Q. Devresse et al.: Eddy-enhanced primary production sustains heterotrophic microbial activities 
5213 
0.83 and 0.76, respectively, p<0.001). However, BR was not 
significantly correlated to autotrophic pico- and nanoplank- 
ton biomass, PPrTorT, and PPpoc (r = —0.05, 0.61, and 0.50, 
respectively, p>0.05). 
4 Discussion 
4.1 
Effect of a cyclonic eddy on the distribution of 
phytoplankton abundance and activity in the 
Mauritanian upwelling system 
In general, coastal Chl @ concentration during this study 
was not as high as observed in earlier studies with strong 
coastal upwelling (e.g. Alonso-Säez et al., 2007; Agusti and 
Duarte, 2013; Aristegui et al., 2020). This might be related 
to the relatively weak upwelling resulting from weak surface 
winds along the Mauritanian coast typically occurring dur- 
ing summer when our samples were collected (Pelegri and 
Pefa-Izquierdo, 2015). Consequently, during summer, fewer 
nutrients reach the euphotic zone. At the same time, off- 
shore surface wind remained strong, enhanced vertical mix- 
ıng, and may explain why coastal Chl a concentration was 
only slightly higher compared to the open ocean. When ex- 
cluding the eddy-influenced stations, there was no marked 
gradient in phytoplankton productivity either, unlike other re- 
gions of the CanUS (Demarcq and Somoue, 2015; Arfstegui 
et al., 2020). PPror and PPpoc rates stayed rather constant 
from the coast to the open ocean and were in the range of 
reported rates in oligotrophic offshore waters of the CanUS 
(Agustf and Duarte, 2013; Lasternas and Agustf, 2014). The 
spatial distribution of SL-DOC was relatively uniform as 
well when considering the coastal and open ocean stations 
only. PER in our study was on average 51.1 +17 % in both 
the open ocean and the coastal stations, which is in contrast 
to previous findings. For example, Agustf and Duarte (2013) 
reported PER to range from — 1 % in “healthy” communities 
from the upwelled waters of the CanUS to — 70 % in “dy- 
ing” communities from the oligotrophic waters of the ETNA. 
PER values have been reported to increase with nutrient de- 
pletion (Obernosterer and Herndl, 1995; Agustf and Duarte, 
2013; Lasternas et al., 2014; Piontek et al., 2019) among 
other factors (see review by Mühlenbruch et al., 2018). Since 
upwelling was weak during our sampling period, low nutri- 
ent concentrations in the surface waters might explain the 
relatively high PER that we observed near the coast. 
The CE broke this rather uniform distribution of phyto- 
plankton productivity from the coastal to the open ocean wa- 
‚ers. Chl a isolines were pushed towards the surface in the 
CE (Fig. 4a). A similar uplifting of Chl a isolines towards 
che surface has been reported for other eddies (Lochte and 
Pfannkuche, 1987; Feng et al., 2007; Noyon et al., 2019) 
and might result from phytoplankton relocation through in- 
tense vertical mixing by strong surface winds (Feng et al., 
2007; Noyon et al., 2019). Before our eddy survey, strong 
https://doi.org/10.5194/bg-19-5199-2027 
{a} 
A 
RE 
Rn 
Depth 
Temperature 
DIN 
PO4 fx 
SICOH)4 
SL-DOC 
Chla 
AutPI 
PPror 
PPosc 
HB 
BR 
BP 
3 
@ ST 
Ei 
Open ocean 
+ Coastal 
& 
za 
a 8 
m 
Mohr 
A 
rare. PL 
a 
(b) 
Fi 
$ 
Depth 
Temperalure 
DIN Be 
PO, 
SICOH), ME“ 
SL-DOC 
Chla ** 
AutPt Ba 
P| Pror 
PPose 
HB 
BR 
BP 
„o* 
Pd 
Cyclonic eddy 
+ Frontal zone 
S 
T 
5 
Sr ZO 
x N 
KT 
Oo 
bb 
& 
gi 
> 
N 
rn 
* 
* 
FF. — 
. 
Y 
FF 
Be 
0.8 
06 -04 -02 0 02 04 06 
Correlation coefficient (r) 
08 
Figure 7. Pearson correlation matrix of biochemical parameters, 
metabolic activities, and bacterial abundance in the upper 100 m in 
samples not influenced by the cyclonic eddy (ie. coastal and open 
ocean stations) (a) and samples influenced by the cyclonic eddy (b). 
Statistical significance: *** <0.001, ** <0.01, * <0.05. 
Biogeosciences, 19, 51995219, 2022