Biomonitors of atmospheric nitrogen deposition: potential uses and limitations.

Atmospheric nitrogen deposition is the third largest cause of global biodiversity loss, with rates that have more than doubled over the past century. This is especially threatening for tropical regions where the deposition may soon exceed 25 kg of N ha-1 year-1, well above the threshold for physiological damage of 12-20 kg of N ha-1 year-1, depending on plant species and nitrogenous compound. It is thus urgent to monitor these regions where the most diverse biotas occur. However, most studies have been conducted in Europe, the USA and recently in China. This review presents the case for the potential use of biological organisms to monitor nitrogen deposition, with emphasis on tropical plants. We first present an overview of atmospheric chemistry and the nitrogen metabolism of potential biomonitors, followed by a framework for monitoring nitrogen deposition based on the simultaneous use of various functional groups. In particular, the tissue nitrogen content responds to the rate of deposition, especially for mosses, whose nitrogen content increases by 1‰ per kilogram of N ha-1 year-1. The isotopic signature, δ15N, is a useful indicator of the nitrogen source, as the slightly negative values (e.g. 5‰) of plants from natural environments can become very negative (-11.2‰) in sites with agricultural and husbandry activities, but very positive (13.3‰) in urban environments with high vehicular activity. Mosses are good biomonitors for wet deposition and atmospheric epiphytes for dry deposition. In turn, the nitrogen saturation of ecosystems can be monitored with trees whose isotopic values increase with saturation. Although given ecophysiological limitations of different organisms, particular studies should be conducted in each area of interest to determine the most suitable biomonitors. Overall, biomonitors can provide an integrative approach for characterizing nitrogen deposition in regions where the deployment of automated instruments or passive monitoring is not feasible or can be complementary.

Introduction

Nitrogen is one of the essential elements for life and the most abundant in the terrestrial atmosphere, 80% of which is composed of N2 (Soderlund, 1976). Due to the high chemical stability derived from its strong triple bond, this molecule can only be divided by processes involving large quantities of energy or through the action of specialized nitrogen-fixing microorganisms (Galloway ). For this reason, in the pre-industrial age, more than 99% of the atmospheric nitrogen was unavailable for the great majority of organisms, which lack the enzyme nitrogenase required for fixing N2 (White ). However, as a result of our growing human population and its associated demand for food and energy, the biologically available nitrogen has more than doubled in the atmosphere over the last century. Agriculture, industry and the use of automobiles are the main sources of a complex of chemical species known as reactive nitrogen (Nr), originated from the splitting of N2 (Galloway ).

Such an increased deposition of atmospheric nitrogen has adverse effects on biodiversity. Indeed, this form of atmospheric pollution is considered to be the third largest threat to global biodiversity, following only changes in land use and climate (Sala ; Payne ). In particular, a deposition rate of 10 Kg of N ha−1 year−1, which has already been recorded for some ecosystems, is sufficient to cause physiological damage in plants (Fenn ; Bobbink ; Simkin ; Payne ). Global projections of nitrogen deposition are especially threatening for tropical regions, where it could exceed 25 Kg of N ha−1 year−1 during the present century (Galloway , 2008; Phoenix ).

Implementation of monitoring programs that enable evaluation of the status of this phenomenon and its effects on different ecosystems is thus necessary, especially in the tropics where the most diverse biotas occur. However, the deployment and operational costs of automated air quality monitoring networks may exceed the financial capacity of developing countries. One economical alternative is the use of passive collectors, which are effective in tracking pollution over large areas. Another potential alternative for tracking the nitrogen that enters ecosystems is the use of biomonitor organisms, whose spontaneous occurrence in sites of interest allows an integrative assessment of nitrogen deposition even with a single collection event, as could be during an exploratory field campaign, or in extensive exploration efforts such as national forest surveys. A biomonitor, ‘is an organism that contains information on the quantitative aspects of the quality of the environment’ (Markert ). The particular species to be selected in each region of interest (i) should have an ample ecological and geographic distribution, (ii) should be abundant and available throughout the year and (iii) there should be a clear relationship between the variable of interest and the response of the bioindicators (Conti and Cecchetti, 2001).

This paper presents the case for the potential utility of direct measurements of the nitrogen content and isotopic signature of plant tissue for characterizing nitrogen deposition. We start by showing how reactive nitrogen is formed and released to the atmosphere through anthropic activities and discuss the isotopic variation of these chemical species. Next, we explore the advantages and disadvantages of using different types of biomonitors such as mosses and vascular plants, as well as their particular responses to the different forms of nitrogen.

Reactive species of nitrogen in the atmosphere

Agriculture releases reactive nitrogen through the volatilization and leaching of nitrogenated fertilizers (Fig. 1; Cameron ). In turn, husbandry contributes to such reactive nitrogen through volatilized ammonia gas (NH3; Fowler ). Industrial activity and motor vehicles also release reactive nitrogen to the atmosphere through the combustion of fossil fuels and other processes that consume large quantities of energy, which break the triple bond of N2 and form nitrogen oxides (NOx, i.e. NO and NO2; Fig. 1; Galloway ).

Figure 1:

Sources and sinks of nitrogenous emissions. Isotopic composition of different biomonitors including trees, atmospheric plants, lithophytic mosses, grasses and the soil. Green lines and values indicate the isotopic composition of plants from natural environments and ensuing fluxes. Red indicates fluxes derived from exposure to NOx emissions from urban environments, while black NH3 from urban environments. Blue tracks the nitrogenous emissions from agriculture and husbandry.

Such nitrogen oxides and ammonia emitted to the atmosphere are subject to different chemical reactions that lead, for example, to the formation of water-dissolved compounds and gases (NO3−, NH4+, HNO3), and aerosols [(NH4)2SO4 and NH4NO3] (Aneja, 2001). These compounds are subsequently transferred to the surface of the earth either as dry deposition, in which the atmospheric gases or aerosols or deposit by gravity, or as wet deposition, in which the nitrogen ions are deposited in fog, snow or precipitation (Fig. 1; Anderson and Downing, 2006; Decina ).

Isotopic composition of atmospheric reactive nitrogen

The isotopic values of reactive nitrogen in the atmosphere have a direct relationship with the source of emission (Box 1). For instance, biogenic emissions of the soil have very negative δ15N values between −50‰ and −20‰ (Felix ; Felix and Elliott, 2014). Such an ample range of values for gaseous nitrogen species leads to differences in the δ15N of the nitrogenous compounds that dissolve in atmospheric water. In particular, the isotopic values of NH3 from volatilization of ammonia in the soil and animal wastes, tend to be low, as negative as −40‰ (Freyer, 1978, 1991; Kendall ; Felix , 2017). In turn, the δ15N for NO3− and for NH4+ range from −15‰ to 15‰, where NO3− is usually less negative than NH4+ (Hoering, 1957; Heaton, 1990; Liu ). In this respect, the negative values observed for the NH4+ are the result of the very negative NH3 reacting in the atmosphere (Felix , 2017). In addition, land use influences the δ15N of NH4+ from wet deposition are less negative in rural areas, ranging from −7‰ to 1‰, than in urban zones where they range from −16‰ to −5‰ (Ammann ; Stewart ; Xiao and Liu, 2002; Garten, 2006; Liu ; Xiao ; Harmens ; Sheng ).

Isotopes are atoms of an element that have the same number of protons and electrons, but a different number of neutrons; i.e. they are of different atomic mass. Of the known elements, there are at least 300 stable isotopes. Some elements, such as tin, have up to ten, while 21 elements are known to only have one isotope (Sulzman, 2007).

For the case of nitrogen, there are two stable isotopes. 14N is the most common and the lightest, with an abundance on Earth of 99.63%. In turn, the heaviest isotope is 15N, with a terrestrial abundance of a mere 0.37% (Rosman and Taylor, 1997; Sulzman, 2007). A stable isotope is one that remains energetically stable over time; i.e. it neither emits energy nor decays, as it occurs with radioactive isotopes that gradually mutate towards a more stable state. The better known is the radioactive isotope of carbon,14C, which is widely used in archaeological studies (Sulzman, 2007).

