Human activity has more than doubled the quantity of nitrogen (N) fixed in terrestrial ecosystems, due to industrial nitrogen fixation, the mobilization and inadvertent fixation of nitrogen during fossil fuel combustion, and cultivation of nitrogen fixing crops. The mobility of fixed nitrogen within and between terrestrial ecosystems has also been enhanced as a consequence of land clearing, biomass burning, wetland drainage, and other processes.  Enhanced inputs of nitrogen are spread unevenly over Earth, being concentrated in and near areas where intensive agriculture is widespread, and in industrial regions. In regions where nitrogen inputs have been altered substantially, ecosystem properties as fundamental and as disparate as C storage, trace gas exchange, cation leaching, biodiversity, and estuarine eutrophication have been affected.

Until recently, enhanced inputs of nitrogen were concentrated in economically developed regions of the temperate zone and research on the consequences of enhanced nitrogen inputs was even more strongly focused on temperate ecosystems. Temperate and boreal forest and grassland ecosystems are often limited by the supply of fixed nitrogen, in the sense that additions of fixed nitrogen increase plant growth and C storage therein. The term ‘nitrogen saturation’ has been applied to changes in ecosystem functioning that occur as nitrogen additions relieve nitrogen limitation to biological activity. In general, the paradigm of nitrogen saturation predicts that nitrogen-limited forests initially retain anthropogenic nitrogen by using it for vegetation and microbial growth and accumulation in biomass; nitrogen can also be accumulated in soil organic matter.

At some point, however, inputs of nitrogen exceed the biological demands for nitrogen within the ecosystem, and the system begins to lose its ability to retain nitrogen. As this capacity to retain nitrogen is exceeded, excess nitrogen can move from the system via solution losses and gas fluxes. The nitrogen saturation model has been evaluated extensively through observations on nitrogen deposition gradients and through experimental studies in temperate forests. In general, such studies report a delay between the commencement of nitrogen additions and the observation of increased losses. Added nitrogen is more or less quantitatively retained within systems for a number of years, but with continued additions, nitrogen dynamics within systems are altered progressively and the ability of forests to retain added nitrogen is reduced.

The concentration of research on the consequences of nitrogen deposition into north-temperate forests made sense until recently – most nitrogen deposition occurred there. Now, however, 40% of global applications of industrial nitrogen fertilizer take place in the tropics and subtropics, and over 2/3 areexpected to occur in now-developing regions by 2010. Similarly, fossil fuel combustion is increasing dramatically in less economically developed parts of Earth, including much of the tropical and subtropical region. Galloway et al. (1994) estimate that by 2020, nearly 2/3 of Earth’s energy-related nitrogen inputs will take place in the tropics and subtropics.

Finally, nitrogen emissions associated with biomass burning are already heavily concentrated in the tropics, and likely will remain so for decades. We suggest that most tropical forests function quite differently from most temperate forests with regard to nitrogen cycling, and that the effects of anthropo-genic nitrogen inputs on tropical ecosystem processes may also differ. There is substantial although mostly indirect evidence that nitrogen supply does not limit plant production in the majority of tropical forests, while most temperate systems are nitrogen-limited. Thus, additions of nitrogen may have little direct effect on plant production and carbon storage, but may substantially affect the rate and timing of nitrogen losses. Also, many tropical forests soils are highly acidic; additions of anthropogenic nitrogen may increase that acidity, leading to increased losses of cations and decreased availability of phosphorus and other limiting nutrients, ultimately reducing plant production and other ecosystem functions. In this paper, we assess the likely consequences of enhanced nitrogen inputs in tropical forests. Before enhanced deposition becomes more widespread, it is important to identify the direct and indirect effects of nitrogen additions to tropical forests, and to determine their probable consequences to aquatic systems and the atmosphere locally, regionally, and globally. In this analysis, we identify unknowns and uncertainties that constrain our ability to predict the consequences of this already significant – and rapidly growing component of environmental change.

Direct Effects of Altered Nitrogen
on Tropical Forest Processes

Whereas NPP and NEP in temperate ecosystems are largely nitrogen-limited, a range of evidence suggests that biological activity in many tropical forests, especially those on highly weathered soils, is not limited by nitrogen but rather by some other nutrient (e.g., P, Ca) or by other resources. The one forest fertilization experiment we are aware of that was carried out on a highly weathered tropical soil demonstrated limitation by P and not nitrogen. Studies that have used the growth response of roots to split additions of nutrients have yielded responses to nutrients other than nitrogen, except on white sand soils and montane forest soils. Unlike the situation in temperate ecosystems, where anthropogenic nitrogen stimulates plant growth, probably creating a significant sink for excess atmospheric CO₂.

