankton bloom develops during May–June, coinciding with a rise in SST, which leads to a reestablishment of the surface thermochline after winter convection. The Seine field-samplings (March 2003 and July 2004) coincided with the two extreme situations of the phytoplankton annual cycle: lowest SST/highest Chl (March 2004), and highest SST/lowest Chl (July 2004). At Sedlo we could not occupy the seamount during the phytoplankton bloom, but we sampled at two contrasting situations: low SST/moderate Chl (November 2003) and high SST/low Chl (July 2004). Seine presented warmer and saltier surface waters than Sedlo in July 2004 (Fig. 6). However, a strong near-surface thermocline was developed in the two cases, extending from about 30 m depth to more than 200 m depth. In March (Seine) and November (Sedlo)

Fig. 2. Map of annual-average Sea WiFS chlorophyll (year 2005), showing the position of Seine and Sedlo seamounts. Whitish tones indicate higher chlorophyll concentrations. Notice the offshore extension of the upwelling filament at Cape Guir towards the Seine seamount region.

surface temperatures were lower, and the mixed layers extended down to 90 m in Sedlo and 4160 m in Seine (Fig. 5). 3.2. Inter-seamounts variability The average (7SE) depth-integrated (0–100 m) daily rates of Pg and Rd (Fig. 3) were higher in Seine (Pg ¼ 158770, Rd ¼ 235778 mmol O2 m2 d1) than in Sedlo (Pg ¼ 78727, Rd ¼ 185724 mmol

ARTICLE IN PRESS J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

2649

Table 1 Integrated (0–100 m) values of gross production (Pg), net community production (Pn), dark community respiration (Rd), chlorophyll a (Chl) and microplankton proteins (Pt) from stations around Seine seamount (A, C, E and F), compared to farfield stations (H and I; see Fig. 1 for locations). Seine seamount A

C 2

O2 m2 d1), although the large variability in the rates makes the differences non-significant. As explained below, there was a large degree of temporal and spatial variability at each seamount superimposed to the regional variability (Tables 1 and 2). Pn was on average negative in the two cases, indicating net heterotrophy around the seamounts during the times of study. The average Pg/Rd ratio – which sets the degree of heterotrophy of a system – was o1 in both seamounts, but more variable in Sedlo (ranging from 0.1 to 1.4) than in Seine (0.4–1.1). According to this, the two ecosystems would be in metabolic imbalance during the sampling dates. The apparent differences in microplankton metabolic rates were not matched by similar differences in Chl and Pt, the biomass proxies of phytoplankton and total microplankton, respectively (Tables 1 and 2). Indeed, average integrated Chl and Pt were rather constant, with low standard deviations (Fig. 4), suggesting that variability in Pg and Rd were controlled by changes in specific metabolic rates, rather than from changes in microplankton biomass. The average (7SE) Chl/Pt ratio (mg/g) – an index of the autrotrophic state of the system – was low both in Sedlo (5.570.5) and Seine (5.670.6). The values were characteristic of oligotrophic regions (Dortch and Packard, 1989), pointing to an inverted trophic pyramid with dominance of hetrotrophic organisms over autotrophs.

3.3. Temporal variability Average temporal differences in plankton metabolism between cruises were apparent in the two seamounts, but frequently lower than the spatial variability at each seamount. The vertical distributions of the oxygen production rates were not always clearly dependent on the depth of the mixed layer (compare Figs. 5 and 6). Pn in Seine was significantly (Mann–Whitney rank sum test; MW, Po0.05) higher in March (lower negative values) than July 2004, showing in all cases positive values (1 mmol O2 m3 d1) in the upper 25 m and negative values (1 mmol O2 m3 d1) below 40 m depth (Fig. 5A). Integrated Rd, and hence Pg, were, however, on average, about 2-fold higher in July than in March (Table 1), being the Rd differences between the two periods statistically significant (MW, Po0.05). In Sedlo, average temporal differences also were marked in P and Rd, with lower (more negative) Pn, lower Pg and higher Rd values in November than July (Table 2), although the differences were not significant at the Po0.05 level. Comparing average vertical profiles from summer 2004 (the only common period in which the two seamounts were sampled), Seine presented lower Pn but higher Pg rates than Sedlo,

E

F

Average (SE)

I

) 83 601

56

142 72

Pn (mmol O2 m2 d1) March 28 55 July 130 128

120

7 79

138 729

176

135 151

128 (9) 315 (139)

30 14

14

27 18

32 (4) 16 (1)

Rd (mmol O2 m2 d1) March 110 July 202

Chl (mg m2) March 38 July 17

Pt (g m2) March July

H

1

Pg (mmol O2 m d March 83 July 72

Fig. 3. Inter-seamounts variability. Comparison between Seine and Sedlo of average (SE) water column (0–150 m)-integrated values (all units in mmol O2 m2 d1) of gross oxygen production (Pg), net community production (Pn) and dark community respiration (Rd).

