Nitrate dynamics in natural plants: insights based on the concentration and natural isotope abundances of tissue nitrate.

The dynamics of nitrate (NO(-) 3), a major nitrogen (N) source for natural plants, has been studied mostly through experimental N addition, enzymatic assay, isotope labeling, and genetic expression. However, artificial N supply may not reasonably reflect the N strategies in natural plants because NO(-) 3 uptake and reduction may vary with external N availability. Due to abrupt application and short operation time, field N addition, and isotopic labeling hinder the elucidation of in situ NO(-) 3-use mechanisms. The concentration and natural isotopes of tissue NO(-) 3 can offer insights into the plant NO(-) 3 sources and dynamics in a natural context. Furthermore, they facilitate the exploration of plant NO(-) 3 utilization and its interaction with N pollution and ecosystem N cycles without disturbing the N pools. The present study was conducted to review the application of the denitrifier method for concentration and isotope analyses of NO(-) 3 in plants. Moreover, this study highlights the utility and advantages of these parameters in interpreting NO(-) 3 sources and dynamics in natural plants. We summarize the major sources and reduction processes of NO(-) 3 in plants, and discuss the implications of NO(-) 3 concentration in plant tissues based on existing data. Particular emphasis was laid on the regulation of soil NO(-) 3 and plant ecophysiological functions in interspecific and intra-plant NO(-) 3 variations. We introduce N and O isotope systematics of NO(-) 3 in plants and discuss the principles and feasibilities of using isotopic enrichment and fractionation factors; the correlation between concentration and isotopes (N and O isotopes: δ(18)O and Δ(17)O); and isotope mass-balance calculations to constrain sources and reduction of NO(-) 3 in possible scenarios for natural plants are deliberated. Finally, we offer a preliminary framework of intraplant δ(18)O-NO(-) 3 variation, and summarize the uncertainties in using tissue NO(-) 3 parameters to interpret plant NO(-) 3 utilization.

Plant nitrate (NO−3) in a natural context

Nitrogen (N) is a key factor in the control of the primary productivity in terrestrial plant ecosystems (Vitousek and Howarth, 1991; LeBauer and Treseder, 2008). Among the N species available to plants, ammonium (NH+4) is dominant in the inorganic N of unfertilized soils (Schimel and Bennett, 2004) and atmospheric N deposition (Stevens et al., 2011). Some plants prefer NH+4 (Britto and Kronzucker, 2013) while the roots of a few plants directly absorb organic N (Chapin et al., 1993; Näsholm et al., 2009; Hill et al., 2013). However, nitrate (NO−3) is an important N source for all plants because of its versatile functions in both plant nutrition and physiological regulations (Raven, 2003; Wang et al., 2012). The utilization of NO−3 (mainly uptake and reduction/assimilation) has been investigated intensively in plants through characterization of related enzymes including nitrate reductase (NR) and nitrite reductase (NiR) and their activities (NRA and NiRA, respectively) in response to different environmental conditions (Beevers and Hageman, 1969; Atkin et al., 1993; Kronzucker et al., 1995; Campbell, 1999). The framework of plant NO−3 studies has expanded in the past few decades due to the availability of molecular techniques. A few model plants have been used in understanding the transporters responsible for NO−3 uptake and transportation (Wang et al., 2012). Besides its function in nutrient supply, plant NO−3 and its metabolism contain unique information related to the mediation of plant physiology, diversity, and the ecosystem N cycle (Crawford, 1995; Tischner, 2000). However, evolution has yielded diverse strategies by which plants acquire N and NO−3 from natural environments to adapt to changes in ecosystem N availability (Chapin, 1980; Raven and Yin, 1998; Nacry et al., 2013). Therefore, there are considerable uncertainties in assessing the utilization of NO−3 by plants in natural habitats, which cannot be explained fully by laboratory-based mechanisms because of methodological constraints. Consequently, a great need exists for a straightforward estimation of plant NO−3 availability and a mechanistic understanding of the processes controlling plant NO−3 uptake and reduction. These can enhance our understanding of the role of plant NO−3 utilization in the ecosystem N cycle and the changes of plant growth and diversity with ecosystem N status (Lambers et al., 2008; Bloom et al., 2010; Boudsocq et al., 2012).

Denitrifier method for NO−3 in natural plants

Natural abundance of stable isotopes in natural plants can integrate the information related to N sources and physiological processes (Högberg, 1997; Robinson, 2001; Craine et al., 2009). The stable isotopes include δ15N, δ18O, and δ17O for NO−3; 15N:14N, 18O:16O, and 17O:16O ratios expressed relative to atmospheric N2 and standardized mean ocean water (VSMOW), respectively (Coplen, 2011). These isotopes have been broadly used for studying plant N strategies and enzymatic dynamics in natural settings (Evans, 2001; Tcherkez and Farquhar, 2006; Granger et al., 2010). Nevertheless, it is difficult to measure the concentration and isotopes (δ15N and δ18O) of NO−3 in plant tissues precisely using traditional methods (Liu et al., 2012a). The use of the denitrifier method for measuring low (sub-nanomole) concentrations of NO−3 ([NO−3]) started during the mid-1980s (Lensi et al., 1985). The method has high sensitivity and is especially applicable for samples with low [NO−3] but with high dissolved organic carbon (DOC) (Christensen and Tiedje, 1988; Binnerup and Sørensen, 1992; Aakra et al., 2000). The denitrifier method developed for both δ15N and δ18O analysis is based on the isotopic analysis of nitrous oxide (N2O). The N2O is converted from sample NO−3 by cultured denitrifying bacteria (Pseudomonas aureofaciens; ATCC 13985) that lack N2O reductase activity (Sigman et al., 2001; Casciotti et al., 2002). The method was initially performed on seawater with 20–50 nmol NO−3. Since then, the application has been expanded widely to accommodate isotopic analysis of NO−3 in fresh water (e.g., groundwater, stream water, precipitation), soil and sediment water, soil extracts, as well as dissolved organic N (DON) in seawater and DON bound to diatoms as described by Koba et al. (2010a) and McIlvin and Casciotti (2011), respectively. This method has recently been used for measurements of NO−3 in natural plants and crops (Liu et al., 2012a, 2013a; Laursen et al., 2013; Bloom et al., 2014; Mihailova et al., 2014). The established protocol facilitates the Δ17O (Δ17O = [1 + δ17O] / [1 + δ18O]0.5247 − 1; Kaiser et al., 2007) analysis of leaf NO−3 to diagnose atmosphere-derived NO−3 in leaf uptake (Mukotaka, 2014).

The denitrifier method enables more precise measurements of subnanomole amounts of NO−3 (Binnerup and Sørensen, 1992; Højberg et al., 1994) as compared to traditional methods that use flow injection analysis, ion chromatography, high-performance liquid chromatography, and Kjeldahl distillation. Thus, the denitrifier method overcomes the difficulties in determining NO−3 in plant, soil, and sediment samples (Norwitz and Keliher, 1986; Anderson and Case, 1999; Alves et al., 2000). Moreover, it greatly simplifies the pretreatment procedures and reduces the risk of contamination during plant NO−3 isotopic analysis (see the old δ15N protocol in Volk et al., 1979 and Evans et al., 1996). The denitrifier method especially avoids the influence of DOC in plant extracts (Haberhauer and Blochberger, 1999) on the δ18O of NO−3 (Figure 1) that was previously measured as carbon monoxide with TC/EA-IRMS (Michalski, 2010).

Figure 1

Assigned isotopic ratios (.

Compared with NRA assays, concentrations and isotopic signatures of tissue NO−3 provide more authentic evidence related to NO−3 uptake and reduction under in situ N availability. In vitro and in vivo NRA measurements (Stewart et al., 1992, 1993) do not reflect the in situ ability of plant NO−3 reduction. This is because firstly, the added amount of NO−3 (often at the micromolar level) during NRA assays is uniform. Moreover it is much higher than normal NO−3 availability and the endogenous NO−3 in natural plants. The synthesis of the NR enzyme or the activation of NRA, however, is substrate-inducible (Beevers and Hageman, 1969; Somers et al., 1983; Campbell, 1999). Secondly, the reagents used in the assay can affect the estimation of NRA. Different analytical settings (e.g., with or without ethanol) can alter the fluxes of NO−3 and photosynthate, resulting in different estimations (Ferrari and Varner, 1970; Aslam, 1981). Thirdly, NRA might be altered by pH adjustment and vacuum infiltration during the NRA analysis. High DOC concentrations in the plant extract also easily destroy the precision of the colorimetric determination of NO−3 or nitrite (NO−2) (Alves et al., 2000).

Since natural isotope analysis does not require artificial N addition, it presents no risk of changing the soil N pools and plant N-uptake kinetics (Liu et al., 2012b). The natural abundance approach does not disturb the N pools in plants and provides information related to the NO−3 behavior in plant tissues based on isotopic compositions and fractionations. In fact, the field application of 15NO−3 tracer is advantageous in terms of the total and short-term incorporation of NO−3 into plants (e.g., McKane et al., 2002; Wanek and Zotz, 2011). However, the added tracer cannot bypass the influence of soil microbial activity, which can greatly change the picture of N uptake and preference over time (Harrison et al., 2007). Measurements of cytosolic and vacuolar NO−3 concentrations have been conducted to explore factors controlling uptake, intracellular transport and assimilation. However, related techniques such as compartmental radiotracer (e.g., 13N; Kronzucker et al., 1995), efflux analysis, nuclear magnetic resonance, cell fractionation, and NO−3-selective microelectrodes showed high cost and low field operability (Zhen et al., 1991; Miller and Smith, 1996). The calculated [NO−3] is especially sensitive to the small error of the estimation of cytosolic and vacuolar volumes, the precisions of which are difficult to ascertain.

Major sources and processes of NO−3 in natural plants

Root NO−3 uptake from the soil is achieved by active transportation (Wang et al., 2012). The extracellular NO−3 enters the cytosol of plant cells where it is either reduced by NR to NO−2 or stored in the vacuoles (Figure 2). The NO−2 will be transported into plastids (in root) or chloroplasts (in leaf) and reduced further by NiR to reduced N (Figure 2). Both NRA and NiRA are well known to be substrate-inducible, meaning that the de novo synthesis of the enzyme results from the presence and increase of the NO−3 in plants (Beevers and Hageman, 1969; Campbell, 1999). The induction of NRA by both soil and airborne NO−3 is an important mechanism to elucidate the interactions among NO−3 uptake, translocation/allocation, and reduction dynamics (Norby et al., 1989; Scheible et al., 1997a; Tischner, 2000).

Figure 2

Schematic map showing major NO.

The NO−3 transported by the xylem flow, either directly from soil or partially processed by root NR, is the initial NO−3 reaching leaves and shoots (Peuke et al., 2013). This is especially true for plants growing at some pristine sites (e.g., arctic tundra) where the atmospheric NO−3 availability is negligible. However, in regions with substantial NO−3 deposition, both atmospheric NOx and NO−3 serve as potential sources of NO−3 in leaves (Wellburn, 1990; Raven and Yin, 1998; Sparks et al., 2001), especially for non-vascular plants such as mosses, which rely more on atmospheric nutrients (Liu et al., 2012c). Leaf NO−3 acquisition from the atmosphere is conducted through passive diffusion mechanisms wherein uptake through the stomata is dominant (Wellburn, 1990; Raven et al., 1992; Gessler et al., 2002) (Figure 2). The leaf-accessible NO−3 in the atmosphere includes an array of inorganic and organic ions and compounds (Wellburn, 1998; Teklemariam and Sparks, 2004; Vallano and Sparks, 2008). Although, previous tracer studies have described their incorporation into leaves (Hanson and Garten, 1992; Yoneyama et al., 2003; Lockwood et al., 2008), it is rather difficult to apply the natural abundance method for estimating field contributions of atmospheric NO−3. This can be attributed to the heterogeneity in chemical and deposition forms, and temporal and spatial distributions (Sievering et al., 2007; Sparks, 2009).

Concentration levels and implications of NO−3 in natural plants

Nitrate cannot be produced in photoautotrophic plants, except in a few legumes (Hipkin et al., 2004). The presence of NO−3 in any part of a plant constitutes evidence of NO−3 uptake by the plant and reflects that external NO−3 is available; and that the rate of uptake is higher than the rate of reduction. The NO−3 that is extractable from a plant organ is often a sum of the amounts from the extracellular pool, cytosolic pool, and vacuolar pool (Figure 2). These pool sizes and turnover rates are regulated by both environmental and physiological factors (Zhen et al., 1991; Miller and Smith, 1996), which determine the isotopic signatures of the extracted NO−3. Generally, the concentration level and distribution of NO−3 in vascular plants and the variations among species is a complex result of two important factors: external availability (previously often evaluated through NO−3 concentration and net nitrification rate in soil) and physiological strategies (mainly including uptake, translocation, and reduction dynamics). Moreover, the external factors also consider the availability of NO−3 relative to NH+4 or other N sources because it can influence both plant NO−3 uptake and assimilation (Boudsocq et al., 2012; Liu et al., 2012c; Britto and Kronzucker, 2013) while the physiological factors include the affinity of plants to different soil NO−3 levels (Wang et al., 2012; Kalcsits and Guy, 2013).

First, the distribution of organ-specific NO−3 concentrations among plants under different growing conditions (Figures 3, 4A) showed that plants growing in natural soils might also have a high NO−3 accumulation. In natural forests, leaf NO−3 concentrations of some species can be as high as 1000–10000 μ g-N g−1 dw (Figure 4A; Gebauer et al., 1988; Koyama et al., 2013), which was even higher than those of some crops (e.g., Bloom et al., 2014) and N-polluted natural plants (Figure 3). Plant NO−3 concentrations are indicators or predictors of the soil N cycle (e.g., soil nitrification and soil NO−3) and forest N pollution (Stams and Schipholt, 1990; Aber et al., 1998; Fenn and Poth, 1998; Koba et al., 2003). Such concentrations show higher sensitivities than bulk N and NRA parameters in revealing species-level responses to N enrichment (Fenn et al., 1996; Jones et al., 2008; Tang et al., 2012). The increase in NO−3 concentration in roots and or leaves with external NO−3 was observed under both natural soil conditions and experimental N addition (e.g., Stewart et al., 1993; Lexa and Cheeseman, 1997; Wang and Schjoerring, 2012). However, the level of leaf NO−3 and its response to soil NO−3 variation differ among species with distinct uptake or accumulation rates. For example, the NO−3 concentrations in plants (mostly as mosses) we recently investigated (Liu et al., 2012a,c, 2013a) were much lower than those reported by Gebauer et al. (1988) or Koyama et al. (2013) on vascular plants (Figure 4A) when compared within a similar soil [NO−3] range (e.g., 0–15 mg-N kg−1 soil, dw). Besides, the correlation between leaf NO−3 and soil NO−3 is apparent for plants with low NO−3 concentrations (Figure 4A). However, synthesis or extrapolation to different plants with distinct NO−3 accumulation abilities should be done carefully when evaluating soil N enrichment or N saturation.

Figure 3

Tissue NO.

