Input Output Balances and Nitrogen Limitation in Terrestrial Ecosystems

Peter Vitousek

Stanford University Stanford. California

Christopher B. Field

Carnegie Institution of

Washington Stanford. California

1. Introduction 217

2. Long- Term Nutrient limitation 218

3. A Simple Model 219

4. Pathways of N Loss 219

5. Constraints on N Fixation 222

6. Conclusions 222

References 223

Why does the availability of N often limit net primary production (NPP) and other processes in terrestrial ecosystems? For N to limit NPP in the long term, two conditions must be met: N must be lost from terrestrial ecosystems by pathways that cannot be prevented fully by N-demanding organisms, and the power of N2 fixation to add new N to N-limited ecosystems must be constrained. We utilize a simple model to explore the consequences of (a) losses by dissolved organic nitrogen, transformation dependent trace gas fluxes, and spatial/temporal variation in the supply versus demand for N, and (b) constraints on N2 fixation caused by disproportionately severe effects of P limitation, grazing, and shade intolerance on symbiotic N, fixers. The results of these analyses suggest that the pervasiveness of N limitation in terrestrial ecosystems is strongly shaped by processes that are not well understood.

1. Introduction

The biological availability of nitrogen—its pattern, dynamics, and regulation — has attracted a great deal of research for several decadcs. Why has there been such intense focus on just one of the 13 or so essential elements that higher plants obtain from soil? There arc a number of reasons that the nitrogen cycle has been and remains particularly interesting to terrestrial ecologists:

1. The supply of biologically available N demonstrably limits ecosystem properties and processes over much of the earth. It controls yield in most intensive agricultural systems and controls plant growth, net primary productivity (NPP), species composition and chemistry, and trophic structure in many managed and natural systems (e.g., Tilman, 1987; Berendse et al. 1993).

2. The global cycle of N has been altered to an astonishing degree by human activity. Humanity more than doubles the quantity of N2 fixed annually on land, and greatly increases fluxes of the N-containing trace gases from land to the atmosphere, and those of nitrate from land to aquatic systems (Galloway etal. 1995; Howarth et al, 1996; Vitousek et al., 1997a).

3. On the one hand, human activities that increase the availability of N in N-limited systems can cause net storage of C; this may be an important component of the "missing sink" for anthropogenic C02 (Schimel et al., 1995; Townsend et al., 1996). On the other hand, N limitation may be an important constraint on the ability of terrestrial ecosystems to store C in response to anthropogenic C02 enrichment (Melillo et al., 1996).

4. Anthropogenic fixed N causes or contributes to a wide range of environmental problems, from forest dieback and the loss of biological diversity on land to acidification and unhealthy concentrations of nitrate in streamwater and groundwater to eutrophication of estuaries and ocean margins to increasing concentrations of the reactive gas nitric oxide regionally and of the greenhouse gas nitrous oxide globally (Schulze, 1989; Nixon et al., 1996; Vitousek et al, 1997a; Aber et al, 1998).

5. N limitation is economically important; humanity spends tens of billions of dollars annually on N fertilizer and its application.

6. There is a fascinating intellectual puzzle concerning N limitation. Given the ubiquitous distribution of N2-fixing organisms that can draw upon the essentially unlimited supply of atmospheric N2, how can N limitation be anything more than a marginal or transient phenomenon (Vitousek and Howarth, 1991)?

The nature of this puzzle is perhaps best appreciated by examining lake ecosystems. Twenty-five years ago, there was an intense controversy in the United States and Canada concerning nutrient limitation — specifically, concerning what controls the anthropogenic eutrophication of lakes. C, N, and P all had their proponents, until experimental studies with whole lakes demonstrated unambiguously that while the supply of C and N may affect photosynthesis and other processes in lakes, the longer-term accumulation of algal biomass is driven by P enrichment. P supply is controlling in the long term because lake surface water is an open system with respect to C and N; the concentration of CO, in surface water can be drawn down by biological uptake, but then more will enter in by diffusion from the atmosphere. Similarly, the supply of fixed N can be drawn down by biological uptake, but then N,-fixing cyanobacteria will have a substantial competetive advantage over other phytoplankton, dominate the producer community, and bring the quantity of fixed N more or less into equilibrium with that of P, at the N:P ratio required by phytoplankton (Schindler, 1977). There is still excellent work being done on the interactions between C, N, and P in aquatic ecosystems (e.g., Elser et al., 1996), but the fact that eutrophication generally is controlled by P supply, and the reasons for that control, are not in dispute.

Why aren't terrestrial ecosystems more lake-like? Their N cycle is open, at least potentially, so why is NPP in many terrestrial systems N-limited? Before addressing this question directly, we should note several points. First, the question is explicitly comparative across elements. Why is N more important than P, Ca, K, or B in controlling NPP, net ecosystem production (NEP), and other processes in many terrestrial ecosystems?