Far field

6.5 2.4

4.5 3.6

103 (20) 200 (133)

141 621

25 (18) 114 (12)

4.2 4.7

56 259

197 362

26 16

5.1 (0.9) 3.6 (0.8)

15 16

4.1 3.2

3.1 3.2

SE ¼ standard error.

Table 2 Integrated (0–100 m) values of gross production (Pg), net community production (Pn), dark community respiration (Rd), chlorophyll a (Chl) and microplankton proteins (Pt) from stations around Sedlo seamount (A to X17), compared to far field stations (F and X1; see Fig. 1 for locations). Sedlo seamount A

B

Pg (mmol O2 m2 d1) November 25 July 98

Far field C

X3

175 143

7 82

Pn (mmol O2 m2 d1) November 259 184 July 96

3 41

211 49

Rd (mmol O2 m2 d1) November 284 195 July 194

172 102

218 132

24 19

26 20

Chl (mg m2) November July

Pt (g m2) November July

27 18

11

D

24

5.0 2.6

4.1 3.6

6.4 3.8

X14 X17 Average (SE)

55 (47) 108 (22)

F

X1

27

163 (66) 54 35 (49)

32

4.1

30

5.0

31

8.2

217 (28) 143 (33)

81

28 (1) 19 (1)

29 26

5.5 (0.7) 3.3 (0.5)

3.6 3.0

27

5.3

SE ¼ standard error.

as result of near 2-fold higher Rd rates throughout the water column (Figs. 5A,B). Chl and Pt displayed also a clear temporal variability, but in these cases their average vertical distributions were closely related to the structure of the water column. In Seine, there was more Chl (about 2-fold) and Pt (about 1.5-fold) during March

ARTICLE IN PRESS 2650

J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

(MW, Po0,05) – when the mixed layer deepened down to 180 m – than July, particularly in the upper 100 m (Fig. 6A). The average Chl/Pt (mg/g) ratio was also higher in March (6.3) than in July (4.4), reflecting a relative increase in phytoplankton biomass respect to the total microplankton biomass. A deep-chlorophyll maximum (DCM) was observed in July at about 75–100 m, at the base of the surface thermochline. Vertical profiles of Pt in March and July showed similar homogeneous patterns of distribution in the upper 150 m, with a slight decreasing gradient from surface to depth (Fig. 6B). The variability in Chl and Pt were not paralleled by comparable changes in metabolic rates in all cases. Lower Pn, higher Rd, and lower Chl and Pt than March characterized July. Since Pt represents an index of biomass of the whole microplankton community, the specific Rd rates would be higher in July than in March. In Sedlo, the temporal variability in Chl and Pt was even higher than in Seine in the upper 50 m, with 3-fold more Chl and 2-fold more Pt in November, with respect to July (MW, Po0.05; Figs. 6C,D). However, the average Chl/Pt (mg/g) ratio for the upper 100 m was similar during the two periods (5.1 and 5.8 for November and July, respectively), suggesting a concomitant variability in the autotrophic and heterotrophic biomasses. A

Fig. 4. Inter-seamounts variability. Comparison between Seine and Sedlo of average (SE) water column (0–150 m)-integrated values of chlorophyll a (Chl; mg m2) and microplankton proteins (Pt; g m2).

sharp and constant DCM (40.3 mg Chl m3) was observed in July, at about 75 m, near the base of the surface thermochline, with decreasing Chl values towards the surface (0.1 mg Chl m3). On the contrary, homogeneous Chl concentrations (0.3 mg Chl m3) were found in the upper 75 m in November, through the thick (90 m) surface mixed layer. A maximum in Pt also was observed coinciding with the DCM during July, although the relative differences in magnitude with respect to the surface waters were smaller. In November, Pt decreased with depth, but showed higher concentrations than in July in the upper 100 m. Like in Seine, there was no coincidence in the variability observed in the Chl and primary production. Lower Pg but higher Chl than July characterized November. However, a good agreement was observed in Rd and Pt, with higher values corresponding to November. 3.4. Spatial variability Large spatial variability was observed around the two seamounts, compared with far-field stations. In Seine, higher Pg and Rd were measured in July at the far-field station ‘‘I’’, compared to most of the seamount stations. However, station C – at the eastern margin of Seine – presented extremely high integrated values of Rd: about 2 times more than station ‘‘I’’, and 4–5 times more than the other seamount stations (Table 1). In March, higher Rd was also observed in ‘‘I’’ with respect to the seamount. These differences suggest that station ‘‘I’’ – closer than the rest to the African coastal upwelling – could be affected by external loadings of organic matter, hence increasing Rd. A concomitant increase in microplankton biomass was, however, not observed at this station. Rather, Chl or Pt values were comparable or lower than those observed at stations around Seine. Increases in Chl, phaeopigments (Pha) and Pt were found at station ‘‘A’’, on the summit of the mount, during March (Figs. 7A–C). Moreover, the Chl/Pt ratio (7.1 mg/g) was 1.5–2 times higher than at the rest of stations around the seamount or in the far field. This was the only time of our samplings, around Seine and Sedlo, in which a clear enhancement on phytoplankton biomass was observed, probably as result of a trapping mechanism rather than due to a local increase in productivity, since the increase in Pha – the degradation products of Chl – was even higher, and Pg and Pn were of the same magnitude as in other stations (Fig. 7B).