70 additions of anthropogenic nitrogen are likely to have little direct effect on carbon uptake in most tropical forests. Nitrogen biogeochemical processes in soils strongly re?ect the greater availability of nitrogen in many tropical soils. Studies of nitrogen cycling and nitrogen trace gas emissions across a range of tropical ecosystems suggest that nitrogen gas fluxes (as a proportion of nitrogen mineralization) are greater in tropical than in temperate systems, and that quantitatively, nitrogen trace gas fluxes are on average much greater in tropical than in temperate systems. While Vitousek and Matson reported a range in microbial uptake of applied 15N depending on tropical soil type, microbial immobilization of 15N generally was much less than in temperate systems where similar measurements have been made. Thus, it seems likely that retention of anthropogenic nitrogen will be much less in most tropical systems than in most temperate systems, due to reduced microbial uptake and retention in soil organic matter as well as reduced uptake and retention in plant biomass.

The rather limited data from upland tropical forests also suggests that most mineralized nitrogen is quickly nitri?ed. In contrast, in many temperate forests, not all nitrogen that is mineralized to ammonium is oxidized to nitrate. Indeed, one of the hypothesized responses of temperate soils to long-term anthropogenic nitrogen inputs is a gradual shift in the nitrogen economy of soils, from ammonium-dominated to nitrate-dominated systems. Upland tropical soils are likely to display little delay in nitri?cation in response to increasing nitrogen additions.

Indirect Effects of Anthropogenic Nitrogen on Ecosystem

The biogeochemical consequences of enhanced nitrogen inputs to terrestrial ecosystems can range well beyond direct alterations of nitrogen cycling and plant response. Numerous studies in natural and agricultural systems of the temperate zone have shown that both atmospheric and fertilizer inputs of nitrogen can lead to soil acidification, depletion of base cations, and mobilization of potentially toxic aluminum (Al) ions. Moreover, many tropical soils have developed in place over a very long time, without rejuvenation by glaciation or loess deposition. Accordingly, these soils can be highly weathered, depleted in primary minerals, poorly buffered, and dominated by variable charge clays (Uehara& Gillman 1981; Sanchez & Logan 1992). These systems are therefore (1) poorly buffered against additional sources of acidity, and (2) potentially quite sensitive to base cation depletion. As nitrogen inputs increase in humid tropical regions, changes in soil acidity, base cation supply, P availability, Al mobility,and carbon storage are all possible, but the rates, consequences, and some-times even the direction of such changes are likely to be markedly different in many tropical sites than what has been observed in the temperate zone.

Soil Acidification

Elevated nitrogen inputs can lead to soil acidification via several pathways, with the overall change depending on the properties of the ecosystem, the form in which nitrogen is added, and the anion or cation associated with the added nitrogen. Both fertilizer and atmospheric nitrogen inputs can lead to soil acidification; our focus here will be upon atmospheric inputs only. Added nitrogen can be acidifying regardless of the inorganic form by which it is added, but the net change in soil acidity due to inputs depends on plant uptake. For example, nitric acid inputs do not necessarily lead to soil acidification: if the nitrate ion is taken up and retained by plants, the proton is neutralized by release of a hydroxyl ion. However, if NO−3 remains mobile and leaches out of the system, the proton’s greater affinity for cation exchange sites in the soil will displace a base cation (potassium, calcium, magnesium, sodium) which will balance the charge of the leaching NO−3 ion. The addition of one molecule of nitric acid would then cause a net increase of one proton in the soil and the net loss of one base cation. Addition of nitrogen as ammonium (NHC4) can produce the greatest per molecule effects. If the NHC4 ion is taken up, one proton is released. (When soil NHC4 is derived from decomposition of organic matter, this released proton is not an increase in acidity as it balances the consumption of a proton during conversion of NH3 to NHC4.) If the added NHC4 ion is nitri?ed, two protons are released. If the resulting NO−3 ion is then taken up by a plant, the net change is still an increase of one proton; if the NO−3 is leached, two protons are added and one base cation is lost.

