Atmospheric nitrogen deposition in Yangtze River Delta: insights gained from the nitrogen content and isotopic composition of the moss Haplocladium microphyllum

ABSTRACT

Simple and inexpensive estimation of the rates and sources of atmospheric nitrogen (N) deposition is critical for its effective mitigation in a region with different land-use types. In this study, the N content and N isotopic composition (δ¹⁵N) of moss (Haplocladium microphyllum) tissues and precipitation at six sites with three land-use types (urban, suburban, and rural) were measured in the Yangtze River Delta. A significant linear relationship between moss N content and wet N deposition, and a consistent decrease trend for moss N content and wet N deposition from urban to suburban to rural areas were observed. More negative δ¹⁵N of suburban and rural mosses indicated N mainly released from agriculture and effluent, while the less negative δ¹⁵N of urban mosses were mainly influenced by fossil fuel combustion and traffic emissions. Although the negative moss δ¹⁵N indicates that reduced N dominates wet N deposition, there was no significant correlation between moss δ¹⁵N and the ratio of ammonium to nitrate (NH₄⁺/NO₃⁻). These results reveal that the moss N content and δ¹⁵N can be used as a complementary tool for estimating the rates and sources of wet N deposition in a region with different land-use types.

Graphical abstract

摘要

简单有效地估算大气氮沉降量及其来源至关重要。本研究同时收集长三角市区, 郊区, 农村的雨水和附近的石生细叶小羽藓，分析苔藓氮含量与氮湿沉降量以及苔藓和雨水的氮同位素（δ¹⁵N）。结果显示：苔藓氮含量与氮湿沉降量显著相关，均呈市区>郊区>农村趋势。苔藓δ¹⁵N表明郊区和农村大气氮湿沉降主要来自农业和人畜废水释放，而市区还受煤炭燃烧和机动车排放影响。该研究表明石生苔藓氮含量和δ¹⁵N能用来指示长三角大气氮湿沉降量及其来源，以弥补传统氮沉降监测数据的缺乏和费时费力的缺点。

KEYWORDS:

• Epilithic mosses
• nitrogen content
• nitrogen isotope
• atmospheric wet nitrogen deposition
• sources

关键词:

• 石生苔藓
• 氮含量
• 氮同位素
• 氮沉降
• 来源

1. Introduction

Atmospheric reactive nitrogen (N) deposition has more than doubled over the past century because of the growing human population and its associated demand for food and energy (Galloway et al. 2008). Excessive N deposition could lead to soil acidification, aquatic eutrophication, and decreased biodiversity (Gu et al. 2015). In China, N deposition is increasing at an average of 0.18–0.41 kg N ha⁻¹ a⁻¹ (Liu et al. 2013a; Lu and Tian 2014) and it is important to understand the rates and sources of N deposition to take appropriate measures. However, mapping wet N deposition using monitoring stations in a region, especially in remote areas and in regions with many land-use types, is costly (Zhu et al. 2015). Further, accurately determining the wet N deposition rates and sources is very difficult because there are many N compounds (Harmens et al. 2011, 2014; Zechmeister et al. 2008). Therefore, the rates and sources of wet N deposition need to be urgently estimated with a simple and inexpensive method.

During the past several decades, many researchers have used biomonitoring organisms, such as mosses, lichen, and vascular plants to investigate the rates, compositions, and sources of atmospheric N deposition (e.g. Bragazza et al. 2005; Christie 1987; Harmens et al. 2014). First, mosses absorb N mainly from atmospheric wet and dry depositions because of the lack of a well-developed root system and protective cuticle and the unique specific leaf area (Harmens et al. 2014). Second, reduced N (NH_x) generally has a more negative N isotopic composition (δ¹⁵N) than oxidized N (NO_y) in atmospheric deposition, and mosses utilize atmospheric N with little isotope fractionation (Liu et al. 2008a, 2012; Xiao et al. 2010b). Third, moss is widely distributed and can be used for biomonitoring N deposition in some regions to compensate for the lack of automated instruments and passive monitoring (Diaz-Alvarez, Lindig-Cisneros, and de la Barrera 2018; Harmens et al. 2011). Therefore, mosses have been broadly used by researchers to investigate N deposition levels and its effects since the 1980s.

