Increasing N Abundance in the Northwestern Pacific Ocean Due to Atmospheric Nitrogen Deposition

See allHide authors and affiliations

Science  28 Oct 2011:
Vol. 334, Issue 6055, pp. 505-509
DOI: 10.1126/science.1206583


The relative abundance of nitrate (N) over phosphorus (P) has increased over the period since 1980 in the marginal seas bordering the northwestern Pacific Ocean, located downstream of the populated and industrialized Asian continent. The increase in N availability within the study area was mainly driven by increasing N concentrations and was most likely due to deposition of pollutant nitrogen from atmospheric sources. Atmospheric nitrogen deposition had a high temporal correlation with N availability in the study area (r = 0.74 to 0.88), except in selected areas wherein riverine nitrogen load may be of equal importance. The increase in N availability caused by atmospheric deposition and riverine input has switched extensive parts of the study area from being N-limited to P-limited.

Rapid growth in human population and industrial activity has led to increases in the concentrations of pollutant nitrogen species (NOx and NHy) throughout the environment (1). However, the changes that result from anthropogenic nitrogen (air-NANTH) deposition into lakes and oceans have yet to be fully explored (24). Several recent studies have reported that the impacts of air-NANTH deposition on lakes and oceans can be locally substantial (58). For example, Elser et al. (8) showed that increases in air-NANTH deposition have increased the ratio of nitrogen to phosphorus (P) in lakes in Norway, Sweden, and the United States, thereby shifting nutrient limitation from nitrogen to P. These shifts can alter the composition of phytoplankton species and, in the long run, the structure of the ecosystem. Modeling studies suggest that air-NANTH deposition can also change the chemistry of coastal and marginal seas located downstream of densely populated East Asian, European, and eastern North American regions (6, 7). In these areas, air-NANTH deposition from urban, agricultural, and industrial expansion is high (1, 6). However, we are unaware of any observational evidence linking changes in air-NANTH deposition to changes in marine nutrient biogeochemistry.

We investigated the relative abundance of N over P in waters of the East Asian marginal seas, including the East China Sea, the coastal waters of Korea, the East Sea (Sea of Japan), and the Pacific coast of Japan. To do so, we analyzed observational data (~173,000 data points) collected from the study area [inset in Fig. 1A and fig. S1 in the supporting online material (SOM) (9)]. The temporal trends in the relative abundance of N over P (N*) were evaluated by calculating N* at each sampling location, where N* = N – (RN:P) × P (10). Here, N and P are the measured concentrations, and RN:P is the mean N:P ratio in the deep (>1000 m depth) East Sea, which has been reported to be 13.06 ± 0.03 (1113).

Fig. 1

(A) Rate of change (μM decade−1) of N* in surface waters (≤50 m) of the study area. The red and yellow boxes indicate regions in which the N* values tended to increase, and the blue box indicates a region in which N* decreased. Boxes with statistically significant N* trends are marked with an asterisk (indicating statistical significance based on both the t test and bootstrapping methods, Group 1) or with the characters “t” or “b” (indicating statistical significance based on either the t test or bootstrapping) next to the box number. CR and HR indicate the locations of the mouths of Changjiang and Han rivers, respectively. (Inset) Locations at which the data were sampled over the past 30 years, from the archive of the Korea National Fisheries Research and Development Institute (red) or the Japan Meteorological Agency (blue). (B) Time-dependent trends in normalized N* anomaly (N* values minus mean values for the observational periods) for all boxes showing statistically significant trends. Numbers on the right match the box numbers in (A); numbers in parentheses after the box numbers indicate confidence intervals for the rates of change of N*.

We divided the study area into 46 boxed regions, each 2° latitude by 2.5° longitude, to provide an optimal grid size such that the distribution of data points permitted the trends in N, P, and N* to be discerned. Within each box, the bimonthly mean values of N, P, and N* (e.g., Jan−Feb, Mar−Apr, …, Nov−Dec) were calculated over the study period. The t test and bootstrapping (P = 0.05) were used to evaluate the significance of the long-term trends in N, P, and N*. The geographic variations of the resulting N* values were also examined (Fig. 1). Two distinctive groups in N* dominated the surface waters (≤50 m) of the study area.

Group 1. ΔN*/Δt > 0

Case 1. ΔN/Δt > 0 and ΔP/Δt ≈ 0

Case 2. ΔN/Δt > 0 and ΔP/Δt < 0

Case 3. ΔN/Δt > 0 and ΔP/Δt > 0; where (ΔN/Δt) > (RN:P) × (ΔP/Δt)

Group 2. ΔN*/Δt ≈ 0 (not significantly different from zero)

In the first group, N increased over time t (ΔN/Δt > 0), regardless of the changes in P (red and yellow in Fig. 1A). Of the total 46 boxed regions, 24 boxes fell into Group 1, accounting for ~52% of the study area. Most boxes within Group 1 were found in the East China Sea, Korean coastal waters, and the East Sea. Group 2 also contained many boxed regions in which N and N* increased over time, but the rates of increase were not statistically significant. Twenty-one boxed regions fell into Group 2 and were mainly located in the Pacific coast of Japan (light yellow in Fig. 1A). Group 2 accounted for approximately 46% of the study area. Only one box showed a decreasing N* trend over time, due primarily to an increase in P, while N remained relatively unchanged.

Over the 30-year study period, the rate at which N* increased varied considerably in space and time. The rate of N* increase was highest in regions of the East China Sea and in the coastal waters of Korea (>1.0 μM decade−1). The trends progressively decreased moving eastward into the East Sea (~0.5 μM decade−1). The rate of N* increase off the Pacific coast of Japan remained positive but was not statistically different from zero. The areas showing an increase in N* were mostly located downwind (to the east and southeast) of the populated regions of the East Asian continent (fig. S2), suggesting air-NANTH deposition as a potential underlying cause of the trend. Air-NANTH deposition over the study area was reported to have increased by a factor of 10 over the past 140 years, a larger increase than the global mean increase of a factor of 3.5 (7, 14). In more recent periods, NOx emissions from East Asia increased 65% over the period 1975 to 1987 and 230% over the subsequent 15 years (15, 16). Therefore, the importance of air-NANTH deposition in the ocean regions examined in our study has been previously recognized as an external nitrogen source (1720).

To evaluate whether air-NANTH deposition was an important driver of ocean N* variability in our study area, we compared the values of air-NANTH deposition measured at four air-monitoring sites with the values of seawater N* (21) obtained from boxed regions in the vicinity of the air-monitoring sites (see SOM for a description of the methods by which air-NANTH deposition was determined). The seawater N concentration measured southwest of Cheju Island has increased in the time since measurements commenced in 1986, whereas the P concentration has remained approximately unchanged (Fig. 2A). The same trends in N and P levels were found off the east coast of Korea between 1983 and 2000 (Fig. 2B). The seawater N concentration southwest of Cheju Island increased considerably from 2001 to 2002; this was the main driver of the increase in N* over the same period. After a brief interval of high N* between 2002 and 2005, N* values rapidly decreased, and then increased. A similar N* pattern was evident off the east coast of Korea, except that N* did not increase during the latter part of the 2000s (2007 to 2010) (Fig. 2B).

Fig. 2

Time series of N* (μM) (red line), N (nitrate, μM) (blue line), and P (phosphorus, μM) (black line) (A) in the region southwest of Cheju Island (box 5), (B) off the east coast of Korea (box 15), (C) off the west coast of Korea (box 6), and (D) in the region east of Oki Island (box 26). The color gradations indicate the confidence interval (P = 0.05) from the mean of the measurements collected over a 2-month interval. The dotted lines in (A) to (D) correspond to N* = 0 μM.

More important, the observed patterns in seawater N* at the two sites were highly correlated with those of total air-NANTH deposition at the closest air-monitoring stations [located on Cheju Island (r = 0.46) (Fig. 3A) and Uljin (r = 0.83) (Fig. 3B)] during the period 2002–2008, for which both seawater N* and air-NANTH deposition data were available. The temporal variations in N* in these two regions were more dramatic than those observed off the west coast of Korea and in the region east of Oki Island. The values of N and N* off the west coast of Korea and in the region east of Oki Island increased steadily over the observational period, whereas the P concentration remained approximately unchanged (Fig. 2, C and D). Compared with the correlations observed southwest of Cheju Island and off the east coast of Korea, the variations in seawater N* values noted off the west coast of Korea and in the region east of Oki Island were more highly correlated with those of total air-NANTH deposition at the air-monitoring stations of Imsil (r = 0.74) (Fig. 3C) and Oki Island (r = 0.88) (Fig. 3D). Finally, the significance of the correlations between seawater N* and air-NANTH deposition levels in all four locations indicated that increasing air-NANTH deposition over the past 30 years most likely caused the observed increase in N concentration and, therefore, N*, over the East Asian marginal seas.