Fig. 5. Seasonal variability in Seine and Sedlo. Vertical profiles of gross oxygen production (Pg, circles), net community production (Pn, squares) and dark community respiration (Rd, triangles) in (A) Seine and (B) Sedlo during the four surveys. Values are average (SE) data from all the stations around each seamount (excluding the far-field stations). All units in mmol O2 m3 d1.

ARTICLE IN PRESS J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

2651

Fig. 6. Seasonal variability in Seine and Sedlo. Vertical profiles of chlorophyll a (Chl; mg m3) and microplankton proteins (Pt; mg m3) in (A and B) Seine and (C and D) Sedlo during the four surveys. Values are average (SE) data from all the stations around each seamount (excluding the far-field stations). Vertical profiles of temperature are plotted in (A) and (C) as thick dark (March and November 2003) and grey (July 2004) lines.

Contrary to Seine, stations around Sedlo showed higher average integrated Pg and Rd than at station ‘‘F’’ at the far-field (Table 2). These higher rates corresponded also with higher Pt values around the seamount, compared to the far-field stations ‘‘F’’ and ‘‘X1’’, although differences in Chl were not apparent. Thus, the Chl/Pt (mg/g) ratio was consistently higher at the far-field station F (8.1 for November and 8.7 for July) than at the seamounts stations (average values: 5.1 for November and 5.8 for July). The representation of vertical profiles of Pt during November, allows the grouping of stations into three clusters with different distributions (Fig. 8A). Stations to the south and southwest of Sedlo (D, X14 and X17) presented the highest Pt values (4100 mg m3) in the upper 25 m. Stations to the north or in the far-field (X1, X3 and F) showed the lowest values in the water column decreasing from surface to depth, while stations B and C, located between the former two groups, presented intermediate values. A similar pattern in the distribution of particulate organic

matter was also observed by Vilas et al. (2009). During July, the integrated Pt values were not significantly different between stations. However, a slight seamount effect was observed in the vertical distribution of the Pt maximum (Ptm). Station D, on the trough of the two seamounts peaks, presented the shallower (50 m) Ptm, while the deeper (100 m) Ptm was found at the farfield station F. Stations A and C, presented lower Ptm at intermediate depths (80 m; Fig. 8B).

4. Discussion 4.1. Heterotrophic metabolic balance around Seine and Sedlo The two seamounts studied – Seine and Sedlo – were different in topography, summit depth, and regional location. However, the plankton metabolic balance around the two topographic features

ARTICLE IN PRESS 2652

J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

Fig. 7. Intra-seamount variability in Seine. Vertical profiles of (A) chlorophyll a (Chl; mg m3), (B) phaeopigments (Pha; mg m3) and (C) microplankton proteins (Pt; mg m3) in Seine during March 2004. Profiles at the summit of Seine (st.A) are compared with averaged values of other seamount stations (C and F) and the far field (H and I).

Fig. 8. Intra-seamount variability in Sedlo. Vertical profiles of microplankton proteins (Pt; mg m3) in Sedlo during (A) November 2003 and (B) July 2004. In November, profiles correspond to average data of stations clustered into three different groups (see text for details).