Previous studies have reported spatial and long-term changes in atmospheric N deposition according to the relationship between moss N content and atmospheric N deposition (Harmens et al. 2011; Poikolainen et al. 2009). Moreover, moss δ¹⁵N, which indicates the species and sources of atmospheric N deposition (Liu et al. 2008b; Qu et al. 2016), has been studied. Recently, δ¹⁵N of moss was used to study the mechanism of moss N utilization which could be more accurate to indicate the N deposition (Liu et al. 2012, 2013b). However, moss biomonitoring of atmospheric N deposition is mainly concentrated in Europe and Southern China. The Yangtze River Delta (YRD) has experienced serious N deposition (Ti et al. 2018), but there are few monitoring sites for atmospheric N deposition. Moreover, the relationship between moss N content and N deposition in the YRD has not been reported.

In this study, moss tissue and deposition were sampled within a distance of 2 km of each other at six monitor sites in the YRD with three land-use types – urban, suburban, and rural. The objectives of this study were to indicate the relationships between moss (Haplocladium microphyllum) N content and wet N deposition, between moss δ¹⁵N and precipitation δ¹⁵N, and to explore the sources and compositions of atmospheric wet N deposition in the YRD.

2. Materials and methods

2.1. Study area

The YRD has a subtropical monsoon climate with a mean temperature and rainfall of 15°C and 1050 mm, respectively. Rainfall has an obvious seasonal pattern, with more than 70% falling during the months of May to October. The land area in the YRD region is 5.63 × 10⁴ km², of which agriculture, urban, aquatic, forest, and grassland types occupy 41.9%, 23.4%, 21.0%, 6.2%, and 5.8% of the land, respectively. Six sites in the YRD were selected to represent the three main land-use types, i.e. urban, suburban, and rural (Figure 1, Table S1). Suiyuan (SY) and Hehuachi (HHC) are close to the centers of Nanjing and Yangzhou, respectively, and are surrounded by residential, transportation, commercial, and catering areas— representing an urban area. Xianlin (XL) and Yangzijin (YZJ) are in university towns of Nanjing and Yangzhou, respectively, mainly in the cultural, educational, and residential areas— representing a suburban area. Tongshan (TS) is surrounded by cropland, mainly by rice and wheat cropping rotations, which receive chemical N fertilizer of 550 kg N ha⁻¹ a⁻¹. Yuyangli (YYL) belongs to Suzhou city and is close to the Taihu Lake drinking water source protection area with few human activities. TS and YYL belong to rural area. A detailed description of the sampling sites can be found in previous studies (Chen et al. 2019).

Figure 1. Map showing (a) the location of Yangtze River Delta (YRD) and (b) sampling sites. The image of each site derived from Google Earth.

2.2. Sample collection and analysis

To avoid possible interspecific and seasonal differences (Dong et al. 2017), the epilithic moss Haplocladium microphyllum, was collected on July 2017, which commendably biomonitors atmospheric N deposition (Liu et al. 2008b). Mature and green tissues from three to five subsampling sites were collected within a 2 km range of each wet N deposition sampling site. Three to five samples were mixed as one composite sample for each site (Figure 1). All mosses were collected from bare rock without canopies to avoid the effect of the canopies and that of falling N compounds. The sampling sites are more than 300 m away from the main roads and are considered to be completely free of the direct impact of stagnant water and surface water splashes, traffic, and other man-made pollution sources (human and animal excrement, fertilization, and stamping).

The collected fresh moss samples were stored in self-sealed bags. After dead branches, leaves, and debris attached to the mosses were removed, the mosses were sonicated and repeatedly washed with deionized water to remove N components from the moss surface. They were dried in an oven at 70°C, ground with liquid N, and then dried to a constant weight and passed through a 100-mesh sieve. The moss N content was measured using an elemental analyzer (PE2400Ⅱ, PerkinElmer, USA) with an analytical precision of 0.1%. N isotope ratios were analyzed on a Finnigan MAT 252 mass spectrometer with an analytical precision of 0.2‰, after combustion at 850°C and purification with liquid N. Analysis of potassium nitrate standard (MOR2386-01, +1.9‰), provided by Shoko Co., (Tokyo, Japan), gave a mean (±SD) δ¹⁵N_air value of 1.9‰ ± 0.2‰ (n = 5). Accuracy and recovery of N and δ¹⁵N were checked analysing a sample of this standard material after each set of eight moss analyses.

