sign in sign up
<<
Home , Limnology

Limnology and Third Edition

ROBERT G. WETZEL

ACADEMIC PRESS 260 1

AHarcoun Science and Technology Cornpony

San Dieeo San Francisco Boston London Svdnev Tokyo 12-11 Estimates of the in Various Compartments in Lake Wingra, Wisconsin, Midsummef

Percs,irnge of rota1 N Nitrogen Compartment (merric tons) A" B'

Soluble and particulate N in 3.2 0.5 50.6 Macr~phyteN 1.2 0.2 19.3 Sediment, upper 10 cm Organic N 138 23.0 - Interstitial water NH,-N 0.74 0.1 11.5 Organic N 0.19 0.03 3.0 Exchangeable NIi-N 1.0 0.2 15.6 Sediment, 10-30 cm Organic N 460 74.3 - Interstitial water NH,-N 4.6 0.8 - Organic N 0.51 0.08 - Exchangeable NH,-N 5.4 0.9 - Total Total N in system 615 Tatal N excluding sediment organic N and all N below 10 crn in sediment 6.3 'Modified from lsirimah etal. (1976). b~otalN insystem. 'Total N excluding sediment organic N and all N below 10 cm in sediment.

conditions (Table 12-11). Much (97%) of the total ni- by differences. Nonetheless, the approxi- trogen occurred in the sediments as organic nitrogen mate calculated budgets are instructive in characteriz- (column A) but was not readily available for metabo- ing the inputs, outputs, and retention of by lism in the lake system (Isirimah et al., 1976). Neglect- ecosystems. ing these relatively unreactive nitrogen , about The nitrogen budget of , Wisconsin, 50% of the "available" nitrogen existed in the water demonstrates the roughly equivalent contributions column, mostly in dissolved form, 20% in the macro- from runoff, ground water, and (Table phytes, and 30% in the interstitial water of surficial sed- 12-12). Major losses of nitrogen occur via sedimenta- iments (column B). Rapid turnover rates of NH4-N tion, denitrification, and outflow. Loss by seepage out occurred in the water but not in the sediments. of the basin was probably small, since most lake basins NO3-N turnover was slower in the water than in the are well sealed (cf. Wetzel and Otsuki, 1974). In Table sediments, where about 80% of the NOsN was 12-13; the nitrogen budget for a large of com- rapidly denitrified to N2. plex morphometry accounts for inflows from all river influents and losses by sedimentation and outflow. Al- though the total nitrogen content of this reservoir de- VI. NITROGEN BUDGETS creased in 1962 because of low water levels, inputs of atmospherically derived nitrogen amounted to roughly Detailed evaluations of the nitrogen cycle that 10,000 tons per year. About 1300 metric tons entered include close-interval quantitative measurements of in- with rain and snow; the remainder may be attributed puts, metabolic dynamics, and outputs are available for largely to biological fixation. Mirror Lake, a kettle lake only a few freshwater ecosystems. The cycle is obvi- located in a crystalline granitic region of the ously complex and requires accurate analyses of the White Mountains of central New Hampshire, receives dynamics of all components for several years. Only a appreciable (half) nitrogen loading from atmospheric few approximate mass balance analyses of the nitrogen sources (Table 12-14). Primary outputs are via perma- budgets of and are available. In these nent sedimentation and river outflow in approximately analyses, the rates of bacterial metabolism are obtained equal quantities of inorganic and organic forms. Some indirectly, by assumptions of microbial processing or 54% of the annual gross sedimentation of organic TABLE 12-12 Estimated Nitrogen Budget of Lake Mendota, Wisconsin"

Nitrogen income Nitrogen losses Sources knNy~' % Losses ka Nyr-' %

Municipal and industrial 21,200 10.4 Outflow 41,300 20.4 wastewater Denitrification 28,100 13.9 13,700 6.8 Ftsh catch 11,300 5.6 Rural runoff 23,500 11.6 Weed removal 3250 1.6 Precipitation on lake surface 43,900 21.6 Loss to ground water b 6 Ground water Sedimentation and otherc 118,850 -58.6 Streams 35,900 17.7 202,800 100 Seepage 28,500 Nitrogen fixation 36,100 Marsh drainage -d 202,800