was in both cases of net heterotrophy (PgoRd) during the times of study, and the differences in average plankton living biomass and metabolic rates were not significant. This was in part due to the large temporal and spatial variability observed around the two seamounts. In particular, Pg ranged from 7 to 175 mmol O2 m2 d1 in Sedlo and from 56 to 601 mmol O2 m2 d1 in Seine. Average depth-integrated (0–100 m) Rd in Seine (235 mmol O2 m2 d1) and Sedlo (185 mmol O2 m2 d1) was up to 2 times higher than average values from the global open-ocean waters (116 mmol O2 m2 d1; Robinson and Williams, 2005), although comparable or even higher Rd values have been reported in other studies. Gonza´lez et al. (2002) measured Rd rates ranging from o200 to 4400 mmol O2 m2 d1 in a latitudinal section across the central Atlantic Ocean, from 501N to 501S, repeated through two different periods (October and May). They observed that higher Rd rates were associated with lower nutrient environments, and suggested

that microbial metabolism would be controlled by the degree of nutrient limitation. Indeed, in a former study, these authors (Gonza´lez et al., 2001) found that Rd was about 25% lower inside mesoscale features than outside. These features included a single station over the Great Meteor Tablemount (GMT), a North Atlantic seamount located about 101 west of Madeira, which was characterized by positive Pn (Pg/Rd41) and lower Rd than the surrounding waters, although no significant differences were found in the bacterial production, Chl concentrations, or the microbial abundances. Surprisingly, the integrated Chl values inside and outside the mesoscale features (11 and 13 mg m2, respectively) were up to 2–3 fold lower than our average estimates for Seine and Sedlo (16–32 mg m2 Chl), or even the values reported by Gonza´lez et al. (2002) from their Atlantic meridional section (20–38 mg m2 Chl). In our study, and contrary to the results obtained by Gonza´lez et al. (2002) from their single

ARTICLE IN PRESS J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

station over the GMT, Pn was lower at the seamount with respect to the far-field in most situations, suggesting nutrient limitation or enhanced respiration. If we assume that Pg represents ‘‘gross primary production’’, the assimilation number (the Pg/Chl ratio) – an index for the productivity efficiency – would be o0.3 mmol O2 mg1 Chl h1 for Sedlo and o0.2 mmol O2 mg1 Chl h1 for Seine. These values would indicate strong light or nutrient-limitation in phytoplankton growth (Falkowski, 1981) during all the sampling periods. However, it is unlikely that phytoplankton would be nutrient limited during March (Seine) and November (Sedlo), when the upper 100 m water column was mixed. Indeed Chl and Pt average values were about 1.5–2-fold higher in these periods with respect to July, when the surface waters were strongly stratified at both seamounts, and presumably nutrient limitation was higher. Rather, it is more likely that Pg does not represent the ‘‘gross primary production’’, but the ‘‘overall gross production’’, since it is based not only on community respiration of autochtonous organic matter, but also on respiration of organic material advected from distant sources, and presumably accumulated at the seamounts (see below). 4.2. Trapping effect around Sedlo Marked enhancements in microplankton Pt and Rd were observed at the seamount stations around Sedlo, compared to the far-field, during November 2003, the only period where farfield stations were sampled. The enhancements could be caused by a retention mechanism, since no evidence of local enhancements in productivity can be inferred from changes in Pn or Chl at the seamount. Mohn et al. (2009) reported the presence of a typical anticyclonic circulation around the summit depth region of Sedlo, driven by a Taylor Column formation process, which would favour the accumulation of organic matter and explain the higher Rd. Indeed, Vilas et al. (2009) found during November 2003 a clear increase in particulate organic carbon (POC) and nitrogen (PON), in the upper 1000 m of the water column at several stations (D, 14 and  17) located at the southern flank of the seamount, where Pt and Rd were also high (Fig. 7; Table 2). During this period, the characteristic anticyclonic flow was disrupted due to the collision of a Meddy with the southern part of the seamount (Bashmachnikov et al., 2009), leading to a counter pairing of vortex cells in the upper 100 m: a cyclone placed at the northeastern sector of the seamount, and an anticyclone located at the south-western sector (see Fig. 8 in Mohn et al., 2009). The opposite flow patterns of the vortices would explain the differential distribution of a passive tracer like the organic matter, and hence of the Pt concentrations and Rd rates.

2653

the African coast. The situation would be similar to that described by Arı´stegui et al. (2003) for the coastal transition zone south of the Canary Islands. Observations of high organic matter concentrations in the vicinity of Seine by Vilas et al. (2009) support this hypothesis. These authors found that POC was significantly more abundant in Seine than Sedlo in the upper 200 m. Moreover, they also observed POC peaks at surface and intermediate layers at the Seine stations, which suggested lateral transport from the NW African coast. Satellite images of ocean colour from the subtropical NE Atlantic illustrate recurrent coastal-open ocean transport of surface Chl linked to the giant upwelling filament of Cape Guir, which at times may extend hundreds of kms into the open ocean, reaching as far as Madeira, and therefore the Seine region (Fig. 2; Herna´ndez-Guerra and Nykjaer, 1997; Pacheco and Herna´ndez˜ oz et al. (2005) recently showed that, Guerra, 1999). Garcı´a-Mun even during upwelling relaxation periods, the Cape Guir filament may transport offshore considerable amounts of organic matter to the open ocean, representing up to 60% of the average primary production from the coastal region. A large part of this organic matter rather than being respired near the upwelling boundary is transported offshore to the ocean domain while is remineralized (A´lvarez-Salgado et al. 2007). Alternatively, filaments from the upwelling region at the southwest coast of Iberia may influence the Seine region. Peliz et al. (2004) using SeaWiFS and SST images, and simulations of Lagrangian trajectories, described a pigment-rich filament originating south of Iberia and stretching south-westward, during winter 2001. These authors argued that the process driving the cross-shelf transport was not related to the wind stress but rather to mesoscale dynamics that generated an advective path between offshore eddies. The filament was about 400 km long and reached as far as 131W and 351N (close to Seine), and presented Chl values up to 1–1.5 mg m3, compared to the much lower background concentrations (0.1–0.2 mg m3) outside the filament. Other seamounts placed in boundary regions, under the influence of coastal regions or upwelling systems, have been described to be affected by advection of organic matter from the ˜ o et al. (2001) analysed the seasonal and spatial far field. Mourin variability in the Chl field through SeaWiFS imagery around the GMT. They observed differences in the seasonality and absolute magnitude of the Chl cycle at the eastern and western sectors of the seamount, which interpreted as caused by lateral advection due to its proximity to the Canary Islands and the African upwelling. Odate and Furuya (1998) found after repeated surveys over Komahashi no. 2, a seamount located on the Northeast Pacific, a Chl maximum resulting both from local production and advection of water with higher Chl content from the upstream flow.

4.3. Coastal advection of organic matter to Seine 4.4. Plankton biomass enhancement over Seine summit Contrary to Sedlo, Rd at the far-field stations of Seine presented higher or similar rates than around the seamount. Duarte et al. (2001), in a compilation of metabolic rates from the subtropical Northeast Atlantic, estimated an average Rd of 239 mmol O2 m2 d1. This value is higher than the average Rd measured around Seine during March (12879 mmol O2 m2 d1), but smaller or comparable to the Rd observed in Seine during July (3157139 mmol O2 m2 d1), the time of the year when the northwest Africa coastal upwelling is stronger (Arı´stegui et al., 2006). The highest Rd rates in July were observed at the eastern stations, with a decreasing gradient towards the west. This suggests that Seine could be affected by lateral transport of organic matter from the eastern far field, which would enhance respiration along the eastern boundary region between Seine and

The only evidence of phytoplankton pigments enhancement around the two seamounts was observed on the summit (station A) of Seine, during March 2004 (Table 1; Fig. 7A). Integrated (0–150 m) Chl and Phae values at st. A were 20% and 38% higher than at other stations around the seamount (C and F), and 49% and 57% higher compared to the far-field stations (H and I). Since Phae result from the degradation of chlorophylls, the higher increase in Phae, with respect to the increase in Chl, could respond to a passive aggregation of senescent phytoplankton rather than to an active growth of phytoplankton cells. The pigment enhancements were paralleled by a similar increase in microplankton Pt, but not on local Pg, supporting the hypothesis of an accumulation effect. Vilas et al. (2009) reported POC values up to 40 mM at station A of

ARTICLE IN PRESS 2654

J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

Seine. These values were on average 4 times higher than those measured at the other stations around the seamount or in the farfield during the same cruise, and could only be explained by local retention of particles over the top of the seamount. A short-period (1 week, March 2003) deployment of a bottom-mounted ADCP was made near the seamount summit rim during the OASIS project, indicating the presence of a tidally rectified flow, consistent with a mean anticyclonic flow. However, a longer duration (March–May 2004) current measurements at two opposite flanks of the seamount indicated strong mesoscale variability present in the flow (Bashmachnikov et al., 2009). Like in Sedlo, the anticyclonic circulation could be driven by the Taylor Column formation process, as tidal rectification is weak at both seamounts. If so, this might have favoured the accumulation of organic material, either produced on top of the seamount or advected from the far-field, over the summit of the mountain. ˜ o et al. (2001) detected also sporadic increases in Chl Mourin and P values around the GMT. Only in one of the five cruises carried out along several years they observed relative maxima of Pn and Chl over the seamount than further away. Although their results were somewhat inconclusive, they suggested that these local increases would be produced by a combined effect of shortterm variability, inducing a shallowing of the pycnocline, and a retention mechanism over the seamount. During some of their samplings, they observed differences in the phytoplankton taxonomic composition, with larger diatoms and dinoflagellates over the GMT. They calculated a mean residence time of the water over the seamount about 2.5 times less than in the surrounding oceanic region, and concluded that the GMT could act as trapping mechanism for living biomass. However, only during one of the cruises they appreciated a clear anticyclonic path from the drogued drifters they deployed over the seamount, questioning the presence of a recurrent anticyclonic cell.

4.5. Seamounts effects: local enhancements in productivity vs biomass accumulation According to the published literature, increases in Chl and productivity seem to be sporadic rather than persistent, and almost all studies – including this one – fail to demonstrate recurrent cause–effect relationships between physical forcing and productivity enhancements over a particular seamount. For example, despite intense occupation of Cobb seamount, on the Northeast Pacific, only once a dramatic increase in water turbidity was observed around its shallow rim, due to a 7-fold increase in Chl (Dower et al., 1992). Rather, relatively low and uniform values on Chl and P were observed over and away from this seamount (Comeau et al., 1995). Also, from a total of 17 surveys around different seamounts across the North Pacific Ocean, in only one case an increase in Chl was observed, as result of nutrient enrichment in the photic layer (Genin and Boehlert, 1985; Genin, 2004). The inconsistencies observed in the biological responses over Seine and Sedlo, as well as other seamounts, are probably the result of the large variability in the impinging flow and the topographic differences among seamounts. Although flow perturbation may at times enhance primary production, if the flow is variable and particles are not retained over the seamount, the biological signal would disappear in a short period. Goldner and Chapman (1997) demonstrated, using a numerical model, that a tall isolated seamount would form a Taylor Column under a steady inflow, retaining passive particles over its summit for a duration that scales inversely with the mean inflow speed. Reported estimations for time-scales of particles’ retention range from days to several weeks (e.g., Genin and Boehlert, 1985; Dower

˜ o et al., 2001), a fact that must have important et al., 1992; Mourin consequences for the productivity of pelagic and benthic seamount ecosystems. Indeed, flows over and around seamounts are far from being steady, and changes in their strength, direction and persistence may preclude the accumulation of local production or allochtonous material advected to the seamount.

5. Conclusions In our study, Seine and Sedlo presented higher Rd values and lower Pn compared to reported average values from the global open ocean, and more specifically the North Atlantic (Robinson and Williams, 2005). Rd rates were particularly high in Seine during summer, presumably due to organic matter loading from the NW Africa upwelling system, as supported by our field data and satellite imagery. Nevertheless, our results show an important degree of both spatial and temporal variability in metabolic rates and plankton biomass, instead of a clear and persistent pattern around the seamounts, probably caused by variability at different scales in the physical environment. Sporadic increases in Chl (and presumably in productivity) may take place, as observed in Seine during the spring survey, although the seamounts studied seem to behave preferentially as trapping mechanisms for organic matter, rather than as local sources of productivity. Accumulation of organic matter would be favoured by the development of anticyclonic circulations on the top of the seamounts, consistent with Taylor Columns formation, as derived from current measurements. In order to quantify precisely the magnitude of the local enhancement versus the accumulation effect, a more detailed sampling would be necessary, addressing different scales of variability in the physical and biological fields, ranging from days to months.

Acknowledgements This research is part of the Project OASIS, funded by the EC (Contract no. EVK3-CT-2002-00073), under the Framework V programme. We thank the crews of the F.S. Poseidon, F.S. Meteor and R.R.S. Discovery, for their professional assistance at sea. We thank Dr. Bernd Christiansen for his invaluable effort coordinating the OASIS project and providing ship time. A. Mendonc- a and I. Polo were contracted under the OASIS project. J.C.Vilas acknowledges a Ph.D. fellowship from the Spanish Ministry of Education and Science (MEC). The satellite image of Fig. 2 was kindly provided by A. Herna´ndez-Guerra (ULPG; Sea WiFS project). Two anonymous reviewers contributed with their comments and suggestions to improve the manuscript. References A´lvarez-Salgado, X.A., Arı´stegui, J., Barton, E.D., Hansell, D.A., 2007. Contribution of upwelling filaments to offshore carbon export in the subtropical Northeast Atlantic Ocean. Limnology and Oceanography 52 (3), 1287–1292. ˜ oz, M., Esca´nez, J., 2003. Arı´stegui, J., Barton, E.D., Montero, M.F., Garcı´a-Mun Organic carbon distribution and water-column respiration in the NW Africa–Canaries coastal transition zone region. Aquatic Microbial Ecology 33, 289–301. Arı´stegui, J., Montero, M.F., 2005. Temporal and spatial changes in plankton respiration and biomass in the Canary Islands: the effect of mesoscale variability. Journal of Marine Systems 54, 65–82. Arı´stegui, J., A´lvarez-Salgado, X.A., Barton, E.D., Figueiras, F.G., Herna´ndez-Leo´n, S., Roy, C., Santos, A.M.P., 2006. Oceanography and fisheries of the Canary Current/ Iberian region of the eastern north Atlantic. In: Robinson, A.R., Brink, K.W. (Eds.), The Sea: Volume 14B—The Global Coastal Ocean: Interdisciplinary Regional Studies and Synthesis. Harvard University Press, pp. 877–931. Bashmachnikov, I., White, M., Mohn, C., Martins, A., Pelegrı´, J.L., Machı´n, F., 2009. Interaction of Mediterranean water eddies with Sedlo and Seine seamounts, subtropical North-East Atlantic. Deep-Sea Research II 56 (25), 2593–2605.

ARTICLE IN PRESS J. Arı´stegui et al. / Deep-Sea Research II 56 (2009) 2646–2655

Beckmann, A., Mohn, C., 2002. The upper ocean circulation at Great Meteor Seamount. Part II: Retention potential of the seamount-induced circulation. Ocean Dynamics 52, 194–204. Boehlert, G.W., Genin, A., 1987. A review of the effects of seamounts on biological processes. In: Keating, B.H., Fryer, P., Batiza, R., Borhlert, G.W. (Eds.), Seamounts, Islands and Atolls. Geophysical Monograph 43, American Geophysical Union, Washington, pp. 319–334. Brink, K.H., 1990. On the generation of seamount-trapped waves. Deep-Sea Research A 37, 1569–1582. Bryan, J.R., Riley, J.P., Williams, P.J.LeB., 1976. A procedure for making precise measurements of oxygen concentration for productivity and related studies. Journal of Experimental Marine Biology and Ecology 21, 191–197. Carrit, D.E., Carpenter, J.H., 1966. Comparison and evaluation of currently deployed modifications of the Winkler method for determining dissolved oxygen in seawater: a NASCO Report. Journal of Marine Research 24, 287–318. Comeau, L.A., Vezina, A.F., Bourgeois, M., Juniper, S.K., 1995. Relationship between phytoplankton production and the physical structure of the water column near Cobb Seamount, northeast Pacific. Deep-Sea Research I 42, 993–1005. Dortch, Q., Packard, T.T., 1989. Differences in biomass structure between oligotrophic and eutrophic marine ecosystems. Deep-Sea Research 36, 223–240. Dower, J., Freeland, H., Juniper, K., 1992. A strong biological response to oceanic flow past Cobb seamount. Deep-Sea Research 39, 1139–1145. Duarte, C.M., Agustı´, S., Arı´stegui, J., Gonza´lez, N., Anado´n, R., 2001. Evidence for a heterotrophic subtropical northeast Atlantic. Limnology and Oceanography 46, 425–428. Eriksen, C.C., 1991. Observations of amplified flows atop a large seamount. Journal of Geophysical Research 96, 15227–15236. Eriksen, C.C., 1998. Internal wave reflection and mixing at Fiberling Guyot. Journal of Geophysical Research 103, 2977–2994. Falkowski, P.G., 1981. Light-shade adaptation and assimilation numbers. Journal of Plankton Research 3 (2), 203–216. Freeland, H., 1994. Ocean circulation at Cobb seamount. Deep-Sea Research 41, 1715–1732. ˜ oz, M., Arı´stegui, J., Pelegrı´, J.L., Antoranz, A., Ojeda, A., Torres, M., 2005. Garcı´a-Mun Exchange of carbon and nutrients by an upwelling filament off Cape Ghir (NW Africa). Journal of Marine Systems 54, 83–96. Genin, A., 2004. Bio-physical coupling in the formation of zooplankton and fish aggregations over abrupt topographies. Journal of Marine Systems 50, 3–20. Genin, A., Boehlert, G.W., 1985. Dynamics of temperature and chlorophyll structures above a seamount—an oceanic experiment. Journal of Marine Research 43, 907–924. Genin, A., Noble, M., Lonsdale, P.F., 1989. Tidal currents and anticyclonic motions on two north Pacific seamounts. Deep-Sea Research 36, 1803–1815. Goldner, D.R., Chapman, D.C., 1997. Flow and particle motion induced above a tall seamount by steady and tidal background currents. Deep-Sea Research I 44, 719–744. ˜ o, B., Ferna´ndez, E., Sinha, B., Esca´nez, J., de Armas, Gonza´lez, N., Anado´n, R., Mourin D., 2001. The metabolic balance of the planktonic community in the North Atlantic Subtropical Gyre: the role of mesoscale instabilities. Limnology and Oceanography 46 (4), 946–952. ˜o´n, E., 2002. Large-scale variability of planktonic net Gonza´lez, N., Anado´n, R., Maran community metabolism in the Atlantic Ocean: importance of temporal changes in oligotrophic subtropical waters. Marine Ecology Progress Series 333, 21–30. Grasshoff, K., Ehrhardt, M., Kremling, K., 1983. Methods of Seawater Analysis. Verlag Chemie, Weinheim. Herna´ndez-Guerra, A., Nykjaer, L., 1997. Sea surface temperature variability off north-west Africa. International Journal of Remote Sensing 18, 2539–2558.

