Description: Eastern boundary upwelling systems (EBUS) are among the most productive marine ecosystems on Earth. The production of organic material is fueled by upwelling of nutrient-rich deep waters and high incident light at the sea surface. However, biotic and abiotic factors can mod- ify surface production and related biogeochemical processes. Determining these factors is important because EBUS are considered hotspots of climate change, and reliable predic- tions of their future functioning requires understanding of the mechanisms driving the biogeochemical cycles therein. In this field experiment, we used in situ mesocosms as tools to improve our mechanistic understanding of processes con- trolling organic matter cycling in the coastal Peruvian up- welling system. Eight mesocosms, each with a volume of ∼ 55 m3, were deployed for 50 d ∼ 6 km off Callao (12◦ S) during austral summer 2017, coinciding with a coastal El Niño phase. After mesocosm deployment, we collected sub- surface waters at two different locations in the regional oxy- gen minimum zone (OMZ) and injected these into four meso- cosms (mixing ratio ≈ 1.5 : 1 mesocosm: OMZ water). The focus of this paper is on temporal developments of organic matter production, export, and stoichiometry in the indi- vidual mesocosms. The mesocosm phytoplankton commu- nities were initially dominated by diatoms but shifted to- wards a pronounced dominance of the mixotrophic dinoflag- ellate (Akashiwo sanguinea) when inorganic nitrogen was exhausted in surface layers. The community shift coincided with a short-term increase in production during the A. san- guinea bloom, which left a pronounced imprint on organic matter C : N : P stoichiometry. However, C, N, and P export fluxes did not increase because A. sanguinea persisted in the water column and did not sink out during the experiment. Accordingly, export fluxes during the study were decou- pled from surface production and sustained by the remain- ing plankton community. Overall, biogeochemical pools and fluxes were surprisingly constant for most of the experiment. We explain this constancy by light limitation through self- shading by phytoplankton and by inorganic nitrogen limita- tion which constrained phytoplankton growth. Thus, gain and loss processes remained balanced and there were few oppor- tunities for blooms, which represents an event where the sys- tem becomes unbalanced. Overall, our mesocosm study re- vealed some key links between ecological and biogeochem- ical processes for one of the most economically important regions in the oceans.

Description: This data is part of the BMBF project CUSCO (Coastal Upwelling Systems in a Changing Ocean). Here we report the dissolved inorganic carbon concentration and total alkalinity during a 35-day experiment, where we enclosed natural plankton communities in in-situ mesocosms off Peru. The experiment investigated the interactive effects of light and upwelling on the Humboldt upwelling ecosystem by mimicking a gradient of upwelling intensities (0%, 15%, 30%, 45% and 60%) under summer-time high light and winter-time low light. Integrated seawater samples from a depth between 0 and 10m were collected using a 5L Integrating Water sampler (IWS; Hydro-Bios, Kiel). Dissolved inorganic carbon (DIC) and total alkalinity (TA) samples were obtained by 0.2µm gentle pressure filtration, poisoned with saturated 7.5 % mercury chloride (HgCl2) solution and frozen at -20°C until measurement. Samples for Total Alkalinity (TA) were measured by means of potentiometric titration with 0.05 M HCl using an automated titration device (862 Metrohm Compact Titrosampler). All DIC samples taken until day 17 were measured using an Automated Infra-Red Inorganic Carbon Analyzer (AIRICA) with a LICOR detector (LI-7000 CO2/H20 Analyzer, MARIANDA, Kiel). Certified reference material (Dickson standard for oceanic CO2 Measurements - CRM Batch 142 with salinity = 33.389 and DIC = 2038,07 µmol/kg) was measured and used to correct measured sample values. Additional DIC samples were measured using gas chromatography-mass spectrometry (GC-MS) to determine the 13C signal. The data of the GC-MS was adjusted to the AIRICA data using a linear transformation. Missing days were filled using an average of the day before and after.