'Modified from Brezanik and Lee (1968),with dara improvemei~tsby Keeney (1972) bUnknawn; likely very small in this lake. 'By difference between total of other estimated losses and sum of income sources. dConsidered significant, but dara unavailable. nitrogen was mineralized in a large, oligotrophic lake namics of internal processing or metabolic controls in Sweden (Jonsson and Jansson, 1997). About two- mechanisms. From an applied viewpoint, however, thirds of the nitrogen mineralized in the sediments, re- such mass balance information is useful in relation to leased to the overlying water, and transported out of nitrogen loading and the effects of this loading on al- the lake as via the outlet during turnovers in terations of , as will be discussed in some and . Only ca. 4% of organic nitrogen detail with (cf. Chap. 13). remained in the sediments; the remainder was lost via Annual elemental mass balance budgets have also denitrification at about 1 mg N m-2 yrr'. been used to characterize the inputs, outputs, and re- Analyses of nitrogen budgets of some 40 lakes of tention (i.e., the difference between nitrogen inputs and very different states of productivity indicated that the outputs) of ecosystems. Hydrological variations removal of nitrogen is usually dominated by denitrifica- can be great in streams and they influence the fluxes. tion, which is directly correlated to Even among more hydrologically stable streams, inter- loadings (Hellstrom, 1996). Where nitrogen fixation is annual variations in can be large (Meyer and significant, the increase in totalnitrogen from N, fixa- Likens, 1979; Meyer et al., 1981; Grimm, 1987). In tion is generally proportional to the concentration of arid regions, flooding events can cause extreme varia- total phosphorus minus a fraction of the estimated ni- tions in annual budgets. trogen concentration without any Nz fixation. General Examination of representative nitrogen budgets of nitrogen budgets of this type leave much to be desired streams indicates that total nitrogen storage was simi- because they provide relatively little insight into the dy- lar even though the distributions among storage

TABLE 12-13 Estimated Nitrogen Budget of the Rybinsk Reservoir, Russia, in Metric Tons N per Year"

Time

1 June 1960- 1 April 1961 - 1 Julie 1962- Factors measured 1 lune 1961 1 Auri11962 1 Tune 1963

A Measured nitrogen balance of water mass + 12,519 + 5399 + 12,170 B Difference between inflow and outflow - 180 - 7329 + 1565 A - B = C Increase in water mass + 12,699 + 12,728 + 10,605 D Precipitated in sediments (20-yr average) + 12,500 + 12,500 + 12,500 C - D = E Theoretical balance in water mass + 199 + 228 - 1895 A-E Nitrogen input from atmosphere and fixation + 12,320 +5171 + 14,065 TABLE 12-14 Average Annual Nitrogen Budget for Mirror components differed (Table 12-15). Nitrogen storage Lake, New Hampshire, 1970-1975' was largely associated with woody debris (59-80%) in forested streams, whereas in an open desert stream as much as 93% was in and autochthonous detritus. Inputs Some nutrients are stored in consumer animals, al- Precipitation though turnover rates in these organisms are relatively Inorganic 112 i 10.7 slow. Dissolved organic nitrogen exports are less than Organic 7 inputs and indicated biological utilization associated Lirter 13.6 with mineralization and potentially some sorption and Dry -15 Fluvial flocculation processes (Triska et al., 1984). Most par- Dissolved ticulate organic inputs increased in nitrogen concentra- Inorganic 35 i 7.2 tions prior to export (Table 12-16). Organic 46 It has been predicted that nitrogen re- Particulate, organic 9 tention would increase as rates of net ecosystem pro- Total 238 duction and accumulation increase and then Outputs decline or increase at a slower rate as biomass Gaseous flux > approaches a steady state (Vitousek and Reiners, 1975; Fluvial Dissolved Grimm and Fisher, 1986). At steady state, Inorganic 61 2 8.8 inputs would equal outputs, even though the state and Organic 54 forms of the nutrients may change. This relationship Particulate, organic 11 has been generally verified in studies that evaluated the Net insect emergence 13 patterns of nitrogen retention during successional Permanent sedimentation 127-139 Total 266-278 recovery of algal and macroinvertebrate communities after they were disturbed by flash-flooding events and Change in lake storage - 6 (decrease in NO3- storage) reduced to virtual zero levels (Grimm, 1987). Biomass and stored nitrogen increased during the early to cModified from Likens (1985). middle successional stagrs of recolonization and devel- opment and were followed by declines in late stages.

