Changes in wet nitrogen deposition in the United States between 1985 and 2012

Anthropogenic nitrogen (N) emissions have caused an increase in global N deposition (Galloway et al 2004), affecting a number of key environmental issues such as loss of biodiversity (Bobbink et al 2010, Clark et al 2013), eutrophication and acidification of terrestrial and aquatic ecosystems (Greaver et al 2012) and nitrogen–carbon–climate interactions (De Vries et al 2008, Erisman et al 2011, Suddick and Davidson 2012). The United States (US) is among the global hotspots of N emissions and deposition (Dentener et al 2006, Clarisse et al 2009, Ghude et al 2009). Emissions of oxidized N (NO_x ) have steadily increased throughout the middle of the 20th century until it has been first regulated under the Clean Air Act (CAA) of 1970 through National Ambient Air Quality Standards (NAAQS) (Houlton et al 2013). Between 1990 and 2008, NO_x emissions in the US have decreased by 36% due to the CAA Amendments of 1990 (Davidson et al 2012). By contrast, emissions of reduced N (NH₃) are currently unregulated under the NAAQS in the US. Reduction of NO_x emissions have led to a decrease in precipitation nitrate (NO₃⁻) concentration and nitrate deposition, especially in the eastern US (Butler et al 2003, 2005), while unregulated NH₃ emissions may increase ammonium (NH₄⁺) deposition. Accurate information on trends in the deposition of both N species is necessary to assess the effectiveness of the existing NO_x regulation policies and the need of NH₃ regulation in view of the resulting environmental effects.

The National Atmospheric Deposition Program (NADP) in the US was initiated to monitor precipitation chemistry in the late 1970s and the network grew rapidly in the early 1980s. There have been a number of studies analyzing the spatial patterns of N deposition (Fenn et al 2003, Holland et al 2005) and temporal trends of precipitation N chemistry using the NADP data (e.g. Lynch et al 1995, Nilles and Conley 2001, Lehmann et al 2005, Stephen and Aneja 2008). Most of these analyses either focus on the spatial pattern of N deposition without considering the temporal trends or only test the trends of precipitation N chemistry but not wet N deposition. There has also been some modeling work on the distributions and historical changes of N deposition in the US (Zhang et al 2012, Lamarque et al 2013). However, the multi-model results of the Atmospheric Chemistry and Climate Model Intercomparison Project showed limited agreement with and even misrepresentation of observational N deposition changes between 1980 and 2000 over the US (Lamarque et al 2013).

With long-term data available, assessing temporal trends of wet deposition is ready to provide an observed evidence of changes in atmospheric N inputs to various ecosystems. The drivers of reduced and oxidized N deposition are different from each other. For instance, ammonium deposition is mainly driven by NH₃ emissions from agricultural activities such as livestock breeding and ammonia fertilizer application (Anderson et al 2003, Sutton et al 2008), whereas nitrate deposition is mostly derived by NO_x emissions from combustion processes (Galloway et al 2003). Moreover, the oxidized and reduced N species can exert different environmental effects. For instance, ammonium deposition is more effective than nitrate deposition in decreasing biodiversity and is more harmful to vegetation (Erisman et al 2007). Therefore, it is essential to consider N species separately when assessing the status and impacts of N deposition. It is also important to note that once deposited, both ammonium and nitrate can be converted to other N species, primarily by microbial action.

Environmental data such as N deposition commonly do not always follow a normal distribution. The Mann–Kendall test (MKT) is a non-parametric test to detect time-serial trends without requiring a normal data distribution (Mann 1945). Recently, a regional Kendall test (RKT) has been proposed to quantitatively combine results of the MKT for individual sites to detect whether a consistent trend occurs over an entire region (Helsel and Frans 2006). This study applied the MKT and RKT to observed time-serial data from NADP to assess the trends of wet deposition of ammonium, nitrate and total dissolved inorganic N (DIN = NH₄⁺ – N + NO₃⁻ – N) over the US between 1985 and 2012. Note that organic N deposition was not included in this study due to a lack of observed data, even though organic N was suggested to account for approximately 25% of total N deposition on a global basis (Jickells et al 2013).

