|
| Organic carbon (OC) synthesised by plankton is exported out of the surface layer as particulate (POC) and
dissolved (DOC) organic carbon. This ‘‘biologicalpump’’ constitutes amajor pathway in the globalmarine carbon
cycle, each year exporting about 10 Pg C into the ocean interior, where it is subsequently remineralised
via biological decomposition. Remineralised inorganic nutrients and carbon are, ultimately, again brought to
the surface by advection and turbulentmixingwhich closes the OC-cycle in the upper ocean. Thus, remineralisationrates
ofOCare a critical component of the biologicalpump. These rates are regulated bythe lability of
thematerial and the environmental conditions in the ambient water. Temperature is particularly important
in regulating the rate of microbial respiration and, thus, remineralisation rates. A significant temperature
dependence of themicrobial metabolic activity in the ocean interior is expected, as this is a feature observed
elsewhere in the biosphere. Such temperature dependence of microbial remineralisation of POC andDOCwill
alter theamount ofmaterial available for transport by the biologicalpump tothe deep ocean.Very fewstudies
on the lability of OC and temperature sensitivity ofmicrobial degradation processes in the mesopelagic zone
(100–1000 m)have, to date, been carried out. Here,we present a comprehensive newexperimental data set
from all major ocean basins and quantify remineralisation rates of OC and their temperature sensitivity in
long-term incubations of water fromthe upper 350 m. Microbial respiration was measured by non-invasive
oxygen optodes and oxygen consumptionwas used as a constraint for determining the remineralisation rates
and temperature sensitivity by two complementary methods. First, we analysed the oxygen consumption
from a multi-component OC-model where the concentration, remineralisation rates and temperature
sensitivity of two bio-available (labile) pools of organic carbon were fitted to the data via a non-linear fitting
procedure. Thereafter, a continuousOC-modelwasfitted to the data through an inversemethod and information
about lability, temperature sensitivity and structural composition of the OC-pool was analysed together
with the results fromthe two-pool solutions.Medianvalues of remineralisation rates fromall experiments on
material characterising sinking POC were found to be 0.6 and 0.05 days1 for the two decomposable pools
corresponding to turnover times of 2 and 21 days, respectively. Accordingly, solutions from the continuous
model resulted in median turnover times between 6 and 11 days. Similar analyses were carried out for the
OC-pool characterising DOC. A significant bio-available OC-pool was found to be present in the surface layer
with amedian value fromall experiments of 30 lMTOC. Themedian values of all remineralisation rates from
the two bio-available OC-pools were found to be 0.2 and 0.02 days1, corresponding to turnover times of 5
and 56 days, respectively, in good agreementwith previous studies. The corresponding temperature sensitivities,
characterised by a Q10-value, were found to be about 1.9 for the POC-fractionwhereas the DOC fraction
was characterised with values above 2.8 for the continuous OC-models. The analysis of the structural
composition indicated that the TOC distribution in the surface layer was characterised by more heterogeneous
material in terms of lability compared with the POC material sampled from the upper 350 m. Finally,
we analyse the potential impact of the calculated temperature sensitivity on the general OC-cycling in the
upper ocean and show that the implied reduction in OC-flux in a warmer oceanmay have significant impact
on nutrient cycling in general and also tends to reduce future ocean carbon uptake significantly. | |
|