Description: Biomass of zooplankton taxa in µg DM per liter as determined by ZooScan, using published area to dry weight relationships (Lehette & Hernandez-Leon 2009). Each data point is one sampling day (date) in one mesocosm (MK). For details on experimental treatments and sampling, refer to Bach et al. 2021 (https://doi.org/10.5194/bg-17-4831-2020) and Ayon et al. 2022 (https://doi.org/10.5194/bg-2022-157). Raw images are stored in https://ecotaxa.obs-vlfr.fr/prj/3784. All taxonomic categories are self-expanatory.

Description: Zooplankton species/groups abundance table per mesocosm and sampling day. Abundances are given as individual per m-3 and individuals per liter.

Description: Fatty acid composition data for the two dominant copepods in the mesocosms (Paracalanus sp. and Hemicyclops sp.).

Description: Gut fluorescence and C/N ratio of Paracalanus sp. determined during two occassions during the mesocosm experiment (Sampling Day 21/22 and 34/35).

Description: The Humboldt Current Upwelling System (HCS) is the most productive eastern boundary upwelling system (EBUS) in terms of fishery yield on the planet. EBUSs are considered hotspots of climate change with predicted expansion of mesopelagic oxygen minimum zones (OMZs) and related changes in the frequency and intensity of upwelling of nutrient-rich, low-oxygen deep water. To increase our mechanistic understanding of how upwelling impacts plankton communities and trophic links, we investigated mesozooplankton community succession and gut fluorescence, fatty acid and elemental compositions (C, N, O, P), and stable isotope (δ13C, δ15N) ratios of dominant mesozooplankton and microzooplankton representatives in a mesocosm setup off Callao (Peru) after simulated upwelling with OMZ water from two different locations and different N:P signatures (moderate and extreme treatments). An oxycline between 5 and 15 m with hypoxic conditions (〈50 µmol L−1) below ∼10 m persisted in the mesocosms throughout the experiment. No treatment effects were determined for the measured parameters, but differences in nutrient concentrations established through OMZ water additions were only minor. Copepods and polychaete larvae dominated in terms of abundance and biomass. Development and reproduction of the dominant copepod genera Paracalanus sp., Hemicyclops sp., Acartia sp., and Oncaea sp. were hindered as evident from accumulation of adult copepodids but largely missing nauplii. Failed hatching of nauplii in the hypoxic bottom layer of the mesocosms and poor nutritional condition of copepods suggested from very low gut fluorescence and fatty acid compositions most likely explain the retarded copepod development. Correlation analysis revealed no particular trophic relations between dominant copepods and phytoplankton groups. Possibly, particulate organic matter with a relatively high C:N ratio was a major diet of copepods. C:N ratios of copepods and polychaetes ranged 4.8–5.8 and 4.2–4.3, respectively. δ15N was comparatively high (∼13 ‰–17 ‰), potentially because the injected OMZ source water was enriched in δ15N as a result of anoxic conditions. Elemental ratios of dinoflagellates deviated strongly from the Redfield ratio. We conclude that opportunistic feeding of copepods may have played an important role in the pelagic food web. Overall, projected changes in the frequency and intensity of upwelling hypoxic waters may make a huge difference for copepod reproduction and may be further enhanced by varying N:P ratios of upwelled OMZ water masses.

Description: Species identification is pivotal in biodiversity assessments and proteomic fingerprinting by MALDI-TOF mass spectrometry has already been shown to reliably identify calanoid copepods to species level. However, MALDI-TOF data may contain more information beyond mere species identification. In this study, we investigated different ontogenetic stages (copepodids C1–C6 females) of three co-occurring Calanus species from the Arctic Fram Strait, which cannot be identified to species level based on morphological characters alone. Differentiation of the three species based on mass spectrometry data was without any error. In addition, a clear stage-specific signal was detected in all species, supported by clustering approaches as well as machine learning using Random Forest. More complex mass spectra in later ontogenetic stages as well as relative intensities of certain mass peaks were found as the main drivers of stage distinction in these species. Through a dilution series, we were able to show that this did not result from the higher amount of biomass that was used in tissue processing of the larger stages. Finally, the data were tested in a simulation for application in a real biodiversity assessment by using Random Forest for stage classification of specimens absent from the training data. This resulted in a successful stage-identification rate of almost 90%, making proteomic fingerprinting a promising tool to investigate polewards shifts of Atlantic Calanus species and, in general, to assess stage compositions in biodiversity assessments of Calanoida, which can be notoriously difficult using conventional identification methods.