F. Große et al.: Looking beyond stratification 
2529 
www.biogeosciences.net/13/2511/2016/ 
Biogeosciences, 13, 2511-2535, 2016 
Figure 10. Mass balances for simulated bottom CD at Cefas North Dogger (see Fig. 2, region 2) during stratification (grey shaded) in (a) 2007 
and (b) 2010. Same legend for (a) and (b). Black y axes apply to processes, magenta y axes apply to CD (saturation) concentrations. Changes 
in concentrations due to different processes are cumulative. 
The model-based analysis of different factors affect 
ing CD showed that besides sufficiently long stratification 
(>60 days), surface layer primary production (driving or 
ganic matter export) and sub-thermocline volume are the key 
parameters influencing the bottom CD evolution. Based on 
this, the North Sea can be subdivided into three different 
zones in terms of CD dynamics: (1) a highly productive, non- 
stratified coastal zone (region A), (2) a productive, season 
ally stratified zone with a small sub-thermocline volume (re 
gion B), and (3) a productive, seasonally stratified zone with 
a large sub-thermocline volume (regions C and D). While 
the zones of types 1 and 3 are unlikely to be affected by 
low CD conditions due to either continuously ongoing ven 
tilation (type 1) or the large sub-thermocline volume dilut 
ing the effect of CD consumption (type 3), type 2 is highly 
susceptible to low CD conditions. This results from the spe 
cific combination of high upper layer productivity and small 
sub-thermocline volume, which causes a strong impact of the 
consumption processes on the decrease in the bottom CD con 
centrations. 
The ODI demonstrates that this regional characterisation, 
based on only three controlling parameters, can be applied 
to most parts of the North Sea. The ODI is rather simple 
compared to the eutrophication risk index (EUTRISK; Druon 
et al., 2004) as it is designed to indicate regions with higher 
risk for 0 2 deficiency. Therefore, it may also allow for an 
operational use as the information on stratification can be de 
rived from operational hydrodynamic al models and informa 
tion on net primary production from satellite data. 
The model-based mass balances showed that pelagic bac 
terial remineralisation constitutes the largest CD consuming 
process within the sub-thermocline volume. In the bottom 
layer, benthic remineralisation constitutes the major CD sink. 
which consistently contributes more than 50 % to the over 
all bottom CD consumption. Pelagic remineralisation consis 
tently contributes to more than 20 % of the overall bottom 
CD consumption. Zooplankton respiration and nitrification 
are less important, however, can contribute to up to 14 and 
8 %, respectively. In addition, the results suggest that the rel 
ative contribution of the different CD consuming processes 
in the bottom layer at a certain location depends on the water 
column depth and is independent of the overall consumption. 
The mass balances also showed that differences in the 
surface layer primary production drive variations in the 
sub-thermocline and bottom CD evolution between different 
years. Increased primary production directly enhances the 
export of dead phytoplankton into the deeper layers. Further 
more, it enhances zooplankton growth which causes an addi 
tional increase in organic matter production and export. This 
enhanced zooplankton growth further increases CD consump 
tion due to respiration. The overall increase in organic matter 
export results in stronger bacterial remineralisation which in 
turn triggers nitrification due to the stronger release of am 
monium. 
Our analysis suggests that advection usually only has a mi 
nor effect on the bottom CD dynamics in most North Sea re 
gions which contradicts the interpretation by Queste et al. 
(2013). However, during years of especially low bottom 0 2 
concentrations it may play an important role for the recov 
ery of the 0 2 levels before the breakdown of stratification in 
autumn. In addition, we showed that during the summer pe 
riod only very strong mixing results in a net increase in CD 
as the enhanced nutrient supply triggers primary production, 
eventually increasing the biological CD consumption, which 
balances or even exceeds the enhanced CD supply.