The TN (total N), NH₄⁺ (ammonium), and NO₃⁻ (nitrate) concentrations of rainwater samples were analyzed using a UV–3600 spectrophotometer (Shimadzu UV–VIS–NIR Spectrophotometer, Shimadzu Co., Kyoto, Japan) (Chen et al. 2019). The δ¹⁵N-NH₄⁺ and δ¹⁵N-NO₃⁻ of precipitation were determined using a diffusion method (Zhang et al. 2017).

2.3. Statistical analysis

Statistical analysis was conducted using SPSS 18.0 (SPSS Inc., Chicago, USA), and graphs were created with OriginPro 9.0 (OriginLab Corporation, Northampton, MA, USA). Linear curve fit analyses were used to examine the relationship between moss N content and N deposition rates, between moss δ¹⁵N and precipitation δ¹⁵N. Here, p < 0.05 is considered statistically significant unless stated otherwise.

3. Results

3.1. Moss nitrogen content

The average N contents of urban, suburban, and rural mosses were 2.72% ± 0.07%, 2.41% ± 0.07%, and 2.05% ± 0.27%, respectively (Figure 2(a)). The lowest value of moss N content was observed at YYL in the rural area, and the highest value occurred at the center of Yangzhou (HHC). A significant decrease in moss N content was found from urban to suburban and rural areas (Figure 2(a)). The decrease trend corresponded with the wet N deposition rates across the three land-use types (Figure 2(c)). There was a significant correlation between moss N content and wet TN (p < 0.05) and NH₄⁺ (p < 0.01) depositions (Figure 3(a,b)). The wet NH₄⁺ deposition had a greater influence on the N content, which can be seen from the stronger correlation between the two than that between wet NO₃⁻ deposition and the N content (Figure 3(b,c)).

Figure 2. (a) Moss N content, (b) moss δ¹⁵N, (c) wet TN deposition rate, and (d) precipitation δ¹⁵N-IN across the three land-use types, i.e. urban, suburban, and rural: different small letters above the bar stand for a significant difference at the 0.05 level across the three land-use types (the δ¹⁵N of different sources is cited from Xiao et al. (2010b) and Freyer (1978)). The primary values of δ¹⁵N-NH₃ from fertilizer vary widely, for example, ranged from −51‰ to −35‰ (Elliott et al. 2019), ranged from −5.9‰ to +6.6‰ with mean −0.2‰ ± 2.1‰ (n = 29) (Bateman and Kelly 2007). What’s more, the δ¹⁵N of fertilizer use ranges from −5‰ to 0 (Freyer 1978), which has been used in the previous studies on mosses monitoring rates and sources of atmospheric N deposition (e.g. Liu et al. 2012; Qu et al. 2016; Xiao et al. 2010b). So, we selected the later.

Figure 3. The relationship between moss N content and (a) TN, (b) NH₄⁺, and (c) NO₃⁻, and the relationship between moss N isotope and (d) δ¹⁵N-IN, (e) δ¹⁵N-NH₄⁺, (f) δ¹⁵N-NO₃⁻, and (g) NH₄⁺/NO₃⁻ of precipitation.

3.2. Nitrogen isotopic composition in moss

All epilithic mosses had negative δ¹⁵N signatures from −10.21‰ to −1.58‰ in the YRD. The moss δ¹⁵N was in the order: urban (−2.35‰ ± 0.77‰) > rural (−3.81‰ ± 1.84‰) > suburban (−6.14‰ ± 2.85‰) sites (Figure 2(b)). The moss δ¹⁵N of urban was significantly higher than that of suburban. There was a significant correlation between moss δ¹⁵N and precipitation δ¹⁵N-IN (inorganic N, i.e. NH₄⁺ + NO₃⁻), δ¹⁵N-NH₄⁺, and δ¹⁵N-NO₃⁻ (Figure 3(d–f)). However, there was no significant correlation between moss δ¹⁵N and the ratio of ammonium to nitrate (NH₄⁺/NO₃⁻) in this study (Figure 3(g)).

4. Discussion

4.1. Moss N content for indicating atmospheric N deposition

In this study, the moss N content (1.85%–2.79%) was similar to that of south China— for example, the Yangtze River drainage basin (1.19%–4.05% (Xiao et al. 2010a)), Guiyang (1.85%–2.48% (Dong et al. 2017), 0.85%–2.97% (Liu et al. 2008c)), Changsha (1.80%–4.05% (Xiao et al. 2010b)), and Nanchang (1.61%–4.64% (Xiao et al. 2010b)). However, it was much higher than that of Europe, e.g. 0.76%–1.99% in Austria (Zechmeister et al. 2008), 0.41%–1.79% in Finland (Poikolainen et al. 2009), and 0.8%–1.6% in 16 European countries (Harmens et al. 2011). The moss N content of YYL (1.85%) was higher than that of the background (Gongga Mountain) in China (1.5% (Liu et al. 2008b)), indicating that wet N deposition was higher in the YRD. There was a linear relationship between moss N content and wet TN deposition (Figure 3(a)). The coefficient was higher than that in Guiyang (0.05) (Liu et al. 2008b; Xiao et al. 2010a) and Europe (0.08) (Harmens et al. 2011). This is mainly because previous studies used a bulk N deposition, while only wet N deposition was used in this study. The type of N deposition affects moss N content. Dry deposition can cause a 0.04%–0.07% increase in moss N content for each 1 kg N ha⁻¹ a⁻¹, reaching up to 4% in regions with high dry ammonia (NH₃) deposition, while reaching only up to 1.6% in regions with wet deposition (Diaz-Alvarez, Lindig-Cisneros, and de la Barrera 2018). This coefficient decreases to a similar value of 0.05, if one estimates that the wet N deposition accounts for 45.5% of the total N deposition in the YRD (Ti et al. 2018). Moreover, when the N deposition is less than 20 kg N ha⁻¹ a⁻¹, the moss N content is more sensitive to N deposition (Bragazza et al. 2005; Diaz-Alvarez, Lindig-Cisneros, and de la Barrera 2018).

As in previous research, the correlation between moss N content and wet NH₄⁺ deposition was stronger than that between moss N content and wet NO₃⁻ or TN deposition (Liu et al. 2012) (Figure 3(a–c)), because NH₄⁺ had a greater effect on moss N content than NO₃⁻ or TN. Previous studies have revealed that moss prefers NH₄⁺ and dissolved organic N over NO₃⁻ due to an energetic economy; the studies have also revealed the inhibition of nitrate reductase activity by reduced dissolved N (Liu et al. 2012; Wiedermann et al. 2009). Therefore, it would be better to use moss N content to estimate N deposition in the regions with the dominance of NH_x than the dominance of NO_y.

4.2. Moss δ¹⁵N for indicating main sources of atmospheric N deposition

Previous studies have reported that the two main atmospheric N sources have different δ¹⁵N values, with δ¹⁵N-NH₄⁺ tending to be negative and δ¹⁵N-NO₃⁻ tending to be positive (Pearson et al. 2000). NH_x deposition should be predominant in the YRD according to the lower moss δ¹⁵N (−10.21‰ to −1.58‰) and have values similar to the δ¹⁵N-NH₄⁺ of wet deposition (−15.00‰ to 0) (Figure 2(b)). This is similar to the directly measured result that the NH₄⁺/NO₃⁻ of atmospheric N deposition was 1.5 in this region (Chen et al. 2019; Ti et al. 2018). Moreover, there was a significant negative linear relationship between moss δ¹⁵N and NH₄⁺/NO₃⁻ in western Germany (Solga et al. 2005) and in the Yangtze River drainage basin of China (Xiao et al. 2010a). However, there was no significant correlation between the δ¹⁵N of moss and NH₄⁺/NO₃⁻ (Figure 3(g)), which was similar to that in 11 European countries (Bragazza et al. 2005). Moreover, there was no significant correlation between moss δ¹⁵N and NH₄⁺/NO₃⁻ when using all the data (Figure 3(g)). Therefore, more monitoring data are needed to verify the relationship between moss δ¹⁵N and NH₄⁺/NO₃⁻, and moss δ¹⁵N should be used with caution while determining component of wet N deposition.

More positive δ¹⁵N values were observed in urban mosses in the YRD (−3.41‰ to −1.58‰), which is similar to that of urban Changsha in China (−6.75‰ to 1.94‰) (Xiao et al. 2010b) and the center of the Ferrara municipality in Italy (−5.27‰ to 1.62‰) (Gerdol et al. 2002). Many previous monitoring-based studies have indicated that the main source of atmospheric N deposition is fossil fuel combustion and vehicle emissions in urban areas (Wang et al. 2017). Furthermore, Pearson et al. (2000) found a very good correlation between moss δ¹⁵N and traffic exposure, where the average δ¹⁵N of moss near busy urban roads is +3.66‰ (−1‰ to +6‰). The mean moss δ¹⁵N in suburban areas (−6.14‰ ± 2.85‰) was found to be close to that near farm buildings in the vicinity of poultry or cattle pens (−7.8‰ (Pearson et al. 2000)) and the Guiyang urban site (−6.5‰ ± 1.1‰ (Liu et al. 2008b)) with atmospheric N deposition primarily from sewage and wastes. More negative moss δ¹⁵N values in suburban areas indicate the contribution of NH₃ from untreated sewage and wastes, whereas the relatively less negative moss δ¹⁵N values in rural areas mainly come from agricultural NH₃ (Figure 2(b)). The moss δ¹⁵N in rural areas (−3.81‰ ± 1.84‰) was similar to that in rural Nanchang (−3.1‰ ± 2.2‰) and Changsha (−4.7‰ ± 1.1‰) (Xiao et al. 2010b). Previous emission inventory studies have indicated that synthetic fertilizer application contributed 33%–40% of the total NH₃ emissions in the YRD (Fu et al. 2013). In addition, small farms and poultry breeding households, which are very common in Chinese small villages, were considered an important source of NH₃ in the rural areas (Fu et al. 2013). Therefore, the δ¹⁵N of moss can be used as an indicator of the sources of atmospheric wet N deposition.

5. Conclusions

The directly measured moss N content and wet N deposition rate demonstrates that wet N deposition can be estimated by the N content of Haplocladium microphyllum in a region with different land-use types. Moss N content significantly decreased from urban (2.72% ± 0.07%) to suburban (2.41% ± 0.07%) to rural (2.05% ± 0.27%) areas, and the trend was similar with wet N deposition rate across the three types of land use. Moss δ¹⁵N values can effectively decipher sources of wet N deposition. More negative δ¹⁵N of suburban (−6.14‰ ± 2.85‰) and rural (−3.81‰ ± 1.84‰) mosses indicated N mainly released from agriculture and effluent, while the less negative δ¹⁵N of urban (−2.35‰ ± 0.77‰) mosses were mainly influenced by fossil fuel combustion and traffic emissions. The negative moss δ¹⁵N indicated that the main form in the N deposition was NH_x in the YRD. However, more monitoring data are needed to verify the relationship between moss δ¹⁵N and NH₄⁺/NO₃⁻ of N deposition in the future.

Acknowledgments

We sincerely thank Hejuan ZHAO at Tongshan and Sheng JIANG at Yuyangli stations for their contributions to the rainfall sampling.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be access here.

Additional information

Funding

This study was funded by the National Natural Science Foundation of China [grant numbers 41571324, 41673108, 41701329, and 41771380]; the National Program on Key Basic Research Project (973 Program) [grant number 2014CB953800]; the Natural Science Foundation of the Jiangsu Higher Education Institutions of China [grant number 16KJD170001]; the Natural Science Foundation of Hainan Province, China [grant number 317190]; and the CAGS Research Fund [grant numbers YYWF201637, 201724].