In temperate ecosystems, addition of excess nitrogen from the atmosphere has led to soil acidification and base cation depletion, but strong plant demand for nitrogen slows the rate of change. As discussed above, however, moist tropical systems frequently are rich in nitrogen relative to other essential elements. Thus, the majority of excess nitrogen inputs are likely to end up as NO−3 , and much ofthis NO−3 may not be taken up by plants. Leaching of NO−3 from surface soils mobilizes a cation. As the base cation supply is progressively depleted, the leached cation will either be a proton or mobilized aluminum ion, both of which have potentially serious consequences for downstream aquatic systems.

Anion Adsorption Capacity

The fate of excess nitrogen and the rate of soil acidification in tropical systems may also be affected by the surface charge properties of many tropical soils. Soil surface charge in the vast majority of temperate zone soils is both permanent and negative, so that anion adsorption capacity is usually quite low. However, iron (Fe) and Al oxide rich tropical soils are often of variable charge, meaning that overall net charge is a function of pH and soil solution chemistry. At the low pH’s typical of these soils, net charge is frequently positive. In addition, free Fe and Al oxides create significant sources of positive charge across a large range of pH values. Thus, electrostatic adsorption of may be important incontrolling losses of excess NO−3 to aquatic systems.

Such adsorption may be especially important in deeper soil horizons. High concentrations of organic matter in surface horizons create negatively charged surfaces that make nitrate highly mobile. However, SOM decreases with depth while the density of positively charged sites increases. Matson etal. (1987) showed that significant amounts of nitrogen lost from surface horizons in a Costa Rican soil were retained as exchangeable NO−3 in soil layers below 40cm.
Soils in the lowland tropics frequently are very deep (>10 m, at times much greater, thereby creating a long hydrologic path over which NO−3 can be adsorbed. Thus ,increases in NO−3 leaching from surface soils due to greater nitrogen deposition in the tropics may not lead to similar increases in NO−3 loading of aquatic ecosystems. Prediction of the role of anion exchange in controlling NO−3 losses is further complicated by the biogeochemistry of other elements. Where excess S inputs occur in addition to increasing nitrogen, ligand exchange of sulfate ions will decrease NO−3 adsorption capacities (Zhang & Yu 1997). In addition, there is evidence that NO−3 adsorption may vary considerably with the identity of its charge-balancing cation in soil solution. For example, Wang (1987) found that NO−3 adsorption in both a Brazilian and a Chinese oxisol was15–20% greater when the accompanying cation was calcium as opposed to potassium. Given that soil acidification should result in base cation losses, NO−3 retention may vary with time as the balance of cations in soil solution shifts.

Effects on carbon storage Unlike the temperate zone, where increasing nitrogen deposition may cause at least a transient increase in carbon storage, we suggest that higher nitrogen inputs to moist tropical systems may lead to lower productivity and reduced carbon storage. This may occur via several mechanisms. First, because plant growth in many tropical systems appears limited by some combination of P and/or base cation availability, losses of base cations due to increased leaching of nitrate may cause reductions in plant growth and carbon storage. Further-more, cation exchange capacity (CEC) in many tropical soils is largely a function of soil organic matter content. Thus, increasing nitrogen inputs could induce a positive feedback in which higher losses of cations decrease growth rates and organic matter pools, which in turn reduce the soil’s capacity to retain base cations, further reducing productivity and C storage. 1 Second, decreases in productivity and carbon storage may also occur due to the effects of increasing soil acidity on phosphorus availability. Phosphorus is the element most commonly associated with nutrient limitation in the lowland tropics. This deficiency occurs for two reasons.

First, as with the base cations, the combination of extremely old soils and high weathering rates has led to severely depleted primary mineral pools of P. Second, acidic soils rich in iron and aluminum oxides react with labile inorganic P, and fix some of that P into insoluble forms. As a consequence, even tropical soils that have large total soil P pools may have scarce plant-available P. Additions of excess nitrogen to such soils may exacerbate P limitation, because rates of P fixation tend to increase with decreasing soil pH2. In many tropical systems, lowered P availability is likely to result in less C uptake and storage.

Finally, the prevalence of aluminum oxides and low pH in many tropical soils means that even slight additional decreases in pH can lead to significant increases in the release of mobile Al ions into soil solution. The solubility of Al in soils is a nonlinear function of pH, and many tropical soils have pH values well below 5, and at times below 4.

Sharp increases in soluble Al occur at soil pH’s below 4, thus Al mobilization is likely to occur following a much smaller increase in nitrogen deposition in the tropics than has been seen in temperate systems. Increases in soluble Al have a number of potentially deleterious effects, ranging from inhibition of plant and microbial activity to poisoning of fish and other aquatic organisms in downstream systems.