Until 1850, natural BNF, cultivation-induced BNF (e.g., planting of leguminous crops), and incorporated organic matter were the only sources of N for agricultural production (Smil 2001). Near the turn of the century, Nr from guano and sodium nitrate deposits was harvested and exported from the arid Pacific islands and South American deserts (Smil 2001). By the late 1920s, early industrial processes, albeit inefficient, were commonly used to produce NH3 (Galloway 2003). Due to the efforts of Fritz Haber and Carl Bosch, the Haber-Bosch process became the largest source of nitrogenous fertilizer after the 1950s, and replaced BNF as the dominant source of NH3 production (Smil 2001). From 1890 to 1990, anthropogenically created Nr increased almost ninefold (Galloway 2003). During this time, global population more than tripled, partly due to increased food production.
Since the industrial revolution, an additional source of anthropogenic N input has been fossil fuel combustion, which is used to generate energy (e.g., to power automobiles). During combustion of fossil fuels, high temperatures and pressures provide energy to produce NO from N2 oxidation (Galloway 2003). Additionally, when fossil fuel is extracted and burned, fossil N may become reactive (i.e., NOx emissions, Galloway 2003). During the 1970s, scientists began to recognize that N inputs were accumulating in the environment and affecting ecosystem functioning (Galloway 2003).
Between 1890 and 1990, global reactive nitrogen (Nr) creation had increased nearly 50% (Galloway and Cowling 2002). During this period, atmospheric emissions of Nr species reportedly increased 250% and deposition to marine and terrestrial ecosystems increased over 200% (Galloway and Cowling 2002). Additionally, there was a reported fourfold increase in riverine dissolved inorganic N fluxes to coasts (Galloway and Cowling 2002). N is a critical limiting nutrient in many systems, including forests, wetlands, and coastal and marine ecosystems (Vitousek and Howarth 1991, Rabalais 2002); therefore, this change in emissions and distribution of Nr has resulted in substantial consequences for aquatic and terrestrial ecosystems.
Atmospheric Nr inputs mainly include oxides of N (NOx), ammonia (NH3), and nitrous oxide (N2O) from aquatic and terrestrial ecosystems (Schlesinger 1997), and NOx from fossil fuel and biomass combustion (Galloway et al. 2003).
In agroecosystems, fertilizer application has increased microbial nitrification (aerobic process in which microorganisms oxidize ammonium [NH4+] to nitrate [NO3-]) and denitrification (anaerobic process in which microorganisms reduce NO3- to atmospheric nitrogen gas [N2]). Both processes naturally leak nitric oxide (NO) and nitrous oxide (N2O) to the atmosphere (Schlesinger 1997). Of particular concern is N2O, which has an average atmospheric lifetime of 114-120 years (Houghton et al. 2001), and is 300 times more effective than CO2 as a greenhouse gas (Schlesinger 1997) . NOx produced by industrial processes, automobiles and agricultural fertilization and NH3 emitted from soils (i.e., as an additional byproduct of nitrification, Schlesinger 1997) and livestock operations are transported to downwind ecosystems, influencing N cycling and nutrient losses. Galloway et al. (2003) cited six major effects of NOx and NH3 emissions: 1) decreased atmospheric visibility due to ammonium aerosols (fine particulate matter [PM]); 2) elevated ozone concentrations; 3) ozone and PM affects human health (e.g., respiratory diseases, cancer); 4) increases in relative forcing and global climate change; 5) decreased agricultural productivity due to ozone deposition; and 6) ecosystem acidification (e.g., Houdijk et al. 1993) and eutrophication.
Terrestrial and aquatic ecosystems receive Nr inputs from the atmosphere through wet and dry deposition (Galloway 2003). Atmospheric Nr species can be deposited to ecosystems in precipitation (e.g., NO3-, NH4+, organic N compounds), as gases (e.g., NH3 and gaseous nitric acid [HNO3]), or as aerosols (e.g., ammonium nitrate [NH4NO3]; Galloway 2003). Aquatic ecosystems receive additional nitrogen from surface runoff and riverine inputs (Rabalais 2002). Increased N deposition can acidify soils, streams, and lakes and alter forest and grassland productivity. In forest and grassland ecosystems, Nr inputs have produced initial increases in productivity followed by declines as critical thresholds are exceeded (Galloway 2003). Nr effects on biodiversity, carbon cycling, and changes in species composition have also been demonstrated. In highly developed areas of near shore coastal ocean and estuarine systems, rivers deliver direct (e.g., surface runoff) and indirect (e.g., groundwater contamination) N inputs from agroecosystems (Rabalais 2002). Increased N inputs can result in freshwater acidification and eutrophication of marine waters.