Second, by saying that NPP is N-limited, we do not assert that only N is limiting; multiple resource limitation is the rule in ecosytems (Bloom et al., 1985; Field et al., 1992). Biomass accumulation in lakes is limited by light as well as P; NPP in terrestrial systems can be limited by water, C02, light, and one or more soil-derived nutrients simultaneously. However, it would be surprising to find that N, P, Ca, Mg, K, B, and all other soil-derived elements were equally limiting in any ecosystem; in practice the supply of one or two elements is controlling at any given time. Soil-derived elements can neither be obtained independently nor readily traded off for each other (Rastetter and Shaver, 1992; Gleeson and Tilman, 1992).

Finally, our analysis will be focused on inputs and outputs of elements at the ecosystem level and on their controls. In the short term, nutrients can limit NPP or other processes when organisms' demands for an element exceed the supply of that element; for N and P, that generally means that potential uptake exceeds mineralization. A particular element may be limiting because it cycles more slowly than another; for example, biochemical mineraliza tion of P by extracellular enzymes can allow P to cycle more rapidly than N, driving ecosystems toward limitation by N (McGill and Cole, 1981). However, in the longer term (centuries to millennia), the balance between inputs to and outputs from ecosystems determines the quantity of elements that can cycle within ecosystems.

2. Long-Term Nutrient Limitation

Ecosystems are open systems, with the potential for inputs and outputs of all biologically essential elements. In the long term, for any element to limit NPP or other ecosystem processes, one essential condition must be met:

1. The element must be lost from the system by some pathway^) in addition to the loss of "excess available nutrients" (defined below); these additional losses must be large enough to balance element inputs at a point where the supply of that element (within the system) remains limiting to NPP.

For N, a second condition also is essential:

2. Some process(es) must constrain rates of biological N2 fixation to the extent that N2 fixers cannot respond to N deficiency sufficiently to eliminate it.

For the first condition, many ecosystem models, conceptual and others, assume that losses of elements occur from a pool of excess available nutrients—nutrients that remain in the soil when plants and microbes have taken up all the nutrients that they can use— and that this pool (of a particular nutrient) is vanishingly small when that nutrient limits NPP or other ecosystem processes (Vitousek and Reiners, 1975). Simple models of N saturation and its consequences are based on this approach (Agren and Bosatta, 1988; Aber, 1992). If this conceptual model were accurate, however, and if nutrient supply and demand were relatively constant in space and time, then no nutrient could remain limiting indefinitely. In any real ecosystem, losses of a limiting nutrient would be near zero, inputs from outside the system would accumulate, and eventually the pool of that limiting nutrient within the system would increase to the point where it no longer limited NPP (or did so only marginally) (Hedin et al, 1995; Vitousek et al, 1998). Nutrient limitation can be sustained in the long run by loss pathways that are independent of excess available nutrients, or where spatial or temporal variation allows losses that are not wholly preventable. What are these pathways? (Note that this necessary condition implicitly includes rates of nutrient input; alternative pathways of element loss may be sufficient to sustain nutrient limitation when inputs are low, but not in the face of high rates of input.)

While the first condition applies to all essential elements, the second is specific to N. Biological N2 fixation is capable of adding tens to hundreds of kg ha- 1 year- 1 to ecosystems (Sprent and Sprent, 1990), more than enough to meet plant and microbial demand for N in a short time, and to overwhelm the capacity of alternative N loss pathways, and so rapidly offset N limitation. It is this potential to respond to deficiency with large, biologically-controlled inputs that makes the widespread nature of N limitation such a puzzle. Biological processes within ecosystems can affect inputs of other elements—for example, plant and microbial activity can enhance rates of rock weathering (Davis et al, 1985; Cochran and Berner, 1997)—but not in a regulatory way, not with the ability to enhance inputs of a particular element when it is deficient within the system. What processes constrain N, fixation in N-limited ecosystems, and so sustain N limitation?

With these conditions and questions in mind, why is P more often limiting in lakes than on land, at least in the temperate zone? P limits lake productivity because (a) unlike C and N, there are no mechanisms that can increase inputs of P when it is in short supply, as discussed above; (b) P is relatively immobile within and through terrestrial ecosystems, so inputs of P to lakes are small; and (c) lakes have an uncontrollable loss of P, in the sinking of particulate organic matter out of the euphotic zone.

In contrast, terrestrial ecosystems include soils that develop from parent material containing large quantities of P (and Ca, Mg, K, and other elements). The supply of P and other elements via weathering of parent material is large relative to the requirements of organisms, for thousands to hundreds of thousands of years after unweathered parent material begins its development into soil (Walker and Syers, 1976).

Once the weathering source of P and other elements is depleted, limitation by P or another rock-derived element becomes possible (Walker and Syers, 1976, Vitousek el al, 1997b); atmospheric inputs of P in particular are very small (Newman, 1995). Accordingly, while sustained P limitation to NPP is unlikely in ecosystems of the north temperate or boreal zones, where the frequency of glaciation should suffice to maintain weathering as a source of minerals within soil, P (and base cation) limitation could be more frequent on geologically old substrates in the tropics or subtropics (Vitousek et al, 1997b, Kennedy et al, 1998, Chadwick et al, 1999).

