Katja Fennel

Model parameterizations of particulate organic carbon (POC) flux are critical for simulating the strength and future evolution of the biological carbon pump (BCP) but remain poorly constrained because direct observations are sparse. Here, we ask whether the Biogeochemical (BGC)‐Argo proxy observations of POC can help distinguish between these parameterizations by objectively comparing two common parameterizations, which reproduce the observed slowdown of flux attenuation with depth by either decreasing the remineralization rate or increasing the sinking velocity. Both can well reproduce the BGC‐Argo observations in top 1,000 m but predict different POC concentration below, making them possible to be distinguished if BGC‐Argo observations were available there. Therefore, an integration of backscatter sensors into the Deep Argo program is recommended to provide full depth proxy measurements. If the parameterization is known, POC flux can be determined from POC concentration. Thus, the BGC‐Argo proxy observations of POC concentration provide new insights into the BCP.

Dinitrogen (N 2 ) fixation was investigated at a pelagic station in the oligotrophic waters of the northern Gulf of Aqaba, Red Sea, between February 2016 and December 2018. In situ 15 N 2 and 13 C incubations were used to evaluate photic and aphotic N 2 fixation rates and diazotrophic contribution to water column productivity. N 2 fixation rates were typically low (below detection to <0.5 nmol N L −1 d −1 ). Maximal rates of 3.1 nmol L −1 d −1 , measured at 100 m when conspicuous slicks of the cyanobacterium Trichodesmium appeared on the surface water. Amplicon sequencing of nifH demonstrated that non‐cyanobacterial diazotrophs, mostly α‐ and γ‐proteobacteria comprised the majority (82–100%) of amplicon sequence variants retrieved from photic and aphotic depths when low N 2 fixation rates were measured, while amplicons representing cyanobacteria were nearly absent, but appeared in low abundance (~ 2%) when maximal rates were measured. During the stratified summer‐period, water‐column N 2 fixation rates (10–75 μ mol m −2 d −1 ) comprised ~ 1–40% of new production (NP). During the winter mixing period N 2 fixation rates were considerably higher (11–242 μ mol m −2 d −1 ) but made up only <1% of NP. This is because, during this period, nitrate supplied to the photic zone by the vertical mixing becomes the major N source for NP. We conclude that on an annual average, diazotrophy plays a minor role in the NP of the Gulf. The major “new” nitrogen sources are cross‐thermocline turbulent diffusion of nitrate during summer stratification and vertical mixing during the fall–winter.