Heme b is an iron-containing cofactor in hemoproteins that participates in the fundamental processes of photosynthesis and respiration in phytoplankton. Heme b concentrations typically decline in waters with low iron concentrations but due to lack of field data, the distribution of heme b in particulate material in the ocean is poorly constrained. Here we report particulate heme b distributions across the Atlantic Ocean (59.9°N to 34.6°S). Heme b concentrations in surface waters ranged from 0.10 to 33.7 pmol L −1 (median = 1.47 pmol L −1 , n = 974) and were highest in regions with a high biomass. The ratio of heme b to particulate organic carbon (POC) exhibited a mean value of 0.44 μmol heme b mol −1 POC. We identified the ratio of 0.10 µmol heme b mol −1 POC as the cutoff between heme b replete and heme b deficient (anemic) phytoplankton. By this definition, we observed anemic phytoplankton populations in the Subtropical South Atlantic and Irminger Basin. Comparison of observed and modelled heme b suggested that heme b could account for between 0.17-9.1% of biogenic iron. Our large scale observations of heme b relative to organic matter provide further evidence of the impact of changes in iron supply on phytoplankton iron status across the Atlantic Ocean.

Heme b is an iron-containing co-factor in hemoproteins. Heme b concentrations are low (<1 pmol L -1 ) in iron limited phytoplankton in cultures and in the field. Here, we determined heme b in marine particulate material (>0.7 μm) from the North Atlantic Ocean (GEOVIDE cruise – GEOTRACES section GA01), which spanned several biogeochemical regimes. We examined the relationship between heme b abundance and the microbial community composition, and its utility for mapping iron limited phytoplankton. Heme b concentrations ranged from 0.16 to 5.1 pmol L -1 (median = 2.0 pmol L -1 , n = 62) in the surface mixed layer (SML) along the cruise track, driven mainly by variability in biomass. However, in the Irminger Basin, the lowest heme b levels (SML: median = 0.53 pmol L -1 , n = 12) were observed, whilst the biomass was highest (particulate organic carbon, median = 14.2 μmol L -1 , n = 25; chlorophyll a : median = 2.0 nmol L -1 , n = 23) pointing to regulatory mechanisms of the heme b pool for growth conservation. Dissolved iron (DFe) was not depleted (SML: median = 0.38 nmol L -1 , n = 11) in the Irminger Basin, but large diatoms ( Rhizosolenia sp.) dominated. Hence, heme b depletion and regulation is likely to occur during bloom progression when phytoplankton class-dependent absolute iron requirements exceed the available ambient concentration of DFe. Furthermore, high heme b concentrations found in the Iceland Basin and Labrador Sea (median = 3.4 pmol L -1 , n = 20), despite having similar DFe concentrations to the Irminger Basin, were attributed to an earlier growth phase of the extant phytoplankton populations. Thus, heme b provides a snapshot of the cellular activity in situ and could both be used as indicator of iron limitation and contribute to understanding phytoplankton adaptation mechanisms to changing iron supplies.

We investigated trace element stoichiometries of the nitrogen-fixing marine cyanobacterium Crocosphaera subtropica ATCC51142 under steady-state growth conditions. We utilized exponentially fed batch cultures and varied iron (Fe) concentrations to establish nutrient limitation in C. subtropica growing at a constant growth rate (0.11 d-1). No statistical difference in cell density, chlorophyll a, particulate organic carbon (C), nitrogen (N) and phosphorus (P) were observed between consecutive days after Day 14, and cultures were assumed to be at steady state with respect to growth for the remaining 11 d of the experiment. Cultures were limited by P in the highest Fe treatment (41 nmol l-1) and by Fe in the 2 lower-concentration Fe treatments (1 and 5 nmol l-1). Cell size and in vivo fluorescence changed throughout the experiment in the 1 nmol l-1 Fe treatment, suggesting ongoing acclimation of C. subtropica to our lowest Fe supply. Nevertheless, Fe:C ratios were not significantly different between the Fe treatments, and we calculated an average (±SD) Fe:C ratio of 32 ± 14 µmol mol-1 for growth at 0.11 d-1. Steady-state P-limited cells had lower P quotas, whilst Fe-limited cells had higher manganese (Mn) and cobalt (Co) quotas. We attribute the increase in Mn and Co quotas at low Fe to a competitive effect resulting from changes in the supply ratio of trace elements. Such an effect has implications for variability in elemental stoichiometry in marine phytoplankton, and potential consequences for trace metal uptake and cycling in marine systems.