2.1. Data set

Observed data of annual wet deposition between 1985 and 2012 were provided from NADP (http://nadp.sws.uiuc.edu/), including information on wet deposition of ammonium, nitrate and DIN and site locations (latitude and longitude). Samples were gathered using buckets opening at the start and closing at the end of each precipitation event and were pooled to one weekly sample for chemical analyses by a central laboratory with a good quality control (more information available at: http://nadp.sws.uiuc.edu/NTN/). Wet N deposition (kg N ha⁻¹ yr⁻¹) was calculated as an annual sum of N fluxes in precipitation. Sites were excluded if data were missing for one or more years. Totally 151 sites were used, including 20 sites in the Northeast, 42 sites in the South, 38 sites in the Midwest and 51 sites in the West region (figure 1). Sites in Alaska and Hawaii were not considered in the analysis of this study.

Zoom In Zoom Out Reset image size

2.2. Statistical analysis

3.1. Current patterns of wet N deposition

3.2. Patterns in temporal trends of wet N deposition

The results of the MKT showed that patterns of temporal trends in wet N deposition differed significantly by species and by region between 1985 and 2012. Wet ammonium deposition showed an increase mainly in the West, Midwest and the eastern part of the South region, while in the Northeast region and western part of the South region most sites did not show a significant trend (figure 2(a)). Wet nitrate deposition showed a decrease mainly in the Northeast, South and the Eastern part of the Midwest region, while most sites in the West and the Western parts of the Midwest regions did not show a significant change (figure 2(b)). Taken together, sites showing a significant increase in wet DIN deposition were mostly located in the northern part of the West region and western part of the Midwest region, while sites with a significant decrease were mostly located in the Northeast region (figure 2(c)). The NH₄⁺–N : NO₃⁻–N ratio showed a significant increase across the US with the highest rate of change occurring in the Midwest, West and Eastern part of the South region (figure 2(d)).

Zoom In Zoom Out Reset image size

3.3. Overall trends of wet N deposition in different regions

Our results showed that current wet deposition of ammonium and nitrate (2011–2012) in the US had distinct regional patterns and magnitudes, which generally agreed with the results simulated by the GEOS-Chem global chemical transport model based on the period 2006–2008 (Zhang et al 2012). Ammonium dominated wet DIN deposition in the Midwest, South and West regions, while in the Northeast region nitrate and ammonium both accounted for half of wet DIN deposition. This is because emission sources of NO_x and NH₃ are distributed differently in these regions. Highest ammonium deposition occurred in the Midwest region, being the major food production region with large NH₃ emissions from livestock breeding and chemical N fertilizer application. The regions showing highest nitrate deposition were characterized by developed manufacturing industry. Considering the contribution of dry deposition (1.3 ∼ 1.6 times of wet DIN deposition; Zhang et al 2012) and organic N deposition (∼25% of total N deposition; Jickells et al 2013), total N deposition was estimated to be 2.9–3.3 times of wet DIN deposition. In the Midwest, South and Northeast regions, total N deposition was thus estimated to be 9 ∼ 26 kg N ha⁻¹ yr⁻¹. This is comparable to the range of empirical critical loads for forest ecosystems (3 ∼ 26 kg N ha⁻¹ yr⁻¹) (Pardo et al 2011) and exceeded that of lakes (2 ∼ 8 kg N ha⁻¹ yr⁻¹) (Baron et al 2011, Pardo et al 2011). This result agrees with a recent review, which summarized the ecological effects of N and S deposition in the US, concluding that current air-quality standards do not protect ecosystems from the effects of N deposition in many parts of the nation (Greaver et al 2012). The absence of regulation policy for NH₃ emissions should have contributed to the failure of current NO_x regulation policy to protect many ecosystems in the US.

Wet ammonium deposition increased by 50% at national scale between 1985 and 2012, with the highest increase in the Midwest and Eastern part of the South region. Currently 80% of total NH₃ emissions in the US are from livestock manure management and application of chemical N fertilizers (Reis et al 2009, Davidson et al 2012). Because NH₃ emissions from agricultural activities are rarely regulated, these growing NH₃ emissions significantly drove the increase of ammonium deposition during the recent decades (table 2; figure 3(a)). The increase in ammonium deposition in the Midwest region was likely attributed to larger animal breeding (e.g. hogs and cattle) and N fertilizer application, whereas the increase in the Southeast was mostly due to smaller livestock (e.g. chickens) (Stephen and Aneja 2008). The Midwest, South and West regions showed very high rates of increase in wet ammonium deposition, suggesting the urgency to regulate NH₃ emissions in these regions.