Differences in the isotopic composition of some materials are so small that they are reported in parts per thousand (‰), relative to an international standard. The standard used for the isotopic analyses of nitrogen is the N2 of the air. The isotopic abundance of a material is determined using the following formula: where δ15N is the isotopic proportion of the sample relative to the standard, R is the proportion between the heavy isotope and the light isotope, so that Rsample is the proportion in the sample and Rstandard is the proportion in the standard (Evans, 2001).

In chemical reactions, the differences in the δ15N of the substrate and the product result from a process known as isotopic fractionation through which the lighter isotope is favoured over its heavier counterpart. This process is described by (Δ) where δ15Ns is the isotopic composition of the substrate and δ15Np is the isotopic composition of the product (Evans, 2001). One tissue will be more enriched than another when it has a greater proportion of 15N, and depleted in the opposite case. For the case of biological reactions, accumulated fractionation is known as isotopic discrimination.

Almost all chemical processes are subject to some degree of isotopic fractionation, in consequence relative abundances of an isotope can reveal the nature of the process from which it comes. Biological organisms are not the exception, all their metabolic reactions reveal their interaction with the environment, allowing track biogeochemical processes. In this case, stable isotopes, particularly of nitrogen, become an excellent integrative tool for understand the organism-environment interactions.

With respect to dry deposition, the isotopic signature of atmospheric NOx is the result of the synergy of various factors. For example, in gasoline and diesel vehicles, both the isotopic composition (positive or negative) and the nitrogen concentration in the fuel interact with the amount of isotopic fractionation during combustion following the mixing of N2 with O2, which depends on the operation of the engine (Moore, 1977; Felix ). The burning of coal and trash can also result in an ample range of δ15N values, depending on various factors, including the isotopic composition of the material burned, temperature, pressure and time of the reaction that influence fractionation (Box 1; Moore, 1977; Liu ; Felix , 2013; Felix and Elliott, 2014). For instance, the NOx emitted by electrical energy plants (stationary source) through the combustion of coal has δ15N values between 6‰ and 13‰ in South Africa and between 5‰ and 26‰ in China (Heaton, 1990; Li and Wang, 2008). Similarly, the δ15N of the combustion of gasoline, diesel, natural gas and the incineration of trash in France yield values of 4.6–7.7‰ (Widory, 2007). In turn, studies of roadside vehicular emissions have δ15N of 3.7–15.0‰ (Moore, 1977). In contrast, the combustion of coal and fuel oil in the European country range from −7.5‰ to −5.3‰ (Widory, 2007). And the NOx from the combustion of gasoline in vehicles (mobile source) in South Africa reach isotopic values of between −13‰ and −2‰ (Widory, 2007).

A framework for biomonitoring atmospheric nitrogen deposition

The use of biomonitors can provide an integrative assessment of ecosystem responses to nitrogenous pollution with consideration of the physiological, ecological and atmospheric conditions of the region of interest (Fig. 1; Sutton ; Harmens ; Pinho ). Species composition and the physiological responses of biomonitor species following experimental manipulations have been amply utilized (Bobbink ; Ochoa-Hueso ; Lu ; Jones ). Here, we propose that a better approach to biomonitoring of nitrogen deposition is the determination of total nitrogen content and δ15N from plant tissue, which can help characterize both the rate of deposition and the source of the nitrogenous pollution (Sutton ; White , Díaz-Álvarez and de la Barrera, 2017). Indeed, various biogeochemical and physiological processes, as well as the determination of nitrogen sources, have been studied through measurements of the isotopic values of soil and plants, including for trees, herbaceous plants, mosses and vascular epiphytes (Emmett ; Stewart ; Wang and Pataki, 2011; Craine ; Díaz-Álvarez ; Felix ). In this case, the simultaneous consideration of an ensemble of biomonitors of different functional groups is necessary.

The total nitrogen content indicates the rate of nitrogen deposition

The total nitrogen content of biomonitors can help estimate the rate of atmospheric deposition in an ecosystem. In this case, epiphytic and litophytic mosses are the best potential biomonitors because their tissue nitrogen content is determined by the prevailing atmospheric deposition. Mosses growing on the forest floor are also suitable biomonitors but to a lesser extent, given that the soil can contribute up to 37% of their tissue nitrogen content (Liu ). Estimation of atmospheric deposition is thus possible from the nitrogen content of tissues, which increases by ca. 1% (dry weight) for each 10 Kg N ha−1 year−1 of deposition (Pitcairn ; Liu ). This can be observed in natural areas of Europe, where the nitrogen content of mosses ranges between 0.5% and 0.7% and can double in polluted sites (Harmens ). However, the nitrogen content of mosses only increases linearly up to a threshold of 20 Kg N ha−1 year−1, after which it decreases progressively (Pitcairn ; Shi ). Moreover, when the main form of nitrogen in deposition is NH4+ such a saturation is reached when this ion exceeds only 12 Kg N ha−1 year−1 (Pitcairn ; Wiedermann ). For instance, the nitrogen content of mosses decreases along pollution gradients in China, from 3.0% to 0.9% in urban areas and from 2.3% to 1.6% as pollution increases in rural areas (Liu ,b; Xiao ). Given that the inherent nitrogen content of mosses varies amply among species, ranging from 0.1% to 0.5% for different species of pleurocarpus mosses (Pitcairn ; Wiedermann ; Harmens ), it is important to determine dose-response curves for the particular candidate biomonitors in each region of interest.

An important environmental factor that influences the relationship between nitrogen content of the mosses is precipitation. Indeed, the nitrogen content is better correlated with the rate of nitrogen deposition when the annual precipitation is above 1000 mm (Zechmeister ). The type of atmospheric deposition (wet or dry) also influences the nitrogen content of mosses. While wet deposition can cause a 0.01% increase in nitrogen content, dry deposition can lead to an increase of nitrogen content between 0.04% and 0.07% for each 1 Kg N ha−1 year−1, reaching up to 4% in sites with high rates of dry deposition of ammonia, but just up to 1.6% in sites with wet deposition (Hicks ; Solga ; Pitcairn ; Liu ; Harmens ).

Vascular plants can also be utilized as biomonitors of the rate of nitrogen deposition, although care must be taken in their consideration as their responses are not linear. For example, the nitrogen content of the epiphytic orchid Laelia speciosa (Kunth) Schltr., 1914, amounts to 1.2% (dry mass) under a deposition of 10 Kg N ha−1 year−1, but 80 Kg N ha−1 year−1 are required to double the nitrogen content (Díaz-Álvarez ). This response has also been observed for seedlings of the tree species Cryptomeria japonica (Thunb. Ex L.f) and Pinus densiflora (Siebold & Zucc) and for adult individuals of Pinus resinosa Aiton. and Schima superba (Reinw. ex Blume) (Nakaji ; Zhang ). In this respect, an increased nitrogen availability often leads to the development of new tissue in vascular plants, rather than to increased levels in the existing cells, thus diluting what otherwise could amount to luxury nitrogen (Taiz and Zeiger, 2002).

Vascular plants can be an excellent complement to mosses for biomonitoring nitrogen deposition. Vascular plants prevail in environments that can be extreme for mosses to prosper, such is the case for urban heat island and arid regions. Additionally, given that vascular plants conform most of the plant cover, they are ideal for using other technologies such as remote sensing which can provide information about biomass and chlorophyll content variations as a result of alterations on atmospheric deposition (Schmidtlein ).

The isotopic composition discerns among natural, agricultural and urban nitrogen sources

The δ15N of plants depends on multiple factors, including mycorrhizal associations, form of nitrogen used, soil depth accessed, but most importantly atmospheric sources (Fig. 1; Table 1). Indeed, epiphytic and litophytic plants growing in natural sites without exposure to nitrogenous pollution have δ15N that are negative but very close to zero (Wania ). In contrast, volatilization and leaching from agricultural and husbandry activities alters the isotopic composition of the vegetation, making it very negative (Craine ).