2655

Kunze, J.M., Toole, J.M., 1997. Tidally driven vorticity, diurnal shear, and turbulence atop Fieberling Seamount. Journal of Physical Oceanography 27, 2663–2693. Longhurst, A., 1998. Ecological Geography of the Sea. Academic Press, San Diego, CA, 398pp. Lowry, P.H., Rosenbrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with a Folin phenol reagent. Journal of Biological Chemistry 193, 265–275. Lueck, R.G., Mudge, T.D., 1997. Topographically induced mixing around a shallow seamount. Science 276, 1831–1833. Mohn, C., White, M., Bashmachnikov, I., Jose, F., Pelegrı´, J.-L., 2009. Dynamics at an elongated, intermediate depth seamount in the North Atlantic (Sedlo Seamount, 401200 N, 261400 W). Deep-Sea Research II 56 (25), 2582–2592. ˜ o, B., Fernadez, E., Serret, P., Harbour, D., Sinha, B., Pingree, R., 2001. Mourin Variability and seasonality of physical and biological fields at the Great Meteor Tablemount (sub tropical NE Atlantic). Oceanologica Acta 24, 1–20. Mullineaux, L.S., Mills, S., 1997. A test of the larval retention hypothesis in seamount-generated flows. Deep-Sea Research I 44, 745–770. Navatov, U.N., Ozmidov, R.V., 1988. A study of turbulence over underwater mounts in the Atlantic Ocean. Oceanology 28, 210–217. Noble, M., Mullineaux, L.S., 1989. Internal tidal currents over the summit of Cross Seamount. Deep-Sea Research 36, 1791–1802. Odate, T., Furuya, K., 1998. Well-developed subsurface chlorophyll maximum near Komahashi No. 2 Seamount in the summer of 1991. Deep-Sea Research I 45, 1595–1607. Owens, W.B., Hogg, N.G., 1980. Oceanic observations of stratified Taylor columns near a bump. Deep-Sea Research 27, 1029–1045. Pacheco, M.M., Herna´ndez-Guerra, A., 1999. Seasonal variability of recurrent phytoplankton pigment patterns in the Canary Islands area. International Journal of Remote Sensing 20, 1404–1418. Peliz, A., Santos, A.M.P., Oliveira, P.B., Dubert, J., 2004. Extreme cross-shelf transport induced by eddy interactions southwest of Iberia in winter 2001. Geophysical Research Letters 31, 1–4. Parsons, T.R., Maita, Y., Lalli, C.M., 1984. A Manual of Chemical and Biological Methods for Seawater Analysis. Pergamon Press, Oxford. Peterson, G.L., 1983. Determinations of total protein. In: Methods of Enzymology, vol. 91. Academic Press, pp. 95–119. Robinson, C., Williams, P.J.leB., 1993. Temperature and Antarctic plankton community respiration. Journal of Plankton Research 15 (9), 1035–1051. Robinson, C., Williams, P.J.leB., 2005. Respiration and its measurement in surface marine waters. In: del Giorgio, P.A., Williams, P.J.leB. (Eds.), Respiration in Aquatic Ecosystems. Oxford University Press, Oxford, pp. 147–180. Roden, G.I., 1987. Effects of seamounts and seamount chains on oceanic circulation and thermohaline strucuture. In: Keating, B.H., Fryer, P., Batiza, R., Borhlert, G.W. (Eds.), Seamounts, Islands and Atolls. Geophysical Monograph 43, American Geophysical Union, Washington, pp. 335–354. Rogers, A.D., 1994. The biology of seamounts. Advances in Marine Biology 30, 305–349. Royer, T.C., 1978. Ocean eddies generated by seamounts in the north Pacific. Science 199, 1063–1064. Uchida, R.N., Tagami, D.T., 1984. Groundfish fisheries and research in the vicinity of seamounts in the North Pacific Ocean. Marine Fisheries Review 46, 1–17. Vilas, J.C., Arı´stegui, J., Kiriakoulakis, K., Wolf, G.A., Espino, M., Polo, I., Montero, M.F., Mendonca, A., 2009. Seamounts and organic matter—is there an effect? The case of Sedlo and Seine seamounts; Part 1. Distributions of dissolved and particulate organic matter. Deep-Sea Research II 56 (25), 2618–2630. Williams, P.J.leB., Jenkinson, N.W., 1982. A transportable microprocessor-controlled precise Winkler titration suitable for field station and shipboard use. Limnology and Oceanography 27, 576–584.