Figure 4

(A) Relation between NO−3 concentrations in soil and natural plants. Plant NO−3 data in the left panel are shown for individual samples in Guiyang, southwestern China and western Tokyo, Japan reported by Liu et al. (2012a, 2013a). Plant NO−3 data in the right panel show organ-specific and whole-plant concentrations (averages of different species) in ecosystems of Central Europe (see details in Gebauer et al., 1988), and leaf NO−3 of different species (H. hirta, P. japonica, L. stellipilum, L. triloba) in a temperate forest of central Japan (Koyama et al., 2013). (B) Relations between total N, C/N, and tissue NO−3 concentration in natural plants. Mosses include different species in different habitats of Guiyang, Southwestern China, and Western Tokyo, Japan (cited from Liu et al., 2012a,c). Vascular leaves I, petioles and roots were reported for a coniferous and a broadleaved plant in western Tokyo, Japan (cited from Liu et al., 2013a). Vascular leaves II included fern, oak, and pine species at the Camp Paivika and Camp Osceola forest sites in the San Bernardino Mountains of southern California, USA (cited from Fenn et al., 1996).

Second, considerable differences (up to 4–5 orders) exist in the level of NO−3 among plant organs and species (Figures 3, 4A). The organ-specific patterns of NO−3 accumulation among coexisting plants can differ with soil N availability and the plant growing stage (Gebauer et al., 1984; Stewart et al., 1993; Liu et al., 2013a). However, this has complicated the use and selection of proper organs and species to evaluate ecosystem N availability based on tissue NO−3 analysis. McKane et al. (2002) used 15N tracers in the field to show that NO−3 uptake in the tundra plants did not passively follow external availability, but depended on specific ecophysiological traits. NO−3 preference in Carex was determined by the appearance of 15N tracer in Carex biomass, which showed that the NO−3 preference might reflect only the 15NO−3-acquiring efficiency associated with root traits, but not NO−3 assimilation given significantly lower NRA in Carex than in other species (Nadelhoffer et al., 1996). Therefore, additional studies should be conducted to determine the extent of organ-specific and species-specific variability of NO−3 concentration that reflects plant NO−3 strategy, and the heterogeneity of NO−3 available to roots. The available data for natural plants revealed a clear increase in NO−3 concentration with bulk N while a decrease with C/N (a clear turning at the C/N of 20–30) in different organs or tissue types (Figure 4B). Similarly, Zhen and Leigh (1990) reported that shoot NO−3 accumulated as a linear function of bulk N in wheat plants once a threshold N was exceeded. These results reflected the regulation of overall physiological N demand on the NO−3 utilization in natural plants (Imsande and Touraine, 1994). The regulation might be unidirectional because the contribution of NO−3 to bulk N assimilation appears to be much lower than that for other N forms in plants (portrayed in Figure 4B). The complexity of the mutual regulations behind the inverse relation between NO−3 and C/N might be comparable with the multi-scale inverse relation prevailing between NO−3 and organic C observed in different ecosystems (Taylor and Townsend, 2010). So far, little direct and simple evidence has been obtained for the driving mechanisms of C and N metabolism on NO−3 uptake, allocation, and accumulation in natural plants. A clearer relation is that even when external NO−3 is uniform, the NO−3 concentration is often higher in organs (especially for growing leaves) of species with higher NRA than in those with lower NRA (Gebauer et al., 1988; Cruz et al., 1991; Widmann et al., 1993; Min et al., 1998). Mutual induction between the maintenance of high NO−3 concentration and that of NR synthesis or NRA activation were elucidated in view of C metabolism and N demand in response to availability and growing conditions (Stewart et al., 1993; Scheible et al., 1997a,b; Scheurwater et al., 2002). The lower NO−3 concentration and NRA might be associated with lower N metabolism and demand in organs and plants with higher C/N and vice versa. Therefore, except regulation by soil NO−3 concentration, the uptake and distribution of NO−3 in a plant might follow the regime of organ-specific or whole-plant metabolic activities.

Other factors such as light and water regimes might also influence plant NO−3 accumulation through the pathway of photosynthetic regulation (Widmann et al., 1993; Simon et al., 2014). Cárdenas-Navarro et al. (1999) found concurrent and linearly correlated changes in whole-plant NO−3 and water content during the day–night cycle, reflecting a homoeostasis effect of endogenous NO−3 concentration. Besides, as discussed above, the heterogeneity of soil NO−3 available to roots of coexisting species should not be excluded considering the differences in root morphology and spatial distribution. Given the difficulties in determining rhizospheric soil NO−3 concentration and flux, it would be promising to measure NO−3 concentrations in roots to evaluate NO−3 availability to the whole plant or aboveground organs.

Isotopic systematics of NO−3 in plants

Stable isotopes of NO−3 in plants are controlled mainly by NO−3 sources and isotopic effects involved in NO−3 acquisition and reduction processes (Robinson et al., 1998; Comstock, 2001; Evans, 2001; Cernusak et al., 2009).