Table 1:

Isotopic values for different plants from contrasting environments

Life form	Species	Rural δ15N	Urban δ15N	Reference
Moss	Braunia sp.	−3.4‰	3.3‰	Díaz-Álvarez et al. (2016)
Mosses	8 species	−12‰	6.0‰	Pearson et al. (2000)
Mosses	4 species	−1.4‰	−12.5‰	Liu et al. (2008a)
Moss	Haplocladium microphylum (Hedw)	−1.3‰	−6.5‰	Liu et al. (2008b)
Mosses	4 species	−7.9‰	−3.9‰	Liu et al. (2012a)
Annual C3 grasses	4 species	−4.2‰	13.3‰	Wang and Pataki (2009)
Herb	Impatiens sp.	−1.2‰	−6.1‰	Stewart et al. (2002)
Herb	Calluna vulgaris (L.)	−8.6‰	0.2‰	Power and Collins (2010)
Vascular epiphytes	8 species	−3.0‰	−10.9‰	Stewart et al. (2002)
Epiphytic bromeliad	Tillandsia recurvata	−6.0‰	3.0‰	Zambrano et al. (2009)
Epiphytic bromeliad	Tillandsia usneoides	−11.2‰	−2.2‰	Felix et al. (2016)
Epiphytic orchid	Laelia speciosa	−3.1‰	5.6‰	Díaz-Álvarez et al. (2016)
Tree	Eriotheca sp.	−1.6‰	−5.1‰	Stewart et al. (2002)
Tree	Picea abies	−3.0‰	2.0‰	Ammann et al. (1999)

In urban environments the isotopic composition of plants can be positive or negative, depending on the dominant species of reactive nitrogen in the atmosphere (Fig. 1). For instance, in cities where the predominant nitrogen species are gaseous NH3 and rain bound NH4+, the δ15N tend to be very negative (Xiao ; Liu ; Felix , 2017). This has been documented for urban mosses in China (Liu ,b,c, 2012a,b, 2013b; Xiao ) and for urban plants in the vicinity of a fertilizer factory in Brazil, whose δ15N reaches −41‰ (Stewart ; Heaton ).

In contrast, the isotopic signature of urban plants from various functional types is positive when NOx is the main source of nitrogen (Fig. 1). This has been documented for different mosses, including Bryum argenteum (Hedw) and Grimmia pulvinata (Hedw) in London and Braunia sp. and Grimmia sp. in Mexico City (Pearson ; Díaz-Álvarez and de la Barrera, 2017). Such positive values of δ15N have also been measured for grasses in the megalopolis of Los Angeles (Wang and Pataki, 2009). The vicinity of roads, where NOx from motor vehicles are emitted, can also determine the isotopic signature in otherwise natural environments, as positive δ15N have been measured for the needles of the conifers Picea abies (L.) H. Karst. from Norwegian forests (Ammann ) and Pinus edulis (Engelm) within the Grand Canyon National Park in the USA (Kenkel ). A similar response to NOx from motor vehicles has been documented for vascular epiphytes from west-central Mexico such as the orchid Laelia speciosa and the bromeliad Tillandsia recurvata (L.) (Díaz-Álvarez ; Díaz-Álvarez and de la Barrera, 2017).

A group with special potential for biomonitoring nitrogenous pollution in tropical regions is the so called atmospheric plants, a group that includes those epiphytes and lithophytes whose nutrition relies almost exclusive on deposited nutrients. Indeed, given their cosmopolitan distribution atmospheric mosses are widely utilized biomonitors (Markert ). Moreover, they are particularly adequate for tropical regions, where they reach their maximum diversity (Cárdenas and Delgadillo, 2009). However, mosses depend on the availability of water for sustaining metabolic activity, thus their monitoring potential is limited to the rainy season. In contrast, succulent epiphytes, especially those with CAM photosynthesis, can be metabolically active throughout the year, thus providing a continuous record of atmospheric deposition regardless of seasonal weather variations (Ammann ; Andrade ; Zotz ). Orchids, for instance, can be found in multiple ecosystems throughout the tropics from sea level up to the subalpine forest above 3500 m (Ernshaw ). Atmospheric plants thus, allow a relatively accurate determination of both the source and the magnitude of atmospheric deposition with a very low or null isotopic discrimination given a direct water flux into the cells of mosses (Liu ). In this case, nitrogen is subject to foliar uptake, either by direct influx of gaseous or aqueous nitrogen, i.e. NO, NO2, NH3 and HNO3, directly from the atmosphere during gas exchange or when nitrogen particles are deposited on the plant and dissolved in rain or fog allowing the absorption of the ions NO3− and NH4+ (Hietz ; Vallano and Sparks, 2008; Padgett ).