3. A Simple Model

Why does N supply limit NPP and other ecosystem processes in many terrestrial ecosystems? In contrast to P, N is absent from most parent material (not all—see Dahlgren, 1994); it must be accumulated from the atmosphere. Nevertheless, even low inputs of N over thousands of years should more than account for the quantities of N we observe in most ecosystems (Peterjohn and Schlesinger, 1990); still less time is required where N fixers are abundant.

We evaluated alternative pathways of N loss, constraints to inputs via N, fixation, and their interactions and consequences using a simple model (Vitousek and Field, 1999). This model is not intended to represent ecosytem dynamics in detail, but it is useful in examining the logical consequences of plausible assumptions about N inputs and outputs and their controls and consequences.

In its simplest form, the model includes two types of primary producers: nonfixers and symbiotic N2 fixers. Nitrogen becomes available in the soil through N mineralization and (secondarily) atmospheric inputs; N mineralization occurs when decomposition reduces the soil C:N ratio below a threshold. In effect this gives microbes priority over plants for available N. The nonfixer is assumed to take up all available N, up to a ceiling set by light availability. If available N remains in the soil above that ceiling, it is lost from the system—implicitly by nitrate leaching or denitrification. If all of the available N is taken up by the nonfixer and light remains available, then (and only then) the N, fixer can grow and fix N,, up to the ceiling set by light availability. In effect, this gives nonfixers priority for available N and for light in proportion to available N. This assumption is too strong, in that symbiotic N fixers can make use of already-fixed N in the soil (Pate, 1986; McKey, 1994). However, it is conservative in that it tends to eliminate symbiotic N fixers from simulated ecosystems, and so downplay N fixation—and yet (as we will show) it is insufficient to maintain N limitation on NPP (Vitousek and Field, 1999).

We assume that 10% of plant C and N is lost annually as litter; accordingly, plant biomass is close to a 10-year running mean of NPP. A mass balance for N in the system is maintained, so that biologically fixed N ultimately increases the quantity and availability of N in the system.

Results of a long-term run of the model, starting with no plants, C, or N in the system, are summarized in Fig. 1. If N fixation is excluded (set to zero), then the system must depend on a low rate of atmospheric deposition to accumulate N. It takes millennia to accumulate sufficient N to the point where it scarcely limits biomass accumulation (Fig. la), but as long as N can only be lost when it is in excess, N will accumulate to this point. Allowing N2 fixation causes N to accumulate and biomass to equilibrate much more rapidly (Fig. lc), but the equilibrium N accumulation and NPP are the same with or without fixation.

4. Pathways of N Loss

We can identify three pathways that could remove N from terrestrial ecosystems, even though it limits NPP therein. These are loss of dissolved organic N (DON), loss of N trace gases by transformation-dependent pathways, and losses of N as a consequence of temporal or spatial heterogeneity in the supply versus demand for available N within ecosystems.

4.1 Dissolved organic N

Hedin et al (1995) suggested that losses of DON could represent an uncontrollable leak of fixed N from ecosystems, one that could balance the very low atmospheric N inputs in the low-input Chilean temperate forest they studied. While the controls of DON flux are not well understood, DON loss appears to be much less dependent on the N status of ecosystems than is nitrate leaching (e.g., Currie et al, 1996). Where DON loss is substantial (and inputs are small),

FIGURE 1 Biomass of a nonfixer (solid line) and a symbiotic N2 fixer (dashed line) as function of time, starting with no plant or soil C or N. (A) No N, fixation, and losses of N occur only from the pool of excess available N. (B) No N, fixation; N losses (as DON or transformation-dependent trace gases) can also occur in proportion to the quantity of N cycling in the system. (C) N2 fixation can occur; N losses only as excess available N. Note change in x-axis. (D) With N, fixation and the additional pathways of N loss. Revised from Vitousek and Field (1999).

FIGURE 1 Biomass of a nonfixer (solid line) and a symbiotic N2 fixer (dashed line) as function of time, starting with no plant or soil C or N. (A) No N, fixation, and losses of N occur only from the pool of excess available N. (B) No N, fixation; N losses (as DON or transformation-dependent trace gases) can also occur in proportion to the quantity of N cycling in the system. (C) N2 fixation can occur; N losses only as excess available N. Note change in x-axis. (D) With N, fixation and the additional pathways of N loss. Revised from Vitousek and Field (1999).

this pathway of removal of N could keep N from accumulating to the point where it no longer limits NPP (Hedin et al, 1995).

4.2 Transformation-Dependent Trace Gas Flux

In a sense similar to DON, N trace gases that are produced and lost in the course of nitrification could be regarded as leaks of potentially available N (Firestone and Davidson, 1989), whereas losses via denitrification could be more analogous to leaching of nitrate. The nitrification process is internal to the N cycle of many ecosystems, while denitrification utilizes a pool of N that can accumulate when N is available in excess (Vitousek et al., 1998).

Vitousek and Field (1999) evaluated the effects of N losses through DON and transformation-dependent trace-gas fluxes by modifying the model described above so that a constant fraction (5%) of net N mineralization is lost. With that additional loss pathway in place, and with N fixation turned off, N accumulation, NPP, and biomass in the simulated system equilibrated to much lower levels (e.g., Fig lb for biomass) than in the case where only excess N was lost. Vitousek et al. (1998) reported similar results from a simpler model; further, they showed that boosting simulated atmospheric deposition of N from 2 to 10 kg ha-1 year-1 was sufficient to overwhelm the additional loss pathways of N and to take the system to the original, non-N-limited equilibrium. Clearly, alternative loss pathways that are dependent on N transformations rather than excess available N are sufficient to cause sustained and substantial limitation by N in low-input systems. Results are similar if these additional losses are made dependent on the total quantity of soil organic N.

Given the assumptions about inputs, outputs, and their controls, these conclusions are robust despite the simplicity of the model. For N to limit NPP and biomass accumulation in the long term, there must be N losses from pathways other than excess available N, and there must be additional constraints on N fixation even stronger than our (assumed) priority of the nonfixers for fixed N and light, in the model as well as in the world. These same processes cause simulated N limitation in the more complex Century model. Century calculates several pathways of loss of N, including losses by leaching and denitrification, from the pool of available N that is left over after biological uptake; losses of N as DON (calculated as a complex function of decomposition and water flux); and losses of N as nitrification-dependent trace gas flux (calculated as a constant fraction of gross N mineralization (Metherell et al., 1993, Parton et al., 1996)). Schimel et al. (1997) used Century to show that globally the cycles of C, N, and water equilibrate with each other in the long run, with N always in relatively short supply (so that it generally limits NPP within Century). Vitousek et al. (1998) ran a tropical forest version of Century with DON and nitrification-dependent losses turned off; when N losses can only occur via excess available N, Century simulates a system with greater N pools, greater productivity, and no N limitation at equilibrium. The world according to Century is limited by N in part because (a) it includes substantial losses of N by pathways that are independent of excess available N and (b) it does not allow for substantial N fixation.

4.3 Temporal/Spatial Variation in N Supply versus Demand

This third pathway involves loss of available N when it is in temporary or local surplus, even though N supply limits NPP most of the time or over most of the area. Temporary excesses of supply over demand can occur on time scales from centuries (disturbance/regeneration cycles in forests) to seasonal or even day-today. Large-scale disturbance can cause a short-term excess of N supply over demand, leading to losses (Vitousek and Reiners, 1975; Bormann and Likens, 1979). More importantly, fire and harvest themselves cause substantial losses of N; where these are the important agents of disturbance, ecosystem N budgets are characterized by long periods of accumulation (and potentially limitation) punctuated by brief periods of large losses.

Year-to-year variations in climate can also drive temporary imbalances in N supply and demand, particularly in water-limited systems. We evaluated this process by modifying the model above (Vitousek and Field, 1999) to include water as a resource—making production, decomposition, and losses all constrained equiva-lently by water supply. We ran the model with constant but low water availability, no N2 fixation, and only excess N lost; NPP equilibrated at a lower level than in Fig. la, in direct proportion to water supply, but N did not limit production or biomass accumulation at that equilibrium value. We then introduced random year-to-year variation in precipitation. As a consequence of this variation, available N was in excess in some years (e.g., wet years following several dry years) and could be lost. However, N was in short supply in most years, to the point that it limited NPP—and this N limitation was exacerbated by N losses during the years when N was in excess. We quantified the extent of N limitations by simulating additions of N fertilizer each year. Without year-to-year variation in precipitation, added N had little effect on NPP or biomass; when N was added to a system with year-to-year variation, NPP and biomass accumulation were enhanced by 20% on average (Fig. 2).

The model used here is relatively simple, but the results make sense given our understanding of controls on decomposition and mineralization in ecosystems. Also, more complex models yield similar results. Year-to-year variations in NPP in Century are controlled by interactions between precipitation and precipitation-induced variations in N mineralization (Burke et al., 1997); these year-to-year variations can drive losses of N even when it limits NPP in most years. The Pnet model also predicts year-to-year variations in nitrate leaching from deciduous forest watersheds as a consequence of variations in precipitation; these predictions are strongly supported by watershed-level observations (Aber and Driscoll, 1997). This mechanism could help to explain observations that the N cycle appears to be more open in semiarid areas than in mesic forest ecosystems, in the sense that both inputs and outputs of N are larger relative to N pools within ecosystems (Austin and Vitousek, 1998).

5000 5200


5600 5800

6000 5000 5200

5400 5600



FIGURE 2 The effect of year-to-year variation in precipitation on N limitation to biomass accumulation in a simulated system without N, fixation or alternative pathways of N loss. The straight lines represent simulated plant biomass without year-to-year variation; simulated additions of N fertilizer (in part B) increase biomass and production by —2%. With year-to-year variation, biomass and production are lower and variable. Simulated additions of N fertilizer increase biomass and production by —25% on average, to the level of the system without year-to-year variation—demonstrating that year-to-year variation in precipitation induces N limitation. (A) No fertilization. (B) N fertilizer added.