The δ15N of NO−3 in soil is reported mostly within −10 to +10‰ however, the δ15N of newly-produced NO−3 in soil is usually low because of strong isotopic effects of nitrification, on the other hand, the values can be elevated at sites with marked denitrification (Mariotti et al., 1981; Högberg, 1997; Koba et al., 1998, 2003, 2010b; Houlton et al., 2006; Takebayashi et al., 2010). Atmospheric NO−3 has a wider δ15N range (−15 – +15‰) because of its complex production pathways and sources (Heaton, 1990; Felix et al., 2012; Altieri et al., 2013). The δ15N of NO−3 is generally lower in wet than in dry deposition (Heaton et al., 1997; Elliott et al., 2009), but both often show a δ15N range overlapping with that of soil NO−3. The δ18O of initial NO−3 produced in soil is usually estimated using the δ18O of in situ H2O (normally −25 – 4‰) and atmospheric O2 (ca. 23.5‰) in a 2:1 ratio, assuming no exchange and fractionation of oxygen (O) isotopes occurs during nitrification and the NO−3 is produced solely through chemolithoautotrophic nitrification (Amberger and Schmidt, 1987). However, kinetic isotopic fractionation and O exchange between NO−2 and H2O often occur during nitrification, which can eliminate the isotopic signal of O2 effecting lower δ18O than the predicted values (Fang et al., 2012). The O of NO−3 in atmospheric deposition is derived mainly from O2 and O3, which have distinctly higher δ18O and Δ17O signatures than those of soil NO−3. In contrast to the overlapping δ15N for different NO−3 sources, δ18O and or Δ17O provide a clear separation between soil and atmospheric NO−3 sources. The δ18O of soil NO−3 produced by nitrification is distinctly lower (mean = −4.0‰ −7.3 to −0.9‰ Fang et al., 2012) than that of atmospheric NO−3 (60 − 100‰). The latter has high Δ17O values (around 25‰) in contrast to 0‰ of soil-derived NO−3 (Kendall et al., 2007; Michalski, 2010; Costa et al., 2011) (Figure 5).

Figure 5

Preliminary relation between δ. The δ18O and Δ17O values were considered respectively, as −5 to 5‰ and 0‰ for soil NO−3 (black and solid line), 70 and 25‰ for atmospheric NO−3 (red square). Dashed lines show the isotopic range of mixing between atmospheric and soil sources. Dashed lines with arrows show the vectors of δ18O enrichments because of NR reduction.

The process of NO−3 entry into root cells and subsequent transport processes within plants per se cause no isotope effect because of the lack of bond breakage. However, the acquisition of NO−3 through mycorrhizae to root cells potentially causes an isotopic difference between tissue NO−3 in roots and NO−3 in soil. Root NO−3 may be enriched in heavier isotopes relative to soil NO−3 if the NO−3 has experienced reduction during the N assimilation of mycorrhizae associated with the roots. Mycorrhizal fungi have substantial NO−3 reduction capacity (Ho and Trappe, 1975), but the fungal NR is present only in the presence of NO−3 and absence of NH+4 (Cove, 1966). So far, the isotopic effect of NO−3 acquisition through mycorrhizae on tissue NO−3 in natural plants has not been estimated or differentiated. Pate et al. (1993) demonstrated that the bulk δ15N differences between non-mycorrhizal and mycorrhizal species (with significant NO−3 storage and NRA) reflected the utilization of different N sources. There appears to be little or no isotopic discrimination within the plant during or subsequent to uptake of NO−3. Mycorrhizal fungi are expected to show higher bulk δ15N than available N sources [potentially including NO−3, NH+4, and DON (at least amino acids)] in soil and bulk N of host plants. However, the isotopic mechanism differed from that of tissue NO−3 and the isotope effect differed among mycorrhizal types (Högberg, 1997; Craine et al., 2009; Hobbie and Högberg, 2012). Högberg et al. (1999) showed that the ECM fungus had higher bulk δ15N relative to the Pinus sylvestris plant, and the fractionation against 15N was smaller when NO−3 was the source than when NH+4. It caused a marginal decrease in δ15N of the N passing from the substrate through the fungus to the host, which is explained by the small size of the fungal N pool relative to the total N of the plant, i.e., the high efficiency of transfer (Emmerton et al., 2001; Hobbie and Högberg, 2012). The significant shift in δ15N of fungal species was a function of fungal physiology; thus, it is difficult to constrain the N sources (using bulk δ15N) by mycorrhizal fungi or their plant partners in natural conditions (Emmerton et al., 2001).

The efflux of NO−3 from root to soil or the subsequent transport of NO−3 within plants is not expected to discriminate 15N as with the entry of soil NO−3 into root cells (Mariotti et al., 1982; Shearer et al., 1991). This can be attributed to that the diffusion of NO−3 through the membrane carriers of plant cells does not cause bonding breakage or consumption (Werner and Schmidt, 2002; Granger et al., 2004; Needoba et al., 2004). However, isotopic differences can occur between organs if partial NO−3 reduction occurs in roots before transportation. The transport of NR-processed NO−3 from roots to leaves might be misunderstood as isotopic fractionations of NO−3 transport or NO−3 reduction in shoots. So far, isotopic fractionations (ε = (lk/hk − 1) × 1000, where lk and hk respectively stand for the reaction rate constants for lighter and heavier isotopes) during the reduction of NO−3 by NR in leaves were reported as 15‰ for both N in spinach (Ledgard et al., 1985; Tcherkez and Farquhar, 2006) and O in wheat (Olleros-Izard, 1983) (Table 1). Direct measurement of endogenous NO−3 reduction in mosses after N deprivation showed similar values (Liu et al., 2012b) (Table 1). Although, NR isotopic fractionations have not been directly measured in roots, predictions can be made about the net enrichment of NO−3 isotopes in roots relative to those of soil NO−3 (Δroot; expressed as δroot − δsoil). These values should be either negligible if substantial NO−3 reduction did not occur (Scenario 1; Δroot = δroot − δsoil ≈ 0), or be close to the reported ε values of NRA in leaves (εNR) (0 − 27‰ Table 1) if NO−3 reduction occurred in the root (Scenario 2; Δroot = δroot − δsoil ≈ εNR > 0) (Figure 6). However, if the modification of soil NO−3 isotopes by soil microbial activities such as denitrification occurred later than root uptake, the observed isotopic values of root NO−3 can also be slightly lower than those of soil NO−3 despite reduction in roots (e.g., in the fine roots of a conifer investigated in Liu et al., 2013a). Furthermore, the variation of NO−3 isotopes with soil depth directly caused isotopic differences in initial NO−3 sources available to co-existing plants with different root depths. Therefore, considering this fact, soil reference samples should be collected corresponding to root distribution for characterizing the soil NO−3 isotopes available to specific plants.

Table 1

Isotopic effects reported for NO.

Biota	15ε / ‰	18ε / ‰	References
Eukaryotic NR enzymes (from fungus and marine diatoms)	26.6*	24.9*	Karsh et al., 2012
Moss	12.1*	14.4*	Liu et al., 2012b
Strains of prokaryotic plankton	0.4–8.6	0.9–8.1	Granger et al., 2010
Spinach and wheat	15.0*	15.0*	Olleros-Izard, 1983; Ledgard et al., 1985; Tcherkez and Farquhar, 2006
Eukaryotic algae	5.6–20.4	5.1–21.0	Granger et al., 2004
Marine phytoplankton	2.7–15.2	–	Needoba and Harrison, 2004
	4–9 (field) 2.2–6.2 (lab)	–	Needoba et al., 2003
Tomato	11.3–12.9	–	Evans et al., 1996
Leafy vegetable	14.2–18.1	–	Yoneyama and Kaneko, 1989; Yoneyama et al., 2003
Grasses	0.0–3.3	–	Mariotti et al., 1982
Pearl Millet and soybeans	0.0-9.5	–	Mariotti et al., 1980, 1982; Bergersen et al., 1988
Red clover	1.7–6.5	–	Kohl and Shearer, 1980

Figure 6

Schematic showing δ. Long and short solid lines with arrows respectively show the vectors of δ18O-NO−3 and [NO−3] variations. Dashed lines with arrows show the uptake, transportation, and translocation of NO−3 from the soil to roots and or to leaves, from atmosphere to leaves, during which isotope effects were regarded as negligible. Shaded areas (gray for roots, green for leaves) show isotopic enrichment during the mixing of different sources (the δ18O-NO−3 in plants should be distributed between the δ18O values of sources, depending on the fraction of each source) and or the occurrence of NR reduction activities (the δ18O-NO−3 in plants would be higher than the δ18O of sources but the magnitude of enrichment depends on in situ NR dynamics; presumably less than that presented in Table 1). For scenarios that occurred, leaf uptake of atmospheric NO−3 was assumed to be homogeneous. The shaded area, the spatial distance, and length of lines had no quantitative implications. S1–S12 correspond to scenarios 1–12 in the main text. Briefly, S1, no occurrence of NO−3 reduction in roots; S2, (inducible) root NO−3 reduction; S3, no NO−3 was transported from soil to leaves and leaf NO−3 was derived from the atmosphere, but no reduction occurred; S4, no NO−3 was transported from soil to leaves and leaf NO−3 was from atmosphere and (inducible) reduction occurred; S5, leaf NO−3 was taken up directly from the soil, but no reduction occurred; S6, leaf NO−3 was taken up from the soil and reduction occurred therein; S7, leaf NO−3 is completely or partially transported from the root where it has experienced reduction, but no further reduction in the leaf; S8, leaf NO−3 is completely or partially transported from the root where it has experienced reduction, and is further reduced in the leaf; S9, leaf NO−3 was from both atmosphere and soil but no reduction occurred in the leaf; S10, leaf NO−3 was from both atmosphere and soil, and reduction occurred in the leaf; S11, leaf NO−3 is a mixture of atm-NO−3 and root NO−3 but no reduction occurred; S12, leaf NO−3 is a mixture of atm-NO−3 and root NO−3, and reduction occurred in the leaf; S13, leaf NO−3 is a mixture of soil NO−3, atm-NO−3, and root NO−3, but no reduction occurred in the leaf; S14, leaf NO−3 is a mixture of soil NO−3, atm-NO−3, and root NO−3, and reduction occurred in the leaf. The δ18O differences between S13 and S11, between S12 and S14 depend on the fraction of soil NO−3 in the mixed pool of leaves.

