148
et al. 2009). Positive N* values have also been noted in the NW Pacific Ocean
(Gruber and Sarmiento 1997), again an area of higher than average N 2 fixation rates
(Shiozaki et al. 2009, 2015) and dust deposition (Mahowald et al. 2009). Large
negative N* values (net N loss) are associated with the major oceanic ODZs (Gruber
and Sarmiento 1997).
8.2 Basin and Global Scale Inputs
Pelagic Studies
Estimates of inputs of N from N 2 fixation in the current ocean have been developed
for both the major basins and the global ocean. These efforts have used a variety of
approaches including direct extrapolation and scaling of measured rates in the field,
integrated biogeochemical proxies (e.g. N* and P*), isotope mass balances and
other diverse modelling methods (Table 8.1).
Carpenter (1983) derived a relatively small global input by Trichodesmium in the
pelagic realm of 15 Tg N y
−1
. His estimate was based on a cell-specific rate averaged from the accumulating rate measurements of Trichodesmium and applied to
global maps of Trichodesmium density he developed from the much larger set of
observations of its concentration by marine phycologists over the previous century.
An average of available basin-specific rates scaled to all surface waters ≥20 °C
within that basin, with adjustment for seasonal (Table 8.1) had the greatest integrated rates associated with the Pacific Ocean (Galloway et al. 2004). Extrapolated
depth integrated rates of Trichodesmium N 2 fixation based on the C 2 H 2 reduction
method as calibrated with
15
N 2 uptake from 154 stations in the North Atlantic on 6
cruises estimated basin scale inputs of from 22 to 34 Tg N y
−1
(Capone et al. 2005).
Separately, they calculated an overall input for the same system of from 77 to
122 Tg N y
−1
based on the δ
15
N of particulate organic N in surface waters and a
stable isotope mixing model.
A semi-quantitative approximation of global N 2 fixation based on an isotope
mass balance suggested it would need to be well in excess of 100 Tg N y
−1
to balance losses by denitrification and maintain the isotopic signature of deepwater NO 3
−
(Brandes and Devol 2002) (Table 8.1).
Over the last several decades, diverse modeling efforts have probed the spatial
(Fig. 8.3) and temporal trends in oceanic N 2 fixation (among other processes such
as primary productivity and iron cycling) and the regional controls on these processes (Fig. 7.4), with ever increasing resolution and coherence (using the expanding availability of field observations which are used to “sea-truth” models)
(Krishnamurthy et al. 2007, 2010; Moore et al. 2002, 2004).
Several modelling studies have focused on the North Atlantic basin which is rich
in observational data (Benavides and Voss 2015). Using data from a north-south
transect on the Atlantic Meridional Transect (AMT) 17 and from three earlier
8 N 2 Fixation in Ocean Basins
et al. 2009). Positive N* values have also been noted in the NW Pacific Ocean
(Gruber and Sarmiento 1997), again an area of higher than average N 2 fixation rates
(Shiozaki et al. 2009, 2015) and dust deposition (Mahowald et al. 2009). Large
negative N* values (net N loss) are associated with the major oceanic ODZs (Gruber
and Sarmiento 1997).
8.2 Basin and Global Scale Inputs
Pelagic Studies
Estimates of inputs of N from N 2 fixation in the current ocean have been developed
for both the major basins and the global ocean. These efforts have used a variety of
approaches including direct extrapolation and scaling of measured rates in the field,
integrated biogeochemical proxies (e.g. N* and P*), isotope mass balances and
other diverse modelling methods (Table 8.1).
Carpenter (1983) derived a relatively small global input by Trichodesmium in the
pelagic realm of 15 Tg N y
−1
. His estimate was based on a cell-specific rate averaged from the accumulating rate measurements of Trichodesmium and applied to
global maps of Trichodesmium density he developed from the much larger set of
observations of its concentration by marine phycologists over the previous century.
An average of available basin-specific rates scaled to all surface waters ≥20 °C
within that basin, with adjustment for seasonal (Table 8.1) had the greatest integrated rates associated with the Pacific Ocean (Galloway et al. 2004). Extrapolated
depth integrated rates of Trichodesmium N 2 fixation based on the C 2 H 2 reduction
method as calibrated with
15
N 2 uptake from 154 stations in the North Atlantic on 6
cruises estimated basin scale inputs of from 22 to 34 Tg N y
−1
(Capone et al. 2005).
Separately, they calculated an overall input for the same system of from 77 to
122 Tg N y
−1
based on the δ
15
N of particulate organic N in surface waters and a
stable isotope mixing model.
A semi-quantitative approximation of global N 2 fixation based on an isotope
mass balance suggested it would need to be well in excess of 100 Tg N y
−1
to balance losses by denitrification and maintain the isotopic signature of deepwater NO 3
−
(Brandes and Devol 2002) (Table 8.1).
Over the last several decades, diverse modeling efforts have probed the spatial
(Fig. 8.3) and temporal trends in oceanic N 2 fixation (among other processes such
as primary productivity and iron cycling) and the regional controls on these processes (Fig. 7.4), with ever increasing resolution and coherence (using the expanding availability of field observations which are used to “sea-truth” models)
(Krishnamurthy et al. 2007, 2010; Moore et al. 2002, 2004).
Several modelling studies have focused on the North Atlantic basin which is rich
in observational data (Benavides and Voss 2015). Using data from a north-south
transect on the Atlantic Meridional Transect (AMT) 17 and from three earlier
8 N 2 Fixation in Ocean Basins