Much of terrestrial growth in temperate systems is limited by N; therefore, N inputs (i.e., through deposition and fertilization) can increase N availability, which temporarily increases N uptake, plant and microbial growth, and N accumulation in plant biomass and soil organic matter (Aber et al. 1989). Incorporation of greater amounts of N in organic matter decreases C:N ratios, increasing mineral N release (NH4+) during organic matter decomposition by heterotrophic microbes (i.e., ammonification, Aber 1992). As ammonification increases, so does nitrification of the mineralized N. Because microbial nitrification and denitrification are “leaky”, N deposition is expected to increase trace gas emissions (Matson et al. 2002). Additionally, with increasing NH4+ accumulation in the soil, nitrification processes release hydrogen ions, which acidify the soil. NO3-, the product of nitrification, is highly mobile and can be leached from the soil, along with positively charged alkaline minerals such as calcium and magnesium (Schlesinger 1997). In acid soils, mobilized aluminum ions can reach toxic concentrations, negatively affecting both terrestrial and adjacent aquatic ecosystems.
Anthropogenic sources of N generally reach upland forests through deposition (Aber et al. 2003). A potential concern of increased N deposition due to human activities is altered nutrient cycling in forest ecosystems. Numerous studies have demonstrated both positive and negative impacts of atmospheric N deposition on forest productivity and carbon storage. Added N is often rapidly immobilized by microbes (Nadelhoffer et al. 1999), and the effect of the remaining available N depends on the plant community’s capacity for N uptake (Bauer et al. 2004). In systems with high uptake, N is assimilated into the plant biomass, leading to enhanced net primary productivity (NPP) and possibly increased carbon sequestration through greater photosynthetic capacity. However, ecosystem responses to N additions are contingent upon many site-specific factors including climate, land-use history, and amount of N additions. For example, in the Northeastern United States, hardwood stands receiving chronic N inputs have demonstrated greater capacity to retain N and increase annual net primary productivity (ANPP) than conifer stands (Magill et al. 2004). Once N input exceeds system demand, N may be lost via leaching and gas fluxes. When available N exceeds the ecosystem’s (i.e., vegetation, soil, and microbes, etc.) uptake capacity, N saturation occurs and excess N is lost to surface waters, groundwater, and the atmosphere (Aber et al. 1989, Bauer et al. 2004, Magill et al. 2004). N saturation can result in nutrient imbalances (e.g., loss of calcium due to nitrate leaching) and possible forest decline (Aber 1992).
A recently published, 15-year study of chronic N additions at the Harvard Forest Long Term Ecological Research (LTER) program has elucidated many impacts of increased nitrogen deposition on nutrient cycling in temperate forests. Magill et al. (2004) found that chronic N additions resulted in greater leaching losses, increased pine mortality, and cessation of biomass accumulation. Additionally, Bauer et al. (2004) reported that chronic N additions resulted in accumulation of non-photosynthetic N and subsequently reduced photosynthetic capacity, supposedly leading to severe carbon stress and mortality. These findings negate previous hypotheses that increased N inputs would increase NPP and carbon sequestration.
Many plant communities have evolved under low nutrient conditions; therefore, increased N inputs can alter biotic and abiotic interactions, leading to changes in community composition. Several nutrient addition studies have shown that increased N inputs lead to dominance of fast-growing plant species, with associated declines in species richness (Huenneke et al. 1990, Tilman 1997, Wilson and Tilman 2002). Other studies have found that secondary responses of the system to N enrichment, including soil acidification and changes in mycorrhizal communities have allowed stress-tolerant species to out-compete sensitive species (Houdijk et al. 1993, Egerton-Warburton and Allen 2000). For example, both Aerts and Berendse (1988) and Bobbink et al. (1992) found evidence that increased N availability has resulted in declines in species-diverse heathlands. Heathlands are characterized by N-poor soils, which exclude N-demanding grasses; however, with increasing N deposition and soil acidification, invading grasslands replace lowland heath.
In a more recent experimental study of N fertilization and disturbance (i.e., tillage) in old field succession, Wilson and Tilman (2002) found that species richness decreased with increasing N, regardless of disturbance level. Competition experiments showed that competitive dominants excluded competitively inferior species between disturbance events. With increased N inputs, competition shifted from belowground to aboveground (i.e., to competition for light), and patch colonization rates significantly decreased. These internal changes can dramatically affect the community by shifting the balance of competition-colonization tradeoffs between species. In patch-based systems, regional coexistence can occur through tradeoffs in competitive and colonizing abilities given sufficiently high disturbance rates (Hastings 1980). That is, with inverse ranking of competitive and colonizing abilities, plants can coexist in space and time as disturbance removes superior competitors from patches, allowing for establishment of superior colonizers. However, as demonstrated by Wilson and Tilman (2002), increased nutrient inputs can negate tradeoffs, resulting in competitive exclusion of these superior colonizers/poor competitors.
As reviewed by Rabalais (2002), aquatic ecosystems also exhibit varied responses to nitrogen enrichment. NO3- loading from N saturated, terrestrial ecosystems can lead to acidification of downstream freshwater systems and eutrophication of downstream marine systems (Rabalais 2002). Freshwater acidification can cause aluminum toxicity and mortality of pH-sensitive fish species (Rabalais 2002). Because marine systems are generally nitrogen-limited, excessive N inputs can result in water quality degradation due to toxic algal blooms, oxygen deficiency, habitat loss, decreases in biodiversity, and fishery losses (Rabalais 2002).
Atmospheric N deposition in terrestrial landscapes can be transformed through soil microbial processes to biologically available nitrogen, which can result in surface-water acidification, and loss of biodiversity. NO3- and NH4+ inputs from terrestrial systems and the atmosphere can acidify freshwater systems when there is little buffering capacity due to soil acidification (see above discussion, Rabalais 2002). As described by Driscoll (2001), N pollution in Europe, the Northeastern United States, and Asia is a current concern for freshwater acidification. Lake acidification studies in the Experimental Lake Area (ELA) in northwestern Ontario clearly demonstrated the negative effects of increased acidity on a native fish species: lake trout (Salvelinus namaycush) recruitment and growth dramatically decreased due to extirpation of its key prey species during acidification (Mills et al. 2000).
Urbanization, deforestation, and agricultural activities largely contribute sediment and nutrient inputs to coastal waters via rivers (Rabalais 2002). Increased nutrient inputs to marine systems have shown both short-term increases in productivity and fishery yields, and long-term detrimental effects of eutrophication. Tripling of NO3- loads in the Mississippi River in the last half of the 20th century have been correlated with increased fishery yields in waters surrounding the Mississippi delta (Grimes 2001); however, these nutrient inputs have produced seasonal hypoxia (oxygen concentrations less than 2-3 mg L-1, "dead zones") in the Gulf of Mexico (Rabalais 2002, Galloway 2003). In estuarine and coastal systems, high nutrient inputs increase primary production (e.g., phytoplankton, sea grasses, macroalgae), which increase turbidity with resulting decreases in light penetration throughout the water column. Consequently, submerged vegetation growth declines, which reduces habitat complexity and oxygen production. The increased primary (i.e., phytoplankton, macroalgae, etc.) production leads to a flux of carbon to bottom waters when decaying organic matter (i.e., senescent primary production) sinks and is consumed by aerobic bacteria lower in the water column. As a result, oxygen consumption in bottom waters is greater than diffusion of oxygen from surface waters .
Most Nr applied to global agroecosystems cascades through the atmosphere and aquatic and terrestrial ecosystems until it is converted to N2, primarily through denitrification (Galloway 2003). Although terrestrial denitrification produces gaseous intermediates (nitric oxide [NO] and nitrous oxide [N2O]), the last step—microbial production of N2—is critical because atmospheric N2 is a sink for Nr (Davidson and Seitzinger 2006). Many studies have clearly demonstrated that managed buffer strips and wetlands can remove significant amounts of nitrate (NO3-) from agricultural systems through denitrification (e.g., Tate et al. 2000, Jackson et al. 2006). Such management may help attenuate the undesirable cascading effects and eliminate environmental Nr accumulation (Galloway et al . 2003).
Human activities dominate the global and most regional N cycles (Galloway et al. 2004). N inputs have shown negative consequences for both nutrient cycling and native species diversity in terrestrial and aquatic systems. In fact, due to long-term impacts on food webs, Nr inputs are widely considered the most critical pollution problem in marine systems (Rabalais 2002). In both terrestrial and aquatic ecosystems, responses to N enrichment vary; however, a general re-occurring theme is the importance of thresholds (e.g., nitrogen saturation) in system nutrient retention capacity. In order to control the N cascade, there must be integration of scientific disciplines and further work on Nr storage and denitrification rates (Davidson and Seitzinger 2006).