In a closed system, isotopic enrichment occurs with the enzymatic consumption of substrate NO−3 and εNR is expressed as Δ/ln[NO−3]remaining fitted to the Rayleigh isotope fractionation model, where Δ represents the isotopic difference of remaining NO−3 from the initial NO−3 (δremaining − δinitial) (e.g., Granger et al., 2004, 2010). Isotopic enrichment also takes place for NO−3 remaining in plants after deprivation of NO−3 or N supply, because the tissue NO−3 pool is only changed by the NRA in a closed system (e.g., Liu et al., 2012b). Thus far, no experimental work has been done to explain the variability of 18εNR in and among vascular plants. In NO−3-supply studies, shoots tend to have higher δ15N values because of the allocation of root NR-processed NO−3 from roots to shoots (Kalcsits and Guy, 2013) or significantly higher 15εNR (by 3.3–6.9‰) than roots (Yoneyama and Kaneko, 1989; Evans et al., 1996; Yoneyama et al., 2001). Evidence from marine biota showed that both 15εNR and 18εNR can vary with growing conditions and that significantly different ε values exist among species (Table 1). In field conditions, NO−3 in an organ is more likely to be an open system with continuous source inputs (uptake), sinks (reduction), and outputs (translocation) (Figure 2). The uptake and allocation often occur according to the reduction ability and the distribution of NR, for example, a higher concentration and more NR are likely to exist in growing leaves (Gebauer et al., 1988; Cruz et al., 1991; Widmann et al., 1993). Passive or high accumulation as in mosses (Liu et al., 2012c) can happen in some organs such as conifer roots that are unable to reduce it (Liu et al., 2013a). Therefore, δ values of tissue NO−3 might not always follow the normal “Rayleigh type” relation, instead might increase with the increase in tissue [NO−3] or show a non-significant correlation with [NO−3] in the tissues (Liu et al., 2012c, 2013a). In fact, experimental studies have also shown the interplay of plant NO−3 uptake and reduction activity. The 15N discrimination during NO−3 assimilation in several higher plants was positively correlated with the supplied and tissue NO−3 concentrations, and negatively correlated with plant age (Kohl and Shearer, 1980; Mariotti et al., 1980, 1982; Bergersen et al., 1988; Liu et al., 2013a). Accordingly, the Rayleigh relation between NO−3 and its isotopes is not always applicable to examine εNR values and NO−3 reduction in organs of natural plants.

For some plants, NO−3 is not available in soil substrates. It can only be acquired from deposition (e.g., non-vascular plants or epiphytes). Alternatively, it is not available in deposition but can only be taken up from the soil (e.g., plants growing in arctic pristine ecosystems with negligible NO−3 deposition). In these plants, it is also feasible to diagnose leaf NO−3 reduction using Δleaf (the net enrichment of NO−3 isotopes in leaves relative to those of source NO−3) (Scenarios 3–6; Figure 6).

Scenario 3: If no NO−3 was transported from soil to leaves, and leaf NO−3 if any, was completely derived from atmosphere, but no reduction occurred, then:

Scenario 4: If no NO−3 was transported from soil to leaves and leaf NO−3 was acquired from atmosphere; and reduction occurred, then:

Scenario 5: If all leaf NO−3 was taken up directly from the soil and no reduction occurred in roots or leaves, then:

Scenario 6: If leaf NO−3 was transported completely from the soil and reduction occurred only in the leaves, then

The induction of NR by atmospheric-derived NO−3 has been shown in plants exposed to airborne N oxides (e.g., Norby et al., 1989; Wellburn, 1990). Scenarios 3–4 are expected to be true for mosses because atmospheric NO−3 has been assumed as the sole source (Liu et al., 2012a). Nevertheless, isotopic partitioning of N sources (Liu et al., 2013b) and further Δ17O analysis (Figure 5) suggests that moss NO−3, even at epilithic habitats, is actually a mixture of atmospheric NO−3 and soil-derived NO−3. Thus, it is becoming clear that mosses can acquire substantial N from substrates; and moss NO−3 is a valid atmospheric bio-monitor only for species growing on rare N-free substrates. Scenarios 5–6 demonstrated NO−3 dynamics of vascular plants in the tundra of northern Alaska, where the Δ17O of NO−3 in plants with surprisingly high [NO−3] was found as 0‰ (e.g., Polygonum bistorta). However, examining only Δleaf seems insufficient to determine NO−3 reduction location, since, isotopic enrichments of leaf NO−3 might result from root reduction activities before moving up to leaves (Scenario 7).

Scenario 7: If the leaf NO−3 is completely or partly transported from the root where it has experienced reduction, but no reduction has occurred in the leaf; then an isotope mass-balance calculation can be conducted to quantify the amount of leaf NO−3 accumulated directly from soil and indirectly from roots:

The reduction of NO−3 that has experienced reduction in roots can further increase the isotopic enrichment of leaf NO−3 relative to soil NO−3 (Scenario 8) (Figure 6). This has been demonstrated by the δ15N difference between roots and leaves in plants growing with NO−3 with known δ15N values (Yoneyama and Kaneko, 1989; Evans et al., 1996; Yoneyama et al., 2001). This NO3 reduction occurs especially in plants that are capable of reducing NO−3 in both shoots and roots (Stewart et al., 1992).

Scenario 8: If the leaf NO−3 is completely or partially transported from roots where it has experienced reduction; and if it is further reduced in the leaf. In this case, a partitioning similar to scenario 7 can be done by considering the Δleaf in the isotope mass-balance calculation:

Plant NO−3 in scenarios 1–8 was derived either from the soil or atmosphere (Figure 6). A supplemental diagnosis of NR dynamics was to examine the covariance of Δδ18O:Δδ15N ratios (Δ is the isotopic enrichment of plant NO−3 relative to source NO−3; Δ = δplant − δsource). This diagnosis helped determine whether the N–O bond breakage attributable to NO−3 reduction was the single process driving NO−3 15N and 18O enrichments. Theoretically, the dissociation of an O atom from NO−3 predicted that NO−3 isotopes would be fractionated in an O-to-N ratio of ca. 0.6 (Brown and Drury, 1967). However, the NR often had the same O-to-N isotopic imprint on substrate NO−3 in experimental studies. Consequently, the 1:1 trend was considered ubiquitous for biological NO−3 reduction (Granger et al., 2004, 2010). However, for leaves of vascular plants that acquire NO−3 from both atmosphere and soil, it is difficult to constrain leaf NO−3 reduction based only on the Δleaf (δleaf − δsource) and εNR, because the mixing of atmospheric NO−3 can raise the δ values (especially δ18O). Liu et al. (2013a) observed that the δ18O:δ15N ratios in roots of a conifer generally followed the 1:1 rule; although leaf NO−3 showed distinctly higher δ18O:δ15N ratios (2.5:1) because of the mixing of atmospheric NO−3.

As described above, the fraction of atmospheric-derived NO−3 (Fatm) in leaves can be estimated using Δ17O mass-balance calculation (Fatm = Δ17Oleaf / Δ17Oatm < 1). Thereafter, the leaf NO−3 sources and NR dynamics can be further constrained.

Scenario 9: If leaf NO−3 was absorbed from both the atmosphere and soil, but no reduction occurred in the leaf, then the fraction of atmospheric-derived NO−3 calculated using δ18O or δ15N (fatm) is expected to be similar to Fatm, as

Scenario 10: If leaf NO−3 was absorbed from both the atmosphere and soil, and reduction occurred in the leaf, then:

Scenario 11: If leaf NO−3 is a mixture of atm-NO−3 and root NO−3, but no reduction occurred, then:

Scenario 12: If leaf NO−3 is a mixture of atm-NO−3 and root NO−3; and if the reduction occurred in the leaf, then:

Scenario 13: If leaf NO−3 is a mixture of soil NO−3, atm-NO−3, and root NO−3, but no reduction occurred in the leaf, then:

Scenario 14: If leaf NO−3 is a mixture of soil NO−3, atm-NO−3, and root NO−3, and if reduction occurred in the leaf, then:

The parameters in the scenarios 9–14 (fatm, Fatm, Δroot, Δleaf) above, provide theoretical constraints on possible NO−3 sources and reduction dynamics in leaves of field plants. As explained above, δ15N values of NO−3 often overlapped for soil and atmospheric sources, although δ18O and or Δ17O can provide a clear differentiation between them (Kendall et al., 2007; Michalski, 2010). Consequently, the scenarios above are better suited to the δ18O (depicted in Figure 6) than δ15N analysis, particularly when leaf NO−3 was a mixing pool for different sources. The other solution to diagnose atmospheric NO−3 mixing and reduction is the Δ17O-δ18O correlation, which has been used to trace NO−3 sources and dynamics in aquatic environments (Tsunogai et al., 2011). Although preliminary, the Δ17O values in mosses showed clearly higher Fatm than vascular plants, especially in epilithic mosses. Although, the Δ17O in terricolous mosses and vascular leaf samples was as low as 0.0–2.2‰, even at high NO−3 concentration levels (Figure 5), suggesting a 0.0–8.8% of atmospheric contribution to leaf NO−3 pool. The NRA should be responsible for δ18O enrichment relative to the mixing values if plant-absorbed NO−3 has not been influenced by denitrification in soil. Such characterization cannot be warranted by correlation between δ15N and δ18O, or between tissue [NO−3] and isotopes (e.g., Liu et al., 2012c).

Uncertainties in tissue NO−3-isotope methods and future works

Although, the sampling time of plant materials can be controlled, diurnal and seasonal variations in tissue NO−3 and its isotopes should be verified in future works. Until now, no experimental work has directly examined NR enzymatic isotope kinetics in roots and leaves of higher plants. Moreover, it is difficult to mimic in situ NR isotope effects in field conditions. Isotope effects associated with NO−3 uptake and efflux remain unverified for roots. They were measured recently as 1–3‰ in growing cells of marine diatoms, and different O and N fractionations for both uptake and efflux were thought to cause the net 18ε:15ε of NO−3 assimilation above 1 (Karsh et al., 2014). The routes of transformation and entry of inorganic and organic NO−3 sources from the atmosphere into leaf cells and subsequent cellular actions have not been clarified, especially for non-aqueous processes. Consequently, the sources and supply rates of atmospheric NO−3 and their isotope signals should be explored further. Thus far, the Δ17O information of leaf NO−3 was sparse, and is mostly available for leaves with high NO−3 levels. It should be verified whether the atmospheric contribution is higher in low-[NO−3] leaves or not. It is promising to measure NO−3 isotopes in xylem flow and twig samples for NO−3 transportation and translocation. Results of such studies can potentially provide useful insights into intraplant NO−3 transportation and translocation, although the sampling methods of xylem flow are mostly destructive and in-twig NO−3 might be very low. For these reasons, more field works on tissue NO−3 at the organ, stand, and species levels should be done along with source isotope analysis. The scenarios proposed above provide the first conceptual constraint for both sources and NO−3 isotope effects in field plants. In conclusion, the concentration and isotopic analyses of NO−3 in plant tissues together provide new insights for elucidating plant NO−3 sources and strategies. These strategies will be valuable for exploring the communication of plant N utilization with environmental N pollution and altering ecosystem N cycles.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.