TABLE 12-15 Nitrogen Mass Balances of Five Streams in Percenrage of Total in Each CategoryR - - -- Watershed 10, Beaver Creek Sycamore Bear Brook, Mare's Egg Oregon R@e, Quebec Creek, Arrrono New Homprhrre Spnn~,Ore~on

Inputs (%oftotal) Dissolved inorganic N Dissolved organic N Particularc organic N Precipitation and throughfall Litrer N, fixation Pools retained N (% of total) Fine particulate inorganic nirrogcn Large particulate organic nitrogen Producers Consumers Outputs (% of total) Dissolved inorganic N Dissolved organic N Particulate organic N Coarse PON Fine PON Emcrgencc Denitrificarion

"Data exrracred irom Mcycr et al. (1981),Naiman md Melillo (1984),Trirkaet ai. (1984),Dodds and Castenholz (19881, and Grimm (1987). b~ncluderparticulates (16%)and sedimented burial of PON (41%). 230 12. The Nilmgen Cycle

TABLE 12-16 Annual Nitrogen Budget fa!. a Small Stream A. Nitrogen Loading: Effects ol Human Activities (Unnamed)in the Western Cascade Mountains, Oregon, Heavily Forested with Mature Conifer Trees' Enrichment of fresh with nutrients needed for plant growth occurs commonly as a result of losses Sums from agricultural fertilization, loading from sewage g ,~l-~ 8 m-I and industrial wastes, and enrichment via atmospheric pollutants (especially nitrate and ). Sufficient Total nitrogen inputs 15.25 information exists about the general responses of many Hydrologic inputs 11.06 Dissolved inorganic N (NO3-N) 0.50 lakes to loading of major nutrients, especially phospho- Dirsnlved organic N 10.56 11.06 rus and nitrogen, to predict the potential changes in Bioloeical" inouts 4.19 their productivity. The productivity of Terrestrial infertile, oiigotrophic lakes is often limited by the avail- Throughfall ability of phosphorus. As phosphorus loading to fresh Litterfall Needles waters increases and lakes become more productive, ni- Leaves trogen often becomes the nutrient limiting to plant Cones, bark, twigs, wood growth. Excessive loading of these nutrients permits in- Coarse wood debris creased plant growth until other nutrients or light Microparticulate litterfall availability become limiting. Miscellaneous Lateral movement The nitrogen loading concepts in relation to in- Needles creased lake productivity are best discussed simultane- Leaves ously with phosphorus loading and limitations. Both Cones, bark, twigs, wood will be treated in the following chapter. Miscellaneous Aquatic N,-fixation Twigs VII. NITROGEN DYNAMICS IN STREAMS Bark AND Chips Wood debris A. General Nutrient Dynamics In Flowingwaters Moss Particulate organic N pool Nutrients move unidirectionally within running wa- Fine particulate organic matter ters. Dissolved substances move downstream, may be 1 mm-250 pm 250 m-80 pm bound or assimilated for a period of time, and later be 40pm released for further movement downgradient. As materi- Large prriculare organic mlrrer als and chemical mass cycle among biota and abiotic Needles components of the stream ecosystem, they are trans- Leaves ported downstream, in processes that resemble spirals Cones, twigs, bark, wood Miscellaneous and have been termed nuhhentspiraling (Webster and Coarse wood debris Patten, 1979; Newbold et al., 1981; Stream Solute Work- Moss shop, 1990). Although upstream movements of nutrients Consumers can occur in backflows from eddies, migration, and Total nitrogen outputs flight of adult aquatic insects, net fluxes are downsucam. Particulate organic nitrogen Fine particulate organic matter Hydrological processes physically move water con- 1 mm-80 +m taining dissolved and particulate components to reac- <80 pm tive sites. Exchanges at reactive sites include chemical Large part~culateorganic matter ionic transformations, sorption and desorption, and Needles metabolically mediated uptake and assimilation by Leaves Cones, twigs, bark, wood biota (Fig. 12-13). Materials can be transferred from Miscellaneous the to the stationary streambed. Some of Coarse wood debris these materials will be retained, utilized by incorpora- Emergence of insects tion into living organisms, potentially transferred to Drift of insects other organisms, and subsequently released by excre- Dissolved organic nitrogen Dissolved inorganic nitrogen (NO,-N) tions or decomposition to the water column and fur- Denitrificatidn ther transported downstream. Dissolved substances in running waters may be uti- 'Data extracted from Triska et ol. (1984). lized extensively by biota or be reactive abiotically and SURFACE EXCHANGES