5000 5200


5600 5800

6000 5000 5200

5400 5600




FIGURE 2 The effect of year-to-year variation in precipitation on N limitation to biomass accumulation in a simulated system without N, fixation or alternative pathways of N loss. The straight lines represent simulated plant biomass without year-to-year variation; simulated additions of N fertilizer (in part B) increase biomass and production by —2%. With year-to-year variation, biomass and production are lower and variable. Simulated additions of N fertilizer increase biomass and production by —25% on average, to the level of the system without year-to-year variation—demonstrating that year-to-year variation in precipitation induces N limitation. (A) No fertilization. (B) N fertilizer added.

5. Constraints on N Fixation

Pathways of loss that are independent of excess available nutrient pools exist—and where they are quantitatively important, they are sufficient to explain nutrient limitation by elements other than N. Moreover, given the greater mobility of N relative to P and (as nitrate) relative to most other elements, and given the importance of N trace-gas fluxes, it is reasonable to speculate that these pathways of loss would make N more likely than P to limit NPP in many terrestrial ecosystems, in the long term — were it not for N, fixation. However, a system dominated by N fixers has the capacity to add N at least as fast as it can be lost, by all of these pathways. How can N, fixation be sufficiently constrained so that N, fixers do not respond to N deficiency with increased growth and activity?

N, fixation in the model already appears to be strongly constrained, ultimately by a higher cost of N acquisition for fixers, proximately by an absolute priority for fixed N (and the light and water equivalent to that fixed N) given to nonfixers (Vitousek and Field, 1999). Nevertheless, even where alternative pathways of N loss are important, the model simulates sufficient N, fixation to overwhelm N limitation in a very short time (Fig. Id), and to sustain an equilibrium biomass that is barely limited by N. To the extent that N supply limits real terrestrial ecosystems, N, fixation in the world must be constrained more and/or differently than is N, fixation in the model.

Vitousek and Field (1999) explored three additional constraints on N, fixation: P availability, differential herbivory on N, fixers, and a lower shade-tolerance of N2 fixers.

5.1 P Limitation

P limitation on N2 fixation is widely observed in aquatic systems; in terrestrial ecosystems there is good evidence for it from agricultural, pastoral, and some natural systems (Eisele et al, 1989; Smith, 1992; Crews, 1993). The model estimates P availability within a mass-balanced P cycle, with inputs via weathering, outputs via leaching, and a labile adsorbed fraction. Nonfixers are given priority for P, in proportion to the amount of fixed N available. If not enough available P is present to match available N, nonfixers are P limited. If P remains available after nonfixers have taken up what they can, N2 fixers can use it — at a lower C:P ratio than that of nonfixers (Pate, 1986), and up to the overall limit to NPP set by light or water availability (Vitousek and Field, 1999). Limitation by other elements (e.g., Mo; Silvester, 1989) could be treated similarly.

5.2 Grazing

Differential grazing on N, fixers in comparison to nonfixers is often observed (e.g., Hulme, 1994; 1996; Ritchie and Tilman, 1995; Ritchie et al., 1998); it can occur because N, fixers generally have higher concentrations of N and protein than do non-fixers. Indeed, McKcy (1994) suggested that legumes could have evolved the rhizobial symbiosis in part because of their N-demanding lifestyle. High levels of chemical defense can reduce the amount of grazing on fixers, but this in effect further raises the energetic cost of N2 fixation. There is good evidence that preferential grazing by deer on legumes virtually eliminates N2 fixers and is responsible for maintaining N limitation of production and biomass accumulation at Cedar Creek, Minnesota (Ritchie and Tilman, 1995; Ritchie et al., 1998). The model includes this preferential grazing by removing more biomass from fixers than nonfixers, effectively reducing production by N2 fixers early in soil development (Vitousek and Field, 1999). A more realistic demographic analysis of the effects of grazing on N2 fixers could yield a more sustained effect.

5.3 Shade Tolerance

It could be difficult for fixers to colonize under an established canopy of nonfixers, due to their greater cost for N acquisition (Gutschick, 1987; Vitousek and Howarth, 1991). If N2 fixers have systematically lower shade tolerance, they would be unable to respond to N deficiency in a closed-canopy ecosystem, even where N is limiting to NPP and biomass accumulation. The model simulates this effect by suppressing growth of N2 fixers in proportion to a sigmoidal function of the biomass of nonfixers.