Zoom In Zoom Out Reset image size

Wet nitrate deposition decreased substantially by one-third at national scale between 1985 and 2012, with a largest reduction in the Northeast region. The decrease of wet nitrate deposition was attributed to the curb of NO_x emissions by the promulgation of CAA and especially the 1990 amendments. Due to the trade-off of a decrease in NO_x from fossil-fuel burning industries and an increase in emissions from automobiles, national NO_x emissions did not show substantial decrease until the late 1990s (Howarth et al 2002). Accordingly, satellite observations have clearly detected decreases in NO₂ columns since 1999 due to control of NO_x emissions from power plants over the eastern US (Kim et al 2006). Further, NO_x emissions from electrical generation are expected to have decreased by over 50% between 2005 and 2014 based on the Cross-State Air Pollution Rule (Davidson et al 2012). The temporal variations of wet nitrate deposition agree well with the changes of NO_x emissions, showing a rapid decrease since late 1990s (figure 3(b)). Despite significant reduction in wet nitrate deposition in the Northeast region and nearby states, both the wet and dry deposition of oxidized N in the Northeast region is still the highest across the nation (this study; Zhang et al 2012).

Trends of wet DIN deposition varied by regions between 1985 and 2012 and no significant changes in wet DIN deposition has been found at national scale (table 2; figure 3(c)). However, the increase in wet ammonium deposition and the decrease in wet nitrate deposition resulted in a consistent increase of NH₄⁺–N : NO₃⁻–N ratio across the nation (table 2; figure 3(d)). Nitrate dominated wet deposition across the nation in 1985 (mean NH₄⁺–N : NO₃⁻–N ratio = 0.72) and the NH₄⁺–N : NO₃⁻–N ratio had increased to 1.49 in 2012, indicating a transition of dominant N species in wet deposition. The significant decrease of wet nitrate deposition especially in the Eastern US indicates an effective regulation of oxidized N emissions from industrial power generation, which should have useful implications for the developing countries with rapid increase of NO_x emissions, such as China, India and South Africa (Garg et al 2006, Ghude et al 2009, Liu et al 2011). Currently, industrial power generation in the US only contributes to 27% of national NO_x emissions, while vehicle emissions accounts for as high as 62% (Reis et al 2009). Therefore, further reduction in atmospheric deposition of oxidized N will require a regulation of NO_x emissions from motor vehicles. Wet ammonium deposition currently accounts for 60% of wet DIN deposition at national scale, suggesting a current need in regulating NH₃ emissions. Or else, intensive agricultural activities, necessary to support food security for expected population growth, will continually increase NH₃ emissions and ammonium deposition in the food production regions (Erisman et al 2008).

This study analyzed the trends of wet DIN deposition over the US between1985 and 2012 by applying the MKT and RKT to observed time-serial data from NADP. Wet DIN deposition showed no significant change at national scale, but profound changes have occurred in wet deposition of ammonium and nitrate. Due to effective regulation of NO_x emissions, wet nitrate deposition showed a significant decrease at national scale, with the largest reduction in the Northeast region. Without regulation of NH₃ emissions, wet ammonium deposition showed a substantial increase, with the highest increase in the Midwest region and Eastern part of the South region. Mean NH₄⁺–N : NO₃⁻–N ratio across the nation has increased from 0.72 in 1985 to 1.49 in 2012, indicating a transition of dominant N species in wet deposition. Our results imply that the increase of ammonium deposition due to the absence of control policy for NH₃ emissions should be mainly responsible for the current threat of N deposition to many ecosystems in the US. Moreover, distinct magnitudes, trends and patterns of wet ammonium and nitrate deposition suggest the needs to control N emissions by species and regions.

The authors would like to thank all the sponsors of NADP (http://nadp.sws.uiuc.edu/) and the NADP Program Office for providing the N deposition data used in this study. This study was partly supported by the National Natural Science Foundation of China (No. 31021001), and National Basic Research Program of China on Global Change (2010CB950600).