WATER COLUMN

INTERSTITIAL TRANSPORT WATER t I 1 SUBSURFACE EXCHANGES 1 v FIGURE 12-13 Solurc processes in streams. The two spirals represent the conrinuous exchange of solutes and -bound chemical subsrances between the water colum and the streambed and between the streambed aud interstitial water Materials in the water column and interstitial water are moving down- stream, while the streambed materials are starionaty. (From Stream Solure Workshop, 1990.) 232 12. The iVilrogen Cycle are termed nonconsewatiue. Other dissolved sub- stream nutrient fluxes (mass per unit width of river per stances, termed consewatiue, are not generally modi- unit time in the water (W) component, Fw, and in the fied chemically under normal limnological conditions biota (B) component, F,) and the exchange fluxes of or occur in such abundance that concentrations are not biotic utilization (U) of nutrients from the water com- modified substantially by biological removal. Chloride partment or regeneration (R)from the biota in mass ion is an essential nutrient that is nearly always conser- per unit area per unit time (Fig. 12-14). Details of the vative and is often used as a hydrological tracer. It methodology, particularly to evaluate the uptake flux should be noted, however, that differentiation between rates and release rates by the biotic compartments by conservative and reactive solutes is relative; a solute use of isotopes of nutrients, are delineated in Newbold may be conservative at one time or site and may be- et al. (1981,1983). come reactive at another place or time (Bencala and McKnight, 1987). B. Nutrient Limitations and Retention A nutrient atom may be used repeatedly as it passes downstream. The rate of utilization and release Nutrient limitations of primary producers and depends upon physical and biological retentiveness, heterotrophic microbiota in streams and small rivers largely by the microbiota associated with the appear to be uncommon (6. review of Newbold, streambed, that is, the extent to which the downstream 1992). Although cases of phosphorus limitations of transport of the nutrient is retarded relative to that both attached algal growth and production and of het- of the water (Newbold, 1992). The objective is to erotropbic microbial activities have been determined in quantify the average downstream distance (S, in me- a number of streams and rivers, concentrations of solu- ters) for the average nutrient atom to complete a cycle ble reactive phosphorus were consistently < 15 pg that takes an average time while it moves downstream liter-' and often <5 pg liter-'. Similarly, increased at an average velocity. The average downstream veloc- microbial growth is found with inorganic nitrogen en- ity of the nutrient may be near that of the water in richments where natural concentrations were < 50-60 large rivers but is very much slower in streams and pg literT1, hut rarely above this level. Evidence suggests rivers where nutrients reside in the sediments and mi- that in lower-order streams with a greater dominance crobiota for a high proportion of the time. of biota attached to substrata, retention of nutrients is Spiralling length (S) is the average distance a nutri- very high and that these nutrients are intensively recy- ent atom travels downstream during one cycle through cled within the attached communities (cf. Paul and the water and biotic compartment. The S equals the Duthie, 1989; Wetzel, 1993; Burns, 1998; Chap. 19). sum of distance traveled until uptake ("uptake length," Removal of the attached microbial community, as was S,) and the downstream distance traveled within the case in a desert stream following a that elimi- the biota until regeneration ("turnover length," SB) nated most of the biota, reduced the retention of nitro- (Fig. 12-14). The S can be calculated from the down- gen to very low levels, slowed recycling rates, and in- creased spiralling length (Grimm, 1987). Nitrogen retention increased rapidly as the attached algal and bacterial comrnu~lltyreestablished. As rivers become larger, nutrient retention is less and nutrient limitations can become more prevalent. Nutrient loadings to larger rivers, however, are often high because of human activ- ities in the drainage basins. Coupled with high non- algal particulate and light reductions, nutrient availab~lityis usually adequate to support microbial productivity within other environmental constraints. The spiralling length of a nutrient suggests the extent of availability and utilization rates. A shorter spiralling length of one nutrient versus another could FIGURE 12-14 Spiralling in a of two compart- imply that the nutrient cycling more rapidly is in ments: water (Wl and biota (B).The spiralling length (S) is the average greater demand and is possibly limiting the potential distance a nutrient atom travels downstream during one cycle through growth and productivity of the community. For exam- the water and biotic compartments. S = the sum of the uptake length ple, the shorter spiralling uptake length of nitrate nitro- (S,) and the turnover lcngth (S,) estimated from the downstream nu- trient fluxes (Fiand FB)and the exchange fluxes of uptake (U) and re- gen than that of phosphorus in a small stream of south- tention (R). (From Newbold, 1992. The Rivers Handbook, vol. 1, ern Spain was used as evidence of nitrogen limitation Blackwell Science Ltd. Reproduced with permission.) under different hydrological regimes (Maltchik et al., VII. Nilrogen Dynamics in Streams and Rivers 233

