m
Q. Devresse et al.: Eddy-enhanced primary production sustains heterotrophic microbial activities
mary production and carbon export (Cheney and Richard- within eddies may thus lead to spatial heterogeneity of ex-
son, 1976; Aristegui et al., 1997). The sense of rotation and tracellular release rates (e.g. Lasternas et al., 2013; Rao et
their vertical structure classify cyclonic (CEs), anticyclonic al., 2021) and DOM quality (e.g. Wear et al., 2020). DOM
(ACEs; e.g. Chelton et al., 2011), and anticyclonic mode wa- quality impacts bacterial biomass production (BP), bacte-
ter eddies (ACMESs; D’Asaro, 1988). In eastern boundary up- rial respiration (BR), and bacterial growth efficiency (BGE;
welling systems (EBUS), eddies typically form by flow sep- e.g. Neijssel and de Mattos, 1994; Russell and Cook, 1995;
aration along slope boundary currents at topographic head- Robinson, 2008; Lipson, 2015). BGE is the ratio between
lands (D’Asaro, 1988; Molemaker et al., 2015; Thomsen et BP and the bacterial carbon demand (BCD), which is the
al., 2016). Eddies have lifespans from days to months and sum of respired carbon and carbon incorporated into biomass
can travel several hundred to thousands of kilometres across (BP +BR). Lgnborg et al. (2011) observed that BGE de-
ocean basins (Chelton et al., 2011). In the North Atlantic creases with the increasing C/N ratio of phytoplankton-
Ocean, eddies generated in the highly productive Canary Up- derived DOM. BGE is a critical parameter for estimating the
welling System (CanUS) may laterally propagate to the olig- amount of consumed organic carbon used to build biomass
otrophic Subtropical North Atlantic Gyre (SNAG), transport- by heterotrophic bacteria (Anderson and Ducklow, 2001). So
ing nutrients and carbon from the coast to the open ocean far, BGE has been reported for ACEs from the Mediterranean
(McGillicuddy et al., 2003; Karstensen et al., 2015; Schütte Sea (Christaki et al., 2021) but not for CEs and ACMEs. In
et al., 2016). Various studies demonstrated the impact of ed- general, several studies showed a patchy distribution of bac-
dies on primary production (PP) on a global scale. However, terial abundance, BP (Ewart et al., 2008; Baltar et al., 2010),
che effects of eddies vary regionally, and studies with higher BR (Mourifo-Carballido, 2009; Jiao et al., 2014), commu-
spatial resolution of eddies combined with advances in in situ nity respiration (CR; Mourifio-Carballido and McGillicuddy,
observation, remote sensing, and modelling are still needed 2006; Mourifno-Carballido, 2009), and the metabolic balance
to better describe the physical and biological properties of between the production and consumption of organic mat-
the upper ocean (see review by McGillicuddy, 2016, and ref- ter (Maixandeau et al., 2005; Ewart et al., 2008; Mourifo-
erences therein). For example, Couespel et al. (2021) per- Carballido and McGillicuddy, 2006; Mourifo-Carballido,
formed global warming simulations using a representation 2009) within eddies. Yet, insights into the distribution of phy-
of mid-latitude double-gyre circulation. They showed that at toplankton and their activities within mesoscale eddies are
the finest model resolution (1/27°), eddies can mitigate the limited due to insufficient fine-scale vertical- and horizontal-
decline in primary production (—12 % at 1/27° vs. —26 % at resolution studies to adequately describe these distributions.
1°). Modelling studies have long urged consideration of the Thus, data on eddy-induced changes in primary production,
effects of eddies on PP at submesoscale levels (0.1-10 km) extracellular release, and semi-labile DOM concentration, as
to provide more realistic estimates of the oceanic carbon cy- well as the responses of heterotrophic microbial metabolic
cle (Levy et al., 2001). Eddies modulate the mixed layer activities, are scarce. Understanding how eddies modulate
depth by upwelling (CEs), downwelling (ACESs), or fronto- microbial activities will enhance our knowledge about the
genesis from eddy—eddy interaction, thereby creating spatial fate of organic carbon and the overall CO, source/sink func-
varlability in nutrient concentration within and around ed- tion in the ocean, particularly in EBUS, where eddy genera-
dies on the submesoscale (see reviews by Mahadevan, 2016, tion is high (Pegliasco et al., 2015).
and McGillicuddy, 2016). In addition, the nonlinear response Here, we studied the impact of a CE on microbial carbon
of phytoplankton growth to nutrient availability and advec- cycling along a 900 km zonal corridor of the westward prop-
yon of phytoplankton by currents makes plankton distribu- agating eddies between the Cape Verde islands and the Mau-
tion and community composition highly variable within and ritania Upwelling System (13-20° N), a sub-region of the
around eddies (Lochte and Pfannkuche, 1987). As a conse- CanUS (13-33° N; Aristegui et al., 2009). About 146 + 44
quence, the spatial distribution of PP across eddies can be eddies with a lifetime of more than 7d are generated per
highly variable (e.g. Falkowski et al., 1991; Ewart et al.. year in this region (Schütte et al., 2016). Along this corridor,
2008; Singh et al., 2015). a CE was sampled at high spatial resolution to resolve the
Bacterial activity is directly coupled to PP, as autotrophic heterogeneity of microbial processes at the submesoscale.
cells release their main substrate dissolved organic matter We determined phytoplankton (<20 um) cell abundance, pri-
(DOM). DOM release by phytoplankton mainly occurs via mary production, and extracellular release and linked those
two mechanisms: (1) passive leakage of small molecules by measurements of autotrophic activity to semi-labile DOM
diffusion across the cell membrane and (2) active exuda- concentration and heterotrophic bacterial activity. Our study
tion of DOM into the surrounding environment (Engel et al., provides new insights into (1) microbial carbon cycling and
2004). Environmental conditions, such as temperature, nu- (2) factors controlling microbial metabolic activities within
trient availability (e.g. Borchard and Engel, 2012), and light and around CEs formed in EBUS.
conditions (e.g. Cherrier et al., 2015) affect the amount and
the elemental stoichiometry of released DOM. Patchiness of
phytoplankton primary productivity and nutrient availability
Biogeosciences, 19, 51995219. 2027
https://doi.org/10.5194/bg-19-5199-2022
