Q. Devresse et al.: Eddy-enhanced primary production sustains heterotrophic microbial activities
5201
2 Materials and methods
face temperature and salinity (not compensated by density)
different from the adjacent stations (Fig. 1b). Hence, we re-
ferred to that station as “frontal zone”. The classification of
stations is thoroughly discussed in the Supplement, and the
sampling time, location, and distance from the eddy centre
are given in Table S1.
2.2 Chemical analyses
2.1 Study area and eddy characterisation
Sampling was conducted in the ETNA between the Cape
Verde archipelago and the Mauritanian coast during cruise
M156 (3 July to 1 August 2019; Fig. 1a) on the R/V Meteor.
Samples were collected during the relaxation period, which
Is typically from May to July following the upwelling sea-
son (January to March; Lathuiliere et al., 2008). A CE was
sampled at high spatial resolution along two zonal transects
(from 19.1 to 18.2° W at 18.3° N and from 18.5 to 17.1° W
at 18.6° N) and one meridional transect (from 19.4 to 18° N
at 18.4 to 18.1° W). The zonal transect slightly shifted east-
west of the eddy core position. The reason for that was the de-
formed eddy shape (see Fig. la), which made it challenging
to identify the centre of the eddy and required rerouting of the
ship’s track during the survey. In addition, we sampled water
along an 18° N transect, a typical coast to open ocean trajec-
tory of eddies in this region (Schütte et al., 2016). Salinity,
:emperature, depth, and O0, concentration were determined
using a Seabird 911 plus CTD system equipped with two
independently working sets of temperature-conductivity—
oxygen sensors. The oxygen sensor was calibrated against
discrete water samples using the Winkler method (Strick-
land and Parsons, 1968; Wilhelm, 1888). Seawater samples
were collected using 10L Niskin bottles attached to the CTD
Rosette. A total of 25 stations (Table S1) were sampled, 14
of them inside or in the vicinity of the CE. Sampling was
conducted in the epipelagic layer (0-200 m), including sam-
ples from the surface mixed layer, the Chl @ maximum, and
che shallow oxygen minimum zone (OMZ; < 50 umolkg 7!
vetween 0-200 m depth) when present.
Sea surface height (SSH) and acoustic Doppler current
profiler (ADCP) velocity data (Fig. 1) characterised the eddy
as a CE. Based on the angular momentum eddy detec-
tion and tracking algorithm (AMEDA; Le Vu et al., 2018),
che eddy was estimated to be 1.5 months old. The cen-
tre of the eddy and the core radius were determined using
ADCP reconstructions assuming an axis-symmetric vortex
(Fig. 1). On 22 July 2019, the eddy centre was located at
18.69° N, 18.05° W, with a core radius of 40.5 + 5.7 km. The
mean azimuthal velocity in the CE was 19.94 0.7cms”!,
and the absolute dynamic topography (ADT) associated
with the CE core was —23cm on 23 July 2019. How-
ever, as the eddy shape was deformed, the ADCP recon-
struction did not constrain well the physical border of the
eddy (Fig. S1). Therefore, we combined sea surface temper-
ature (23.44 + 0.47 °C), salinity (39.95 + 0.04), and Chl a
(1.35 + 0.73 ug L7!) data to approximate the area influenced
by the eddy (Fig. 1b, c, d). We classified stations into “core”
and “periphery” of the eddy. Stations that were outside and
westward of the eddy influence were referred to as “open
ocean” and those close to the coast as “coastal”. Just beyond
the eddy periphery, at St. E3, a front was observed with sur-
Nutrient concentrations were determined at selected stations
(Table S1). Nutrients were measured on board from duplicate
unfiltered seawater samples (11 mL). Ammonium (NH?)
was analysed based on Solörzano (1969) and phosphate
(POa4), nitrate (NO; ), nitrite (NO, ), and silicate (Si(OH)4)
were measured photometrically with continuous-flow anal-
ysis on an auto-analyser (QuAAtro; Seal Analytical) based
on Hansen and Koroleff (1999). Detection limits for NH7.
PO4, NO3, NO», and Si(OH)4 were 0.1, 0.02, 0.1, 0.02,
and 0.2 umol L7!, respectively. Dissolved inorganic nitrogen
(DIN) was calculated as the sum of NH7, NO; , and NO; .
To estimate the fraction of semi-labile dissolved organic
carbon (SL-DOC), we determined high-molecular-weight
(HMW > 1kDa) dissolved combined carbohydrates (dC-
CHO) and dissolved hydrolysable amino acids (dHAA) as
the main biochemical components of DOM (Carlson, 2002).
For dCCHO analysis, duplicate samples (20mL) were fil-
tered through 0.45 um Acrodisc filters, collected in com-
busted glass vials (8h, 450°C) and frozen (—20°C) until
analysis based on Engel and Händel (2011) with a detection
limit of 1ug L7!. The analysis detected 11 monomers: ara-
binose, fucose, galactose, galactosamine, galacturonic acid,
glucosamine, glucose, glucuronic acid, rhamnose, co-elute
mannose, and xylose. For dHAA analysis, duplicate sam-
ples (4mL) were filtered through 0.45 um Acrodisc fil-
ters, collected in combusted glass vials (8h, 450 °C), and
frozen (—20°C) until analysis. After in-line derivatisation
(2min) with o-phthaldialdehyde and mercaptoethan, dHAA
were separated by high-performance liquid chromatography
(HPLC; Agilent Technologies, USA) and detected fluoro-
metrically. The HPLC was equipped with a C1i8 column (Phe-
nomenex, USA) (Lindroth and Mopper, 1979; Dittmar et
al., 2009). The analysis classified 13 monomers with a pre-
cision <5% and a detection limit of 2nmolL7!: alanine,
arginine, aspartic acid, isoleucine, glutamic acid, glycine,
leucine, phenylalanine, serine, threonine, tyrosine, valine.
and y-aminobutyric acid (GABA). The calculations for the
carbon content of dCCHO and dHAA were based on carbon
atoms contained in the identified monomers. The sum of dC-
CHO and dHAA carbon content is referred to as SL-DOC.
For Chl a, 1 L seawater samples were filtered onto 25 mm
GF/F filters (0.7 um pore size, Whatman, GE Healthcare
Life Sciences, UK) and subsequently frozen (—20 °C) un-
til extraction using 90% acetone for photometric analyses
(Turner Designs, USA) slightly modified based on Evans et
al. (19877.
https://doi.org/10.5194/bg-19-5199-202.
Biosgeosciences. 19. 5199-5219. 2022