1994). Perhaps a more functional application is to de- TABLE 12-17 Distribution of Nitrogen Species (mg liter-') termine the spiralling rates of nutrients among different in Unpolluted Rivers" dissolved, particulate, and animal consumer groups (e.g., Newbold et al., 1983h). Such analyses demon- Mean (mix, max) strate the rapid recycling rates and very short spiralling Values lengths of the attached microbiota in comparison to NH,-N 0.018 (0.005-0.04) abiotic particulate materials and consumer metazoans. NOrN 0.0012 NOI-N 0.101 10.05-0.2) ~issdvedorganic N (DON) 0.260 (0.05-1.0) C. Nitrogen Cycling in Running Waters Ratios The inputs, transformations, and cycling of nitro- DOC:DON 20 (8-41) gen in running waters are influenced, as in standing POC:PON 8.5 (7-10) waters, to a large extent by bacterial, fungal, and other "From many sources but parriculady Malcolm and Durham (1976) microbial metabolism. The transformation processes and Meybeck (1982, 1993b). include both assimilation of nitrogen sources for uti- lization in synthesis of organic matter and growth, as well as utilization of inorganic and organic nitrogen in Amazonia to ca. 200 pg N liter-' in some temperate compounds for energy in oxidation-reduction reac- rivers (Meybeck, 1982, 1993b). Higher nitrate nitro- tions. These processes have been discussed in detail in gen concentrations are found among rivers influenced this chapter and are largely applicable to running water by agricultural runoff. An average NO,-N concentra- ecosystems as well as to lakes. Because many of these tion of nearly 100 pg N liter-' is found in natural processes require anaerobic conditions, which are river waters (Table 12-17). Maximum nitrate concen- found less frequently in rivers, the subsequent discus- trations during storm flows were directly related to the sion evaluates some of the quantitative differences in magnitude of the storms and resulting high discharge flowing waters on these processes. An important initial and inversely related to the frequency of storm events point of emphasis is that the assumptions that moving (Tate, 1990; Triska et al., 1990a). The activity of ter- waters are fully oxygenated is not always true at even restrial vegetation of the riparian zones influences the the macrolevel of hulk water volumes in streams and loadings of nitrate and ammonia to the streams; nitro- rivers and is often not the case in microenvironments gen concentrations are generally higher during periods within communities attached to surfaces and in intersti- of vegetation dormancy or following losses from har- tial waters of stream sediments. vesting or fire (e.g., Likens, 1985; Spencer and Hauer, 1991; McClain et al., 1994). The extensive transfor- D. Dissolved and Particulate Nitrogen in Rivers mation reactions focused at the upland and stream margins of the regulate and diminish The quantities of dissolved inorganic (NO3-, transfers of inorganic nitrogen from ground water to NO2-, and NH4+) and organic nitrogen compounds stream water. are highly diverse and variable because of marked dif- Dissolved organic nitrogen (DON) concentrations, ferences in the inputs from surface and although measured less frequently in natural river sources, particularly as affected by human activities waters, nearly always constitutes a major part (world (e.g., Spalding and Exner, 1993), and because of many average ca. 40%) of the total dissolved nitrogen competing reactions occurring in the nitrogen cycle. (Wetzel and Manny, 1977; Meybeck, 1982). In sub- Ammonia concentrations tend to be low (7-60 pg N arctic and humic tropical rivers, DON can constitute liter-') in natural river waters, as is the condition in over 90% of the dissolved nitrogen. Although aerobic waters of reservoirs and lakes (Meybeck, 1982, dissolved inorganic nitrogen concentrations and dis- 1993b). Ammonia nitrogen can constitute an apprecia- charge can vary widely on a diurnal basis and season- ble portion (15 to >SO%) of the total dissolved inor- ally, dissolved organic carbon and nitrogen are rela- ganic nitrogen (DIN), particularly at low and, among tively constant (Manny and Wetzel, 1973). The ratio polluted rivers, at very high DIN concentrations. As in of dissolved organic carbon to dissolved organic nitro- lakes, nitrogen concentrations are very low gen is, however, variable (8-41) and averages ca. 20 (<3 pg N liter-'; < 1.5% of total dissolved inorganic (Table 12-17). nitrogen, DIN) among well-oxygenated, unpolluted Particulate nitrogen consists of particulate organic waters. nitrogen (PON), adsorbed ammonia, and organic ni- Among major risers, nitrate concentrations range trogen adsorbed to particles (Table 12-17). On a from <25 pg N liter-' in subarctic environments and weight ratio basis, the particulate organic carbon to 234 12. The Nitrogen Cycle particulate organic nitrogen ratio is relatively constant tached microbes (bacteria, fungi, and algae) was the at 8-10 (mean ca. 8.5). primary mechanism controlling spatial and seasonal variations in the water. Diurnal daytime reductions in stream nitrate concentrations at base flow suggest that E. Nitrogen Cycling in Sediments of Flowing Waters uptake by photoautotrophs can be an important reten- Ammonium sorption to channel and riparian sedi- tive process (Burns, 1998), although the coupling of ments can be extensive. For example, in granular photosynthetic oxygen production within the sediment sediments (0.5-2.0-mm grain size) of a third-order microbial communities to diurnal changes in denitrifi- stream in California, exchangeable ammonium ranged cation are not clear in these environments. from 10 peq 100 g-' of sediment where nitrification The physical sorption of ammonia to sediment par- and subsurface exchange with stream water were high ticles is dynamically coupled to sources from ground to 115 peq 100 g-' in the riparian sedi- water and arnmonification and transformations of dis- ments where channel water and groundwater mixing solved inorganic nitrogen by nitrification, denitrifica- and nitrification potential were both low (Triska et al., tion, and nitrate reduction (Fig. 12-15). Experimental 1994). Sorbed ammonium was highest during studies of ammonium uptake by stream sediments indi- summerlautumn base flow and lowest during cated that biotic uptake by attached microflora was storm flows. Similar results were found, with some- quantitatively much greater than physical adsorption what different seasonality, in a small temperate forest (Newbold et al., 1983b; Richey et al., 1985; Aumen stream (Mulholland, 1992) and in the et al., 1995). The duration of the storage of ammonium of desert streams (Holmes et al., 1994; Jones et al., by sorptive processes is variable. Some studies indicate 1995). The riparian zone, when water-saturated, and that appreciable retention can occur in months parafluvia12 and hyporheic zones are major sources of and contribute from 12-25% of nitrate released subse- ammonium and dissolved organic nitrogen to the quently in winter by nitrification (e.g., Richey et al., stream. Once within the stream, nitrogen uptake by at- 1985), but in larger eutrophic river systems much of the ammonium is rapidly nitrified and exported down- 2Paraf7uvial zone refers to the part of the active stream channel stream as nitrate (Lipschultz et al., 1986). without that is connected hydrologically with the sur- The effectiveness of the denitrification and nitrifi- face stream water. cation processes in the hyporheic zone of stream sedi-