Although, atmospheric plants can pick up the isotopic signal of atmospheric deposition, care must be taken when, developing atmospheric biomonitors given the occasional presence of functional roots can obscure the isotopic signal measured from plant tissues (Hietz ; Reyes-García and Griffiths, 2009; Liu ). Indeed, epiphytic plants that root in the canopy soil tend to be enriched in 15N compared with those that grow on thinner branches, where no substrate accumulation occurs, because the decomposition of the accumulated organic matter produces nitrogenous compounds with δ15N close to zero (Wania ). Such a canopy soil originated from debris of the phorophyte is depleted in 15N relative to the forest soil which tend to accumulate 15N as the volatilization and biological uptake of the lighter isotope is favoured (Wania ; Liu ; Craine ).

Trees indicate ecosystem nitrogen saturation

The δ15N of trees is a good indicator of the state of saturation of atmospheric nitrogen in an ecosystem. The leaves and roots of the trees of N-saturated ecosystems tend to have positive δ15N, because saturation increases soil nitrification, a process that involves high rates of isotopic fractionation (Fig. 1; Box 1). In general, plants of ecosystems exposed to low rates of atmospheric deposition tend to present δ15N that are negative but close to zero (Craine ). However, saturation leads to increased rates of nitrate leaching, which in turn causes saturated soils to become enriched with 15N, thus their δ15N can become positive. Saturation also makes the relation between foliar δ15N and nitrification closer than that between foliar δ15N and the δ15N of the nitrogen deposition (Ollinger ; Pardo ; Emmett, 2007). The opposite occurs for translocated nitrogen as a series of isotopic fractionations occurs as it moves from the roots to the branches to the leaves, because a series of enzymes such as nitrate reductase, nitrite reductase and glutamine synthetase are involved in nitrogen transformation, and each one has its own amount of discrimination (Evans, 2001).

Associations with mycorrhizal fungi also influence the δ15N of the plants, and trees in particular, having the potential to alter both the nitrogen relations of the plants and the isotopic signature of the assimilated nitrogen (Craine , 2015). Under natural conditions (lower rates of atmospheric deposition), mycorrhizae supply their hosts with nitrogen that is depleted in 15N (Emmett ). However, saturation can induce species turnover within the mycorrhizal community, from species with high amounts of isotopic discrimination against 15N to species with low discrimination, contributing to the isotopic enrichment of the plants and the homogenization of the isotopic signature of the ecosystem (Emmett ; Craine ; Sheng ).

Metabolic limitations of biomonitors

Biomonitors can become useful tools for detecting nitrogenous pollution over wide areas of terrestrial ecosystems. However, organismal responses are constrained by enzymatic processes. For brevity, this discussion is restricted to the metabolic limitations of mosses, which assimilate NH4+ to a greater extent when supplied simultaneously with NO3−. Likewise, these organisms preferentially assimilate organic compounds such as amino acids. For example, under simultaneous application of glycine with NH4+ and NO3−, assimilation of this amino acid is up to two times greater than that of the nitrate (Wanek and Pörtl, 2008; Wiedermann ). The main reason for this is the high energetic cost of assimilation of NO3−, which requires two consecutive reactions. In the first, NO3− is reduced to NO2− by the enzyme nitrate reductase, consuming two electrons in the process. In the second, NO2− is reduced to NH4+ by nitrite reductase, using six electrons (Heldt and Piechulla, 2011).

