<rahmann et al. 
(19°C) using Shimadzu® 1800 UV—VIS double—-beam 
spectrophotometer within 230-750 nm wavelength range against 
MilliQ water at 1 nm intervals. More details on the spectroscopic 
analyses may be found in Loginova et al. (2015, 2016, 2020). 
Microbial Oxygen Consumption, Nitrogen 
Transformation, and Primary Productivity Rates 
Jinitrogen (N2) and carbon (C) fixation rates were measured 
an nine cruises (Löscher and Mehrtens, 2021b; see Table 2 
and Supplementary Table 17) using shipboard incubation 
experiments, complemented with nutrient and oxygen 
manipulations. During cruises M77/3, M77/4, and M80/2, 
N2 fixation was measured using the bubble addition method 
following Montoya et al. (1996). During M80/2 a novel method 
based on !°N; gas pre-dissolution, which was developed by 
Mobr et al. (2010), was tested in parallel to the method of 
Montoya et al. (1996). An underestimation of N fixation rates 
by the classic method has been observed (Großkopf et al., 2012) 
and therefore the novel ‘pre-dissolution method’ was applied 
during the following cruises (M83/1, M90, M91, M93, M97, 
M104, and M107). Single cell N, fixation rates to differentiate 
the contribution of different clades of N, fixers were measured 
using a NanoSIMS (Martinez-Perez et al., 2016). C fixation was 
determined using !*C-bicarbonate additions (e.g., Großkopf 
at al., 2012; Löscher et al., 2014) and heterotrophic C turnover 
was determined using !*C-glucose additions (Löscher et al., 
2014, 2016). 
Potential rates for microaerobic respiration and aerobic 
organic matter degradation as a source of ammonia (NH4") in 
the Peruvian OMZ was assessed using an !®O>; labeling approach 
suitable for microaerobic respiration (Holtappels et al., 2014). 
Further, the effects of O» depletion associated with marine snow 
particles on microbial respiration was explored by combining 
802 labeling experiments with in situ particle size analysis 
and modeling of aggregate-size dependent respiration (Kalvelage 
2t al., 2015). Anammox, denitrification, and nitrification, as 
well as NO production rates were measured on several cruises 
(Kalvelage et al., 2011; Löscher et al., 2012; Bourbonnais et al., 
2017; Callbeck et al., 2017; Frey et al., 2020; Löscher and 
Mehrtens, 2021a; see Table 2 and Supplementary Table 18) 
using isotope fractionation studies, N tracer additions, and 
inhibitor studies. 
Marine Microbial Diversity Function 
[n order to identify key groups of microbes for C, N, 
and O>» turnover, microbial metabolic rate measurements 
were complemented with analyses of metagenomes and 
metatranscriptomes from the Eastern Tropical South Pacific 
(ETSP) and Eastern Tropical North Atlantic (ETNA). In addition, 
key gene and transcript characterization and quantification for 
aerobic respiration (Kalvelage et al., 2015), N2 fixation (Großkopf 
zt al., 2012; Löscher et al., 2014, 2015, 2016, 2020), anammox, 
denitrification, and nitrification (Löscher et al., 2012, 2015, 2016; 
Kalvelage et al., 2013) were carried out using Sanger sequencing 
and quantitative real time polymerase chain reactions (PCRs) as 
described in Löscher et al. (2012, 2014). To assure high quality 
sampling of nucleic acids, sample filtration times did not exceed 
-rontiers in Marine Science | www.frontiersin.orc 
SFB754 Data Legacy 
20 min and samples were shock-frozen in liquid N, and stored 
at —80°C (e.g., Löscher et al., 2014). Early metagenomic and 
-transcriptomic analyses targeted an understanding of microbial 
communities in the surface waters above the OMZ, the oxyclines, 
OMZ core waters, and sulfidic anoxic waters, as summarized 
in Löscher et al. (2016) and were based on Pyrosequencing 
technology (e.g., Desai et al., 2013; Schunck et al., 2013). Due to 
‘he rapid advance in sequencing technologies, it was possible to 
generate more conclusive metagenomes for targeted studies on 
sulfur, N, and O, cycling during M90-M93. Nine metagenomes 
were sequenced using Illumina HiSeq technology (Callbeck et al., 
2018) from those cruises. On those data sets, genome assemblies 
and phylogenetic classifications were carried out to explore the 
role of a key microbial cluster, SUP05, and its role in OMZ 
sulfur and nitrogen turnover. Metagenomes from the ETNA 
cruise M107 were sequenced in the context of the development 
of anoxic water masses in collaboration with the DFG-funded 
Cluster of Excellence “The Future Ocean’ (Löscher et al., 2015). In 
addition to full metagenomes, targeted community studies were 
carried out using 16S rDNA amplicon sequencing sequenced 
on Illumina MiSeq sequencers from the same anoxic eddy in 
the ETNA and from the Peruvian OMZ (Löscher et al., 2015; 
Scholz et al., 2016). All published sequences were submitted to 
ihe National Center for Biotechnology Information’s archives 
(NCBI'; see Table 2 and Supplementary Table 19). Physical 
DNA libraries are archived at GEOMAR and the University of 
Southern Denmark, and subsamples are available on request 
from C. Löscher. 
In addition to this mainly pelagic work, transcriptomes, and 
genomes of the denitrifying benthic foraminifera Globobulimina 
turgida and G. auriculata from the seasonally hypoxic Swedish 
Gullmar Fjord were analyzed (Woehle et al., 2018). The 
obtained information was used to describe the foraminifera 
unique eukaryotic ability to denitrify and colonize low-oxygen 
environments. Sequences were submitted to the NCBI's 
Sequence Read Archive (accession numbers SRR6202052 - 
S$RR6202078) and to the transcriptome sequencing archive 
(accession numbers GGCE00000000 and GGCD00000000). 
The genome assembly was submitted to NCBI (draft 
genomes PIVH00000000- PIWH00000000; unassigned contigs: 
PJEL00000000). Furthermore, individually amplified 18S$ rRNA 
gene sequences of the two analyzed foraminiferal species were 
submitted to GenBank (MG800664 to MG800667). 
Zooplankton and Particle Distribution 
A Hydrobios Multinet Midi with an aperture of 0.25 m? 
and five nets (mesh size 200 wm) was deployed for vertically 
stratified tows on several cruises (Hauss et al., 2021la; see 
Table 2 and Supplementary Table 20), mostly in paired day- 
night tows to quantify diel vertical migration. Standard depths 
used for these deployments were 1000-600-300-200-100-0 m. 
On cruise M93, a Multinet Maxi (nine nets, 333 wm mesh) 
was used instead. Samples were fixated in 4% formaldehyde 
in seawater solution, scanned at GEOMAR or at the Ocean 
Science Center Mindelo, Mindelo/Cape Verde, and analyzed 
Ohttps://www.ncbi.nlm.nih.gov/ 
deaptember 2021 | Volume 8 | Article 72830-