f \ Atmosphere ------

Nitrate reduction

FIGURE 12-15 Linkages among physical sorption of ammonia to sediment particles, sources from ground water and ammonification, and transformations of dissolved inorganic nitrogen by nitrification, denitrification, and nitrate reduction at the groundwater-surface in- terface of a stream channel. (From Triska eta!., 1994.) ments depends greatly upon the redox conditions ence microbial populations and nitrogen transfor- within the interstitial water (Fig. 12-15). Flow condi- mation rates as well as nitrogen utilization rates by tions within the sediment interstices are quite variable photosynthetic organisms. Although the nitrogen spatially because of localized differences in sediment cycle of fresh waters is understood qualitatively in composition and textures. Internal flows also change appreciable detail, the complex dynamics of quan- over time between precipitation events, which lead to titative transformation rates have not been clearly variations in both rates of groundwater inflow as well delineated, especially at the sediment-water inter- as rates of in-channel flows that penetrate down into face, where intensive bacterial metabolism occurs. the sediments (e.g., Triska et al., 1990b). In organic- 3. Dominant forms of nitrogen in fresh waters in- rich sediments, the redox profiles can be quite similar clude (a) dissolved molecular N2, (b)ammonia to those in lake sediments, in which the zone of nitrifi- nitrogen (NH,+),(c) nitrite (NO2-), (d) nitrate cation can be restricted to the upper 2-3 mm and (NOJ-), and (e) a large number of organic com- intensive denitrification occurs below the stratum of pounds (e.g., amino acids, amines, nucleotides, nitrifier activity (e.g., Cooke and White, 1987a,b; proteins, and refractory humic compounds of low Birmingham et al., 1994). Denitrification also occurs in nitrogen content). shallow riparian sediments connected to the stream by 4. The nitrogen cycle consists of a balance between the hyporheic zone (Duff and Triska, 1990). Denitrifi- nitrogen inputs to and nitrogen losses from an cation increases with greater distance (10-15 m) from . the stream channel were correlated with decreasing a. Sources of nitrogen include (i) nitrogen con- oxygen concentrations and increasing concentrations of tained in particulate "dry fallout" and precipi- reduced compounds. tation falling directly on the lake surface, (ii) ni- Photosynthetic oxygen production by attached trogen fixation both in the water and the algae can diffuse for several millimeters into the sediments, and (iii) inputs of nitrogen from sediments and reduce the rates of denitrification surface and groundwater drainage. (Christensen et al., 1990). Under shaded conditions or b. Losses of nitrogen occur by (i) outflow from the darkness, rates of denitrification increase rapidly as basin, (ii) reduction of NO3- to N2 by bacterial bacterial respiration depletes oxygen in the microzones. denitrification with loss of N2 to the atmos- Clearly, the processes of nitrification and denitrification phere, and (iii) sedimentation of inorganic and can be functioning simultaneously and reciprocally in organic nitrogen-containing compounds to the the sediments, particularly where many microzones of sediments. steep redox gradients occur. 5. Microbial fixation of molecular N2 in by bac- teria is a major source of nitrogen. In lakes and streams, N2 fixation by heterotrophic bacteria and VI11. SUMMARY certain cyanobacteria is quantitatively less signifi- cant, except under certain conditions of severe de- 1. Nitrogen, along with carbon, hydrogen, and phos- pletion of combined inorganic nitrogen com- phorus, is one of the major constituents of the pounds. cellular protoplasm of organisms. Nitrogen is a a: The N2 content of water is usually in equilib- major nutrient that affects the productivity of fresh rium with the N2 of the atmosphere during pe- waters. riods of turbulent mixing. In stratified, produc- 2. A major source of nitrogen of the origi- tive lakes, concentrations of N2 may decline in nates from fixation of atmospheric molecular ni- the because of reduced solubility as trogen (N2). The nitrogen cycle is a complex bio- temperatures rise and increase in the hy- chemical process in which nitrogen in various polimnion from denitrification of NO3-N. forms is altered by nitrogen fixation, assimilation, b. N2 fixation by cyanobacteria is usually much and reduction of nitrate to N2 by denitrification. greater than fixation by heterotrophic bacteria. For all practical purposes, the nitrogen cycle of In cyanobacteria, N2 fixation is light-dependent lakes is microbial in : Bacterial oxidation and usually coincides with the spatial and tem- and reduction of nitrogen compounds are coupled poral distribution of these microbes. NH4-N with photosynthetic assimilation and utilization by assimilation requires less energy expenditure algae, photosynthetic bacteria, and larger aquatic than NO3-N, and NOJ-N less than NrN. plants. The direct role of animals in the nitrogen N, fixation by cyanobacteria increases when cycle is certainly very small; under certain condi- NH4-N and NOJ-N concentrations decrease tions, however, their grazing activities can influ- in the trophogenic zone. 236 12. The Nitrogen Cycle

c. Bacterial N2 fixation in surrounding products, can dominate nitrogen loading. lakes or adjacent to streams can add significant b. Bacterial denitrification is the biochemical amounts of combined nitrogen to freshwater reduction of oxidized nitrogen anions (NO,- ecosystems. and NO2-) concomitant with the oxidation of 6. Ammonia is generated by heterotrophic bacteria as organic matter: NO, + NO2+ N20 + N2. the primary nitrogenous end product of decompo- Nitrous oxide (N20)is rapidly reduced to N, sition of proteins and other nitrogenous organic and has rarely been found in lakes in apprecia- compounds. Ammonia is present primarily as ble quantities. Denitrification is accomplished NH4+ions and is readily assimilated by plants in by many genera of facultative anaerobic bacte- the trophogenic zone. ria, which utilize nitrate as an exogenous termi- a. NH4-N concentrations are usually low in aer- nal hydrogen acceptor in the oxidation of obic waters because of utilization by plants in organic substrates. Denitrification occurs in the . Additionally, bacterial nitrifica- anaerobic environments, such as in the hy- tion occurs, in which NH4+is oxidized through polimnia of eutrophic lakes (Fig. 12-4) or in several intermediate compounds to NO2and anoxic sediments, where oxidizable substrates NO3-. are relatively abundant. b. When the of a eutrophic .lake be- 8. Dissolved organic nitrogen (DON)often consti- comes anaerobic, bacterial nitrification of am- tutes over 50% of the total soluble nitrogen in monia ceases. The oxidized microzone at the fresh waters. sediment-water interface is also lost, which re- a. Over half of the DON occurs as amino nitrogen duces the adsorptive capacity of the sediments compounds, mostly as polypeptides and com- for NH,-N. A marked increase in the release plex nitrogen compounds. of NH,+ from the sediments then occurs. As a b. The ratios of DON to particulate organic nitro- result, the NH4-N concentrations of the hy- gen (PON) of streams and lakes are usually polimnion increase (Fig. 12-4). from 5:l to 10:l. As lakes become more eu- c. Bacterial nitrification proceeds in two stages: (i) trophic, DON:PON ratios decrease. the oxidation of NH4+- NO2-, largely by Ni- 9. The distribution of nitrogen in a lake can change trosomonas but also by other bacteria, includ- rapidly. Examples of the depth-time distributions ing methane oxidizers; and (ii) the oxidation of the different forms of nitrogen demonstrate that NO,- -t NO3-, in which Nitrobacter is the as lakes become more productive from nutrient dominant bacterial genus involved. loading, concentrations of NO3-N and NH4-N d. Nitrite (NO2-)is readily oxidized and rarely ac- in the trophogenic zone can be severely reduced cumulates except in the metalimnion, upper hy- and depleted by photosynthetic assimilation. polimnion, or interstitial water of sediments of Cyanobacteria with the capability of nitrogen eutrophic lakes. Concentrations are usually very fixation may then come to dominate. In anaerobic low (<100 pg liter-') unless organic pollution hypolimnia, NO3-N is rapidly denitrified to Nz, is high. which is either fixed or lost to the atmosphere. 7. Nitrate is assimilated and aminated into organic NH,-N concentrations accumulate from decom- nitrogenous compounds within organisms. This position of organic matter and release of NH,-N organic nitrogen is bound and cycled in photo- from sediments under anaerobic conditions. synthetic and microbial organisms. During the 10. Organic carbon-to-nitrogen ratios (C:N) indicate normal metabolism of these organisms, and at an approximate state of resistance of complex death, much of their nitrogen is liberated as mixtures of organic compounds to decomposition, ammonia. Additionally, organisms release a variety because proteolytic metabolism by fungi and bac- of organic nitrogenous compounds that are resis- teria removes proportionally more nitrogen than tant to proteolytic deamination and ammonifica- carbon. Higher C:N ratios commonly occur in tion by heterotrophic bacteria to varying degrees. residual organic compounds, which are more resis- a. Nitrate (NO3-) is the common form of inor- tant to decomposition. ganic nitrogen entering fresh waters from the a. Organic materials from allochthonous and wet- in surface waters, ground water, land sources commonly have C:N ratios from and precipitation. In certain oligotrophic waters 45:l to 50:l and contain many humic com- in basaltic rock formations, nitrate loading pounds of low nitrogen content. from atmospheric sources, especially if contami- b. Autochthonous organic matter produced by the nated by human-produced combustion emission decomposition of tends to have higher protein content and C:N ratios of about 12:1. 13. In running waters, nitrogen is used repeatedly as it c. Alterations in in the protein-carbohydrate-lipid passes downstream. The rate of utilization and re- ratios alter the particulate C:N ratio as a result lease depends upon physical and biological reten- of phosphorus or nitrogen limitation. A C:N tiveness, largely by the microbiota attached to the ratio of >14.6 often indicates a severe nitrogen streambed. The average distance a nutrient atom deficiency in phytoplankton, between 8.3 and travels downstream during one cycle through the 14.6, a moderate deficiency, and < 8.3, no ni- water, biotic, and substrata compartments is re- trogen deficiency. ferred to as the spiralling length (Figs. 12-13 and 11. Much of the total nitrogen occurs in the sediments 12-14). in forms that are relatively unavailable for biotic 14. Nutrient limitations for biotic productivity are un- utilization. Of the readily available nitrogen, a ma- common in small rivers and streams, where nutri- jority occurs in soluble form in the water and in ents are efficiently retained and recycled. As rivers the interstitial water of surficial sediments (and in become larger, nutrient retention is less and nutri- littoral vegetation in shallow, productive lakes). ent limitations can become more prevalent. Turnover rates of NH,-N are rapid in water but 15. The processes of nitrogen cycling in the water of slower in the sediments. In contrast, NO,-N streams and rivers are similar to those of lakes and turnover is slower in the water than in sediments, are influenced to a large extent by bacterial, fun- where, under anoxic conditions in eutrophic lakes, gal, and other microbial metabolism. Attached NO,-N is rapidly denitrified to N2. bacteria, fungi, and algae are the primary organ- 12. Increased loading of inorganic nitrogen to rivers isms controlling the seasonal spatial and temporal and lakes frequently results from agricultural ac- variations within the water. The processes of nitri- tivities, sewage, and anthropogenic atmospheric fication and denitrification often function simulta- . In unproductive oligotrophic lakes, neously and reciprocally in running water sedi- phosphorus availability is often the principal limit- ments, where many microzones of steep redox ing nutrient for plant growth. As phosphorus load- gradients occur in the hyporheic zone of the ing to fresh waters increases and they become streambed. more productive, nitrogen becomes more impor- tant as a growth-limiting nutrient.