Nitrate reductase can be inhibited by assimilation of NH4+ from atmospheric deposition when the ratio between NH4+ and NO3− is high (Liu ). Furthermore, high rates of atmospheric deposition can reduce or even completely inhibit nitrate reductase activity, whether it is due to the strong relationship between NH4+ and NO3−, or to the increased concentration of NO3− in the deposition of nitrogen. Indeed, while certain concentrations of NO3− are necessary to stimulate nitrate reductase synthesis and activity, an excessive amount of the ion exerts a negative feedback on the enzyme (Heldt and Piechulla, 2011). For this reason, when atmospheric deposition reaches 10 Kg N ha−1 year−1, a significant reduction is observed in the assimilation of NO3− and, on exceeding 30 Kg N ha−1 year−1, the nitrate reductase in the mosses is totally suppressed (Gordon ; Forsum ; Liu ,b). High concentrations of atmospheric NOx (greater than 63 nL L−1) cause suppression of nitrate reductase in mosses of different anthropic environments. Exposure to NO causes nitrate reductase activity to decrease within 24 h, while exposure to NO2 causes such an activity reduction over 21 days leading to the complete loss of inducibility of nitrate reductase even when NO3− is available (Morgan ; Forsum ; Liu ,b).

Reduced assimilation of nitrate forces the mosses to assimilate other nitrogenated compounds in the atmospheric deposition, the different isotopic values of which are presented in Table 1. As a consequence, inhibition of nitrate reductase can cause variation in the isotopic values of moss tissues and can make determination of the source of the nitrogen observed in the tissue differ from the true source by up to 21% (Liu ,b).

Thus, inhibition of nitrate reductase can cause a discrepancy between the nitrogen content of the mosses and the rate of atmospheric deposition on the site they inhabit. This can occur because nitrate that is deposited on the mosses can be partially assimilated or may not be assimilated at all. This will depend on the degree of inhibition of nitrate reductase. Consequently, part of the deposition (which contains the nitrate) will not be accurately recorded. In this case, estimation of atmospheric deposition could be more accurate in mosses when the ratio between NH4+ and NO3− is higher than in deposition with low NH4+ and NO3− ratios (Liu ). It has been observed that the nitrogen content of mosses is lower under wet than under dry deposition (Pitcairn ; Liu ,b). Because mosses lack an epidermal cuticle, the inhibition of nitrate reductase may contribute to the leaching of a fraction of the deposited nitrate instead of being stored in the tissues of these organisms. In contrast, the leaching of unassimilated nitrogen during excessive wet deposition is greatly prevented by the cuticle for vascular plants (Pitcairn ; Liu ,b).

Monitoring nitrogen deposition by means of different organisms can be a useful tool for estimating the rate of nitrogen deposition in many regions. However, caution must be taken because the inhibition of the nitrate reductase above a species-specific threshold can lead to underestimations of actual deposition rate.

Perspectives

The nitrogen content and isotopic values of biomonitors can be suitable to inform environmental policy design for reducing the emissions of nitrogenous compounds, thus contributing to the mitigation of the adverse effects that atmospheric nitrogen deposition may have on priority ecosystems. Mosses can be especially useful because their nitrogen content responds directly to the rate of atmospheric deposition and their isotopic signature to the source. This is true up to certain deposition rate above which N accumulation decreases as a result of nitrate reductase inhibition. With the simultaneous use of different types of biomonitors, a multidimensional evaluation can be carried out regarding the state of ecosystems in the tropics. This could involve biomonitors that indicate the state of saturation, such as trees and shrubs, and those that indicate the source, such as vascular epiphytes with which it is possible to estimate the rate of atmospheric deposition using mosses. Further research should consider the ‘calibration’ and development of potential biomonitors suitable for each region of interest. For the case of tropical regions, atmospheric plants may prove particularly adequate. In any case, caution must be exercised given that biomonitors cannot provide the exact magnitude of atmospheric deposition, but a semiquantiative approximation, including characterizing the nitrogen source. In this case, the simultaneous use of an ensemble of various species can be of great utility in identifying areas subject to pollution by atmospheric nitrogen, especially in regions where nitrogen saturation has not occurred.