Fig. 3

(A to D) Time-series (left column) and correlation (right column) plots of N* (μM) and the total (wet + dry) air-NANTH deposition (NO3 + NH4+, mmol m–2 month–1) measured at the four air-monitoring stations, and the values of the seawater N* measured in boxed regions in the vicinity of the air-monitoring stations (filled red squares) indicated in (E). For the time-series and correlation plots, we used 2-year moving averages of two measures, air-NANTH deposition and seawater N*. The dashed lines in the correlation plots are least-squares linear fits. The filled green and blue circles in (E) indicate the locations of Anmyeon Island and two Japanese air-monitoring sites (Sado and Rishiri), respectively (see SOM). (F) The major pathways of the Changjiang and Han rivers in winter (solid blue line) and summer (solid red line). The climatological salinity map in the East China Sea showing the movement of the Changjiang River plume in summer (37) is included. The dotted contour line corresponds to the 3-μM N level observed in summer 1998, when the Changjiang River discharge was the greatest observed since the 1950s (31).

Across the study area, the rate of increase in the N* inventory of the surface layer (ΔN*/Δt × 50 m) ranged from 2 to 10 mmol m−2 year−1, which accounted for only 5 to 10% of air-NANTH deposition (50 to 110 mmol N m−2 yr−1). This large disparity indicates that most deposited air-NANTH was either assimilated and exported as organic matter or was mixed downward to a depth in excess of 50 m. By comparing the amount of air-NANTH deposition with estimates of export production based on an empirical model (22) and a coupled physical-ecosystem model (19), air-NANTH deposition was found to be a important contributor to export production in this region, accounting for 4% of export production in the East China Sea and 10% in the East Sea (19, 20).

The impact of air-NANTH deposition on seawater N* was not confined to the surface layer but extended to deeper layers (Fig. 4). For the Pacific coast of Japan, air-NANTH deposition only marginally increased N* values in the upper 1000 m, primarily because of the lower rate of air-NANTH deposition. Oscillation patterns in N* observed in the upper ocean were previously reported (23), but the cause was unknown. In contrast, for the East Sea, air-NANTH deposition increased the N* values of waters to depths of 750 m, but the signals were rapidly attenuated with increasing depth. We computed the N* inventory in the East Sea by vertical integration of N* between the surface and 750 m depth and found that its increase over the past three decades was 0.68 × 1012 mol N, in excellent agreement with the temporal integral over this period of the air-NANTH deposition rate, 0.79 × 1012 mol N (table S2). This indicates that the observed N* increase in the East Sea can be attributed solely to air-NANTH deposition; this conclusion is further strengthened by the fact that no major river (potentially a major N* contributor) flows into the East Sea.

Fig. 4

Time series of the 3-year mean N* (μM) for various depth ranges in (A) the East China and Yellow seas, (B) the East Sea (Sea of Japan), and (C) the Pacific coast of Japan. The colors indicate the N* values derived from the data collected at the indicated depth ranges. The dotted lines correspond to N* = 0 μM. The error bars are the confidence intervals of the resulting N*, for P = 0.05. The colored lines in (C) not indicated in the legend correspond to the depths indicated in (A) and (B).

For the East China and Yellow seas, rivers (e.g., the Changjiang River, which is anomalously high in terms of N* level with N* > 100 μM), may also have contributed to the increasing trend in seawater N* (fig. S5). We determined the correlations between N* values in the Changjiang River and those of seawater N* observed in boxed regions 1, 2, 4, and 5, which are located downstream of the river plume, where riverine influence is probably greatest (24). The correlations were estimated to be r = 0.43 to 0.62 in all selected ocean boxes but were statistically significant (P < 0.05) only in boxes 2 and 5. In particular, in box 5, for which air-NANTH deposition data were available, seawater N* was approximately equally correlated with the total air-NANTH deposition (r = 0.46) and with N* values in the Changjiang River (r = 0.58). Such similar correlations indicate that air-NANTH deposition and riverine nitrogen flux contribute equally to seawater N* variability, particularly in the selected areas under the direct influence of the river plume. During the past three decades, the combined total nitrogen flux from the Changjiang River (~2.6 × 1012 mol N) and the air-NANTH deposition (~1.7 × 1012 mol N) into the East China and Yellow Seas have greatly exceeded the increase (~0.07 × 1012 mol N) in the water column N* inventory during the same period (table S2), indicating that most of the added nitrogen is either advected away, buried in sediments, or denitrified. Another external source of nitrogen that may have contributed to an increase in surface N concentration, and therefore N*, is N2 fixation. However, this contribution is likely negligible (25).

