<rahmann et al. 
using automated imaging software (Gorsky et al., 2010) 
allowing taxonomical classification as well as the estimation of 
taxon-specific biomass (Lehette and Hernändez-Leön, 2006) and 
metabolic rates. Scanned image data are available on EcoTaxa'* 
(Picheral et al., 2017) upon request from R. Kiko and H. Hauss. 
Taxon-specific biomass and metabolic rate estimates are publicly 
available on PANGAEA (Kiko and Hauss, 2019; Kiko et al., 2020). 
To expand the ecological knowledge on fragile organisms 
(such as giant rhizaria, medusae, <ctenophores, and 
siphonophores) in situ imaging techniques were employed 
in addition to net sampling. An Underwater Vision Profiler 
5 (UVP5; Picheral et al., 2010) was routinely mounted on the 
CTD/rosette during most SFB 754 cruises since 2012 (Kiko 
zt al., 2021a; see Table 2 and Supplementary Table 21). During 
the cruises in 2012 and 2013 a UVP5 was used that was kindly 
provided by the Laboratoire d’Oceangraphie de Villefranche-sur- 
Mer (France). The instrument consists of one down facing HD 
camera in a steel pressure case and two red LED lights which 
illuminate a 0.88 to 0.93 1 volume (depending on the actual 
set-up). During the downcast, the UVP5 takes 3-20 pictures of 
the illuminated field per second. For each picture, the particles 
are counted and sized immediately and the data are stored in the 
instrument for later analysis. Furthermore, images of particles 
with a size >500 wm are saved as separate “vignettes” — small 
cut-outs of the original picture - which allow for later, computer 
assisted, identification of these particles and their assignment 
into different particle, phyto-, and zooplankton groups. Since the 
JVP5 was integrated in the CTD and has its own pressure sensor, 
fine-scale vertical distribution of particles and major planktonic 
groups can be related to environmental data. UVP5 particle and 
zooplankton data from all cruises can be accessed on EcoTaxa 
(see text footnote 11; Picheral et al., 2017). UVP5 particle 
data have undergone further quality controls since their first 
publication and were merged with data from other international 
collaborators to yield a global data set. This data set, to be found 
at doi: 10.1594/PANGAEA.924375 supersedes the previous 
UVP5 particle data sets and should be used for further research, 
whereas the original data sets are still available for reference. 
Zooplankton Metabolic Rates 
Zooplankton metabolic rates (oxygen respiration and 
ammonium excretion) at different temperatures, oxygen, and 
carbon dioxide partial pressures (Kiko et al., 2015, 2016) were 
measured during three cruises (Kiko et al., 2021b; see Table 2 
and Supplementary Table 22). Zooplankton was collected by 
different nets and the entire catch immediately transferred to 101 
beakers containing pre-cooled seawater. Diel vertical migrators 
were sampled at the surface at night. Individuals for respiration 
rate measurements were isolated immediately and maintained 
in filtered seawater for 1 to 13 h at the chosen experimental 
temperature (13, 18, or 23°C). Only animals appearing unharmed 
and fit were used for experiments. Water for the respiration 
and excretion rate trials was UV-treated, filtered over a 0.2 um 
sterile filter, and supplemented with antibiotics (25 mg 17! 
ampicillin and 25 mg 17! streptomycin). Subsequently, the 
1https://ecotaxa.obs-vlfr.fr/ 
-rontiers in Marine Science | www. frontiersin.orm 
SFB754 Data Legacy 
water was bubbled with different Gas mixtures (N, O2, CO»z; 
see Kiko et al., 2016 for details) adjusted to represent different 
environmental pO, and pCO-„ levels. Incubation bottles (12 to 
280 ml) were pre-filled with the respective incubation water 
and the animals quickly added, transferring as little water as 
possible. The incubation bottles were equipped with a PreSense 
oxygen microsensor spot and readout was conducted from 
:he outside, using a fiber optic cable and a 4- or 10-channel 
Oxy-Mini (PreSens Precision Sensing GmbH, Regensburg, 
Germany). Incubations were conducted in the dark in 10 I 
water baths located inside temperature-controlled incubators. 
Experiments were generally conducted for a maximum of 
16 h to avoid microbial growth, which would have affected the 
ammonium measurements. Generally, three incubations were 
combined with one animal free control incubation, which served 
to estimate microbial background respiration and ammonium 
concentrations in these controls. As oxygen levels within the 
bottles declined, respiration rates could also be estimated at 
other than the pre-set conditions. After an acclimation phase 
of 1 h, respiration rates were calculated for 1-h intervals using 
a linear regression. The microbial background respiration rate 
was subtracted from the experimental incubation respiration 
rate to yield the animal’s respiration rate. Generally, 1 or 15 ml 
water samples were collected at the end of the incubation 
to determine ammonium concentrations fluorometrically 
according to Holmes et al. (1999). Ammonium excretion rates 
were calculated as the difference between the incubation and 
animal-free controls. Animals used in the experiments were 
afterward recovered, frozen at —80°C, and transported to the 
home laboratory, where their dry-weight was determined. The 
rates presented should be considered routine metabolic rates, as 
activity was not monitored continuously (Prosser, 1961). Please 
refer to Kiko et al. (2015, 2016) for further experimental details. 
Nutrient Amendment Experiments 
Bioassays with amendment of DIN, DIP, and various trace 
2lements were conducted in short-term replicated bottle 
ncubations to determine limiting elements for phytoplankton 
growth (Browning et al., 2017; Hauss et al., 2021b; see Table 2 and 
Supplementary Table 23). Shipboard mesocosm experiments 
with a duration from 7 to 11 days were conducted on several 
cruises in the ETNA and ETSP and land-based on Cape Verde 
to determine the impact of N:P stoichiometry on the pelagic 
community (Franz et al., 2012a; Hauss et al., 2012; Czerny et al., 
2016; Meyer et al., 2016) and dissolved organic compounds 
(Engel et al., 2015; Loginova et al., 2015). In austral summer 
2017, a large-scale in situ mesocosm experiment was conducted 
off Callao (Peru) using the KOSMOS facilities. Deep water was 
injected into the mesocosms to simulate an upwelling event and 
che response of the planktonic ecosystem was monitored for 
50 days (Bach et al., 2020). 
Paleoceanography 
One of the objectives of the SFB 754 was the reconstruction 
of the factors controlling the intensity and the spatial extent 
of the OMZ in the Eastern Tropical Pacific, specifically off 
Peru, since the Last Glacial Maximum (21000 years ago). For 
3eptember 2021 | Volume 8 | Article 72330«