A comparison of the initial model (with N fixers, and "with versus without" alternative pathways of N loss) with a revised version of the model that includes these three additional constraints to N2 fixation is displayed in Figure 3. Neither the additional constraints on N, fixation alone (Fig. 3c) nor the alternative pathways of N-loss alone are sufficient to drive more than a marginal N limitation on biomass, at equilibrium (Fig. 3b). However, the combination (Fig. 3d) of both alternative loss pathways and additional constraints on N2 fixation yields a system that is strongly limited by available N at equilibrium. For completeness, one would need to carry out a similar analysis of nonsymbiotic N2 function. Some of the same mechanisms (e.g., energy cost of N2 fixation, P limitation to N2 fixers) could be important, and some others (e.g., decomposers might not be limited by N supply, even in sites where NPP is N limited) could also contribute Vitousek and Hobbie 2000.

6. Conclusions

Overall, we think that Figure 3a displays the logical consequences of the ways that many ecologists think of the N cycle in terrestrial ecosystems—with losses of N occurring primarily when N is available in excess, with N2 fixation constrained to some extent by its energetic cost. If this were a reasonable representation of the world, N limitation would be a transient phenomenon, there would be very little C storage resulting from human alteration of the N cycle, and any stimulation in NPP and/or C storage resulting from increased C02 would not be constrained for long by N availability.

The pervasiveness of terrestrial ecosytems where N availability demonstrably limits NPP and other ecosystem processes suggests that alternative pathways of N loss and additional constraints on N2 fixation should be fundamental parts of our view of the N cycle. To the extent that these are important, anthropogenic

A /


i V



r — t- — ^

1 / x

u v

100 200 300 400 500

100 200 300 400 500 0


FIGURE 3 The effects oil biomass and the extent of N limitation of adding constraints on N2 fixation. The constraints include differentially severe effects of grazing, P limitation, and shade intolerance on N, fixers. In each case, a pulse of simulated N fertilizer was added beginning in year 300 (50 kg N ha ' year ', continued for 20 years) to illustrate the extent of N limitation. (A) No additional constraints to N2 fixation, losses of excess available N only. (B) No additional constraints; losses of N by additional pathways. (C) N, fixation constrained by grazing, P limitation, and shade intolerance; losses of excess available N only. (D) N, fixation constrained, and losses of N by additional pathways. Revised from Vitousek and Field (1999).

changes in the N cycle can have fundamental effects on terrestrial ecosystems; anthropogenic N could increase C storage on land at equilibrium; and the long-term effects of elevated C02 on N availability will depend on how C02 interacts with pathways of N loss and with the processes constraining N2 fixation (Vitousek and Field, 1999). These are mechanisms and interactions that we ought to try to understand.


This research was supported by grants from the National Science Foundation and the Andrew Mellon Foundation.


Aber, J. D.(1992). Nitrogen cycling and nitrogen saturation in temperate forest ecosystems. Trends F.coi. F.volut. 7, 220-223. Aber, J. D., and Driscoll, C. T. (1997). Effects of land use, climate variation, and N deposition on N cycling and C storage in northern hardwood forests. Global Biogeocheni. Cycles 11, 639-648. Aber, J. D„ McDowell, W. II., Nedelhoffer, K„ Magill, A., Berntson, G„ Ka-

makea, M., McNulty, S., Currie, W., Rustad, L., and Fernandez, I. (1998). Nitrogen saturation in temperate forest ecosystems. Bioscience 48, 921-934.

Agren, G. I., and Bosatta, E. (1988). Nitrogen saturation of terrestrial ecosystems. Environ Pollnt. 54, 185-197.

Austin, A. T. and Vitousek, P. M. (1998). Nutrient dynamics on a precipitation gradient in Hawai'i. Oecologia 113,519-529.

Berendse, F., Aerts, R„ and Bobbink, R. (1993). Atmospheric nitrogen deposition and its impact on terrestrial ecosystems. In "Landscape Ecology of a Stressed Environment" C. C. Vos and P. Opdam, (Eds.), pp. 104-121. Chapman & Hall, London.

Bloom, A. J., Chapin III, F. S„ and Mooney, H. A. (1985). Resource limitation in plants—An economic analogy. Annu. Rev. Ecol. Syst. 16, 363-393.

Bormann, F. H„ and Likens, G. E. (1979). "Pattern and Processes in a Forested Ecosystem". Springer- Verlag, New York.

Burke, I. C., Lauenroth, W. K., and Parton, W. J. (1997). Regional and temporal variation in net primary production and nitrogen mineralization in grasslands. Ecology 78, 1330-1340.

Chadwick, O. A., Derry, L. A., Vitousek, P. M„ Huebert, B. J., and Hedin, L. O. (1999). Changing sources of nutrients during four million years of ecosystem development. Nature 397, 491-497.

Cochran, M. F., and Berner, R. A. (1997). Promotion of chemical weathering by higher plants: Field observations on Hawaiian basalts. Chem.l Geol. 132, 71-85.

Crews, T. E. (1993). Phosphorus regulation of nitrogen fixation in a traditional Mexican agroecosystem. Biogeochemistry 21, 141 - 166.

Currie, W. S„ Aber,). D„ McDowell, W. H„ Boone, R. D„ and Magill, A. H. (1996). Vertical transport of dissolved organic C and N under long-term N amendments in pine and hardwood forests. Biogeochemistry 35, 471-505.

Dahlgren, R. A. (1994). Soil acidification and nitrogen saturation from weathering of ammonium-bearing rock. Nature 368, 838—841.

Davis, M. B„ Moeller, R. E„ Likens, G. E„ Ford, M. S., Sherman, J., and Goulden, C. (1985). Paleoecology of Mirror Lake and its watershed. In An Ecosystem Approach to Aquatic Ecology: Mirror Lake and Its Environment." (G. E. Likens, Ed), pp. 410-429. Springer-Verlag, New York.

Eisele, K. A., Schimel, D. S„ Kapustka, L. A., and l'arton W. J. (1989). Effects of available P and N:P ratios on non-symbiotic dinitrogen fixation in tall grass prairie soils. Oecologia 79, 471-474.

Elser, J. J., Dobberfuhl, D. R„ MacKay, N. A., and Schampel, J. H. (1996). Organism size, life history, and N:P stoichiometrv. Bioscience 46, 674-684.

Field, C. B„ Chapin III, F. S., Matson, P. A., and Mooney, H. A. (1992). Responses of terrestrial ecosystems to the changing atmosphere: A resource-based approach. Annu. Rev. Ecol. Syst. 23, 201-235.

Firestone, M. K., and Davidson, E. A. (1989). Microbiological basis of NO and NO: production and consumption in soil. pp. 7-21. In "Exchange of Trace Gases Between Terrestrial Ecosystems and the Atmosphere". (M. O. Andreae and D. S. Schimel, Eds.), Wiley, London.

Galloway, J. N„ Schlesinger, W. H„ Levy 11, H„ Michaels, A., and Schnoor, J. L. (1995). Nitrogen fixation: Atmospheric enhancement—Environmental response. Global Biogeochcm. Cycles 9, 235-252.

Gleeson, S. K„ and Tilman, D. (1992). Plant allocation and the multiple limitation hypothesis. Am. Nat. 139, 1322-1343.

Gutschick, V. P. (1987). "A Functional Biology of Crop Plants." Timber Press, Portland.

Hedin, L. O., Armesto, J. J-, and Johnson, A. H. (1995). Patterns of nutrient loss from unpolluted, old-growth temperate forests: evaluation of bio-geochemical theory. Ecology 76, 493-509.

Howarth, R. W, Billen, G., Swaney, D„ Townsend, A., Jaworski, N„ Lajtha, K., Downing, J. A., Elmgren, R., Caraco, N., Jordan, T., Berendse, F„ Freney. J„ Kudeyarov, V., Murdoch, P., and Zhu Zhao-liang. (1996). Regional nitrogen budgets and riverine N&P fluxes for the drainages to the North Atlantic Ocean: Natural and human influences. Biogeochem-istry 35, 181-226.

Hulme, P. E. (1994). Seedling herbivory in grassland: Relative impact of vertebrate and invetebrate herbivores./. Ecol. 82, 873-880.

Hulme, P. E. (1996). Herbivores and the performance of grassland plants: A comparison of arthropod, mollusc and rodent herbivory. /. Ecol. 84, 43-51.

Kennedy, M. J., Chaduick, O. A., Vitousek, P. M., Derey, L. A., and Hende-icks, D. M., (1998). Changing Sources of base cations during ecosystem development, Hawaian Islands. Geology 26, 1015- 1018.

McGill, W. B. and Cole, C. V. (1981). Comparative aspects of cycling of organic C, N, S, and P through soil organic matter. Geoderma 26, 267-286.

McKey, D. (1994). Legumes and nitrogen: The evolutionary ecology of a nitrogen-demanding lifestyle. "Advances in Legume Systematics: Part 5 — The Nitrogen Factor." J. L. Sprent and D. McKey, (Eds.), pp 211-228. Royal Botanic Gardens, Kew, England.

Melillo, J. M„ Prentice, I. C., Farquhar, G. D„ Schulze, E.-D., and Sala, O. E. (1996). Terrestrial ecosystems: Biotic feedbacks to climate. In "Intergov ernmental Panel on Climate Change 1995: Scientific Assessment of Climate Change". (J. Houghton, L.G. Meira Filho, B. A. Callander, N. Harris, A. Kattenberg, and K. Maskell, Eds.), Cambridge Univ. Press, Cambridge, United Kingdom.

Metherell, A. K„ Harding, L. A., Cole, C. V„ and Parton, W. J. (1993). "CENTURY Soil Organic Matter Model Environment: Technical Documentation." Great Plains System Research Unit Technical Report 4, USDA-ARS, Fort Collins, Colorado.

Newman, E. I. (1995). Phosphorus inputs to terrestrial ecosystems./. Ecol. 83,713-726.

Nixon, S. W, Ammerman, J. W., Atkinson, L. P., Berounsky, V. M., Billen, G., Boicourt, W. C„ Boynton, W. R„ Church, T. M„ Ditoro, D. M„ Elmgren, R., Garber, J. H„ Giblin, A. E., Jahnke, R. A., Owens, N. P. J., Pilson, M. E. Q., and Seitzinger, S. P., (1996). The fate of nitrogen and phosphorus at the land-sea margin of the North Atlantic Ocean. Bio-geochemistry 35, 141-180.

Parton, W. J„ Mosier, A. R„ Ojima, D. S„ Valentine, D. W, Schimel, D. S., Weier, K., and Kulmala, A. E. (1996). Generalized model for N, and N,0 production from nitrification and denitrification. Global Bio-geochem. Cycles 10,401-412.

Pate, J. S. (1986). Economy of symbiotic N fixation. "On the Economy of Plant Form and Function" In (T. J. Givnish, Ed.), pp. 299-325. Cambridge Univ. Press, Cambridge, United Kingdom.

Peterjohn, W. T. and Schlesinger, W .H. (1990). Nitrogen loss from deserts of the southwestern United States. Biogeochemistry 10, 67- 79.

Rastetter, E. B. and Shaver, G. R., 1992. A model of multiple element limitation for acclimating vegetation. Ecology 73, 1157-1174.

Ritchie, M. E. and Tilman, D., (1995). Responses of legumes to herbivores and nutrients during succession on a nitrogen-poor soil. Ecology 76, 2648-2655.

Ritchie, M. E„ Tilman, D„ and Knops, J. M. H. (1998). Herbivore effects on plant and nitrogen dynamics in oak savanna. Ecology 79, 165-177.

Schimel, D. S„ Brassell, B. H., and Parton, W. J. (1997). Equilibration of the terrestrial water, nitrogen, and carbon cycles. Proc. Natl. Acad. Sci. USA 94, 8280-8283.

Schimel, D. S., Enting, I. G„ Heimann, M„ Wigley, T. M. L„ Raynaud, D„ Alves, D., and Siegenthaler U., (1995). C02 and the carbon cycle. In "Climate Change 1994: Radiative Forcing of Climate Change". (J. T. Houghton, L. G. Meira Filho, and K. Maskell Eds.), pp. 39-71. Cambridge Univ. Press, Cambridge, United Kingdom.

Schindler, D. W. (1977). Evolution of phosphorus limitation in lakes. Science 195, 260-262.

Schulze, E.-D. (1989). Air pollution and forest decline in a spruce (Picea abies) forest. Science 244, 776-783.

Silvester, W. B. (1989). Molybdenum limitation of asymbiotic nitrogen fixation in forests of Pacific Northwest America. Soil Biol. Biochem. 21, 283-289.

Smith, V. H. (1992). Effects of nitrogen:phosphorus supply ratios in nitrogen fixation in agricultural and pastoral systems. Biogeochemistry 18, 19-35.

Sprent, J. I., and Sprent, P. (1990). "Nitrogen Fixing Organisms: Pure and Applied Aspects." Chapman 8c Hall, London.

Tilman, D., (1987). Secondary succession and the pattern of plant dominance along experimental nitrogen gradients. Ecological Monogzaphs 57, 189-214.

Townsend, A. R„ Braswell, B. H„ Holland, E. A., and Penner, J. E. (1996). Spatial and temporal patterns in terrestrial carbon storage due to deposition of fossil fuel nitrogen. Ecol. Appl. 6, 806-814.

Vilousek, P. M„ Aber, I. D„ Howarth, R. W„ Likens, G. E„ Matson, P. A., Schindler, D. W„ Schlesinger, W. II. and Tilman, D. (1997a). Human alteration of the global nitrogen cycle: Sources and consequences. Ecol. Appl. 7, 737-750.

Vitousek, P. M., Chadwick, O. A., Crews, T., Fownes, J., Hendricks, D., and Herbert, D. (1997b). Soil and ecosystem development across the Hawaiian Islands. GSA Today 7(9), 1-8.

Vitousek, P. M., and Field, C. B. (1999). Ecosystem constraints to symbiotic nitrogen fixers: A simple model and its implications. Biogeochem-istry 46, 179-202.

Vitousek, P. M„ Hedin, L. O., Matson, P. A., Fownes, J. H., and Neff, J., (1998). Within-system element cycles, input-output budgets, and nu trient limitation. "Successes, Limitations, and Frontiers in Ecosystem Science". (M. Pace and P. Groffman, Eds.), pp. 432-452. Springer-Verlag, Berlin.

Vitousek, P. M„ and Hobbie, S. (2000). Heterotrophic nitrogen Fixation in decomposing litter: Patterns and regulation. Ecology 81, 2366-2376.

Vitousek, P. M., and Howarth, R. W. (1991). Nitrogen limitation on land and in the sea: How can it occur? Biogeochemistry 13, 87-115.

Vitousek, P. M„ and Reiners, W. A. (1975). Ecosystem succession and nutrient retention: A hypothesis. Bioscience 25, 376-381.

Walker, T. W. and Syers, J. K. (1976). The fate of phosphorus during pedogenesis. Geoderma 15,1- 19.

This Page Intentionally Left Blank

Was this article helpful?

0 0

Post a comment