The present study shows that the abundance of N relative to P in northeastern Asian marginal seas has increased since 1980. Air-NANTH deposition has narrowed the deficiency of N relative to P across the study area and has even resulted in an N surplus in the East China Sea, Yellow Sea, and East Sea, commencing in the mid-1990s. In areas located downstream of the Changjiang River plume, contributions from both air-NANTH deposition and riverine nitrogen fluxes appeared to be of equal importance in governing trends in seawater N*. Our findings may have broader implications. Indeed, the observed trends may be extrapolated to the coastal seas of the North American Atlantic Ocean and the North, Baltic, and Mediterranean Seas, which have received ever-increasing amounts of air-NANTH deposition and river-borne nitrogen, comparable to those absorbed by coastal and marginal seas of the northwestern Pacific Ocean (6, 7, 14, 26).

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 to S6

Tables S1 and S2


References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. The low RN:P in the deep waters of the East Sea has not yet been explained. It is possible that the East Sea favors phytoplankton assimilation of N and P at a ratio lower than 16 (2729).
  3. Seawater N* values in the euphotic layer were affected by the supply of N from the upper thermocline (formed by remineralization of organic matter), which occurs on time scales less than 1 year; this is in contrast to the monthly time scale on which the atmospheric nitrogen supply into the euphotic layer was measured. The mismatch in time scales of these two processes (N supply from below versus air-NANTH deposition) could affect the correlations between the values of surface N* and air-NANTH deposition. Therefore, to remove seasonal fluctuations and to highlight the interannual or long-term trends, the data for air-NANTH deposition and seawater N* were smoothed using a 2-year moving average before undertaking comparisons. This data treatment minimized potential biases in the correlations in Fig. 3, enabling comparison of the two parameters.
  4. The riverine influence is largely confined to Chinese coastal waters during winter (October to April), whereas, in summer (May to September), the freshwater plume extends northeast toward Cheju Island (on which an air-monitoring station, marked “A” in Fig. 3E, is located) (30). However, the impact of the river plume rapidly diminished away from the source, as indicated by a rapid decrease in N concentration from 40 μM at the mouth of the Changjiang River to 3 μM in offshore waters, ~300 km distant from the river mouth (Fig. 3F) (31). About 75% of the total nitrogen load from the Changjiang River is added to the East China Sea during summer, with the remainder added during winter (32). In addition, we found no statistically significant correlations between seawater N* values in boxes 6 and 7 (wherein the effect of the Han River is likely strong) and N* values measured over the past 15 years in waters near the mouth of the Han River (fig. S6), indicating a negligible contribution from the Han River nitrogen flux into the Yellow Sea.
  5. Contributions from N2 fixation are unlikely because oceanic conditions in our study area did not favor N2 fixation. With rare exceptions (33, 34), most N2 fixation blooms occur in tropical oceans in which the fixed N concentration is nearly zero and the sea surface temperature is generally higher than 25°C (35, 36). Most of our study areas featured sea surface temperatures below 25°C, except during the summer season, and the surface N concentrations were generally greater than the detection limits.
  6. Acknowledgments: This work would not have been possible without the efforts of many scientists who contributed data to National Fisheries Research and Development Institute and Japan Meteorological Agency for the seawater nutrient data and to Acid Deposition Monitoring Network in East Asia and Korea Meteorological Administration (Climate Change Information Center) for the atmospheric deposition data. The preparation of the manuscript was supported by the Mid-career Researcher Program (2009-0084756) of the Korea National Research Foundation (K.L.). Partial support was provided by the Korea Meteorological Administration Research and Development Program under grant RACS_2010-1006 (K.L.). Support for H.J.J. was provided by the Ministry of Land, Transport, and Maritime Affairs’ Ecological Disturbance Program. Support for R.G.N. was provided by NASA’s Ocean Biology and Biogeochemistry Program (contract NNX08AO25G).
View Abstract

Navigate This Article