Information

5.15: Nitrogen Fixation - Biology

5.15: Nitrogen Fixation - Biology


We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

5.15: Nitrogen Fixation

N2 fixation and cycling in Alnus glutinosa, Betula pendula and Fagus sylvatica woodland exposed to free air CO2 enrichment

We measured the effect of elevated atmospheric CO2 on atmospheric nitrogen (N2) fixation in the tree species Alnus glutinosa growing in monoculture or in mixture with the non-N2-fixing tree species Betula pendula and Fagus sylvatica. We addressed the hypotheses that (1) N2 fixation in A. glutinosa will increase in response to increased atmospheric CO2 concentrations, when growing in monoculture, (2) the impact of elevated CO2 on N2 fixation in A. glutinosa is the same in mixture and in monoculture and (3) the impacts of elevated CO2 on N cycling will be evident by a decrease in leaf δ 15 N and by the soil-leaf enrichment factor (EF), and that these impacts will not differ between mixed and single species stands. Trees were grown in a forest plantation on former agricultural fields for four growing seasons, after which the trees were on average 3.8 m tall and canopy closure had occurred. Atmospheric CO2 concentrations were maintained at either ambient or elevated (by 200 ppm) concentrations using a free-air CO2 enrichment (FACE) system. Leaf δ 15 N was measured and used to estimate the amount (Ndfa) and proportion (%Ndfa) of N derived from atmospheric fixation. On average, 62% of the N in A. glutinosa leaves was from fixation. The %Ndfa and Ndfa for A. glutinosa trees in monoculture did not increase under elevated CO2, despite higher growth rates. However, N2 fixation did increase for trees growing in mixture, despite the absence of significant growth stimulation. There was evidence that fixed N2 was transferred from A. glutinosa to F. sylvatica and B. pendula, but no evidence that this affected their CO2 response. The results of this study show that N2 fixation in A. glutinosa may be higher in a future elevated CO2 world, but that this effect will only occur where the trees are growing in mixed species stands.

This is a preview of subscription content, access via your institution.


NITROGEN BIOGEOCHEMISTRY IN THE OLIGOHALINE ZONE OF A NEW ENGLAND ESTUARY.

Abstract. We investigated nitrogen cycling in the oligohaline zone (the low-salinity region where river water first enters the estuary) of the Parker River estuary in northeastern Massachusetts. We introduced an isotopic tracer ([N.sup.15]-[[NO.sub.3].sub.-]) for 27 days in August 1996 to help determine how watershed-derived nitrogen moves through the upper estuary. The amount of tracer added was sufficient to enrich nitrate isotopically by [tilde]100%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] in the vicinity of the addition but did not influence nitrate concentration appreciably. During typical summer low-flow conditions as occurred during the addition period, essentially all riverine nitrate (including the nitrate tracer) was rapidly removed from the water column by the planktonic diatom Actinocyclus normanii. Export of tracer down-estuary was low during the isotope addition period, in part because of low river discharge. Instead, most of the nitrogen originally assimilated by A. normanii was transferred to sediments in the oligohaline zone. Nitrogen dema nd by phytoplankton during summer exceeded riverine supply by an order of magnitude. The additional nitrogen came mainly from the regeneration of benthic nitrogen, although some may have come from groundwater. The whole-ecosystem isotope tracer approach applied here was a powerful means of investigating the fate of watershed-derived nitrogen in the upper estuary.

Key words: Actinocyclus normanii diatoms estuary land-margin ecosystem Massachusetts (USA) [N.sup.15] nitrogen cycling oligohaline zone phytoptankton silica stable isotope terrestrial- aquatic linkage.

Nitrogen inputs to watersheds in the northeastern United States have increased greatly since European settlement in the early 1600s. Primary sources of the added nitrogen are atmospheric deposition, agricultural fertilizers, and wastewater (Nixon and Pilson 1983, Lee and Olsen 1985, Valiela et al. 1997). Much of this additional nitrogen is retained or denitrified in the watershed, but a substantial amount enters groundwater and rivers and eventually is delivered to estuaries and the ocean (Galloway et al. 1995, Howarth et al. 1996).

The influence of increased nitrogen inputs may be particularly significant in estuaries. Because watershed-derived nutrients are transported to and focused on these ecosystems, estuaries are among the most highly fertilized ecosystems on earth (Nixon et al. 1986, Valiela et al. 1997). Productivity in estuaries is often limited by nitrogen, so increased nitrogen inputs can stimulate algal growth and in turn influence other aspects of ecosystem functioning (Hopkinson and Vallino 1995). For example, high nitrogen loading associated with spring runoff in Chesapeake Bay supports a phytoplankton bloom, and the recycling of the nitrogen retained in the bloom leads to a delayed productivity maximum in the summer (Malone et al. 1988, Harding 1994). Although increased algal production resulting from anthropogenic nitrogen loading may have beneficial consequences such as elevated fish production (Keller et al. 1990), it may also lead to hypoxia in bottom waters, which can be devastating to fish and other biota (Dauer e t al. 1992, Justic et al. 1993, Breitburg et al. 1994, Justic et al. 1995).

Because interactions of hydrology, geomorphology, and biology greatly complicate nitrogen cycling in estuaries, a complete understanding of the biogeochemistry of nitrogen in estuaries is lacking. Although much progress has been made, uncertainties in the details of nitrogen processing hamper our ability to predict the effects of increased nitrogen loading. These uncertainties highlight the necessity of further investigation of nitrogen dynamics in coastal ecosystems and suggest that new approaches may be needed.

The objective of this study was to investigate nitrogen cycling in the upper estuary, or oligohaline zone, of the Parker River estuary in northeastern Massachusetts (Fig. 1). The study consisted of a whole-ecosystem [N.sup.15]-[[NO.sub.3].sup.-] tracer addition designed to investigate (1) biogeochemical cycling of nitrogen and (2) trophic pathways in the upper estuary. This paper focuses on biogeochemical aspects of the study and a companion paper emphasizes food web structure (Hughes et al. 2000). We focused on the upper estuary because it is the region most closely connected to the watershed and therefore is the first reach to process riverine nitrogen inputs. Compared to more seaward reaches of estuaries, the oligohaline zone has received relatively little attention (Anderson 1986, Schuchardt et al. 1993). In addition to the biogeochemical significance of the oligohaline zone as a buffer or ecotone between the watershed and the lower-estuary and ocean, the low salinity waters of the upper estuary are critical t o the life histories of many estuarine organisms (Odum 1988, Deegan and Garritt 1997). Increased study of the upper estuary will yield greater understanding of the functioning of the estuary as a whole as well as facilitate effective management of these waters.

The Parker River estuary (Fig. 1) is part of the Plum Island Sound estuarine system (42[degrees]44' N, 70[degrees]50' W) in northeastern Massachusetts. The watershed of the Parker River estuary, 65 [km.sup.2] above the dam at the head of the estuary, is predominantly forested and has moderate residential development. The total length of the estuary is [tilde]24 km, with the oligohaline zone nominally occupying the initial 5 km. Our definition of the oligohaline zone or upper estuary is based upon considerations such as salinity distribution and species composition. Salinity in the upper estuary is generally [less than]10 (salinity given as a ratio [UNESCO 1985], conductivity [tilde]15 mS/cm) and biota include typically freshwater organisms as well as estuarine and marine species. Mean tidal amplitude in the upper estuary is [tilde]2.7 m and tidal excursion is 2-4 km. Mean annual freshwater input from the Parker River is 1.2 [m.sup.3]/S, but during the summer discharge is lower (generally 0.1-0.5 [m.sup.3]/s) .

The upper estuary has a single main channel that meanders through extensive fresh and salt marshes (Fig. 1). The surface area of the upper 5 km of the estuarine channel is [tilde]223 650 [m.sup.2] (J. Vallino, personal communication). Marsh vegetation consists mainly of cattail (Typha latifolia) and sedges (Scirpus americana and Carex sp.), with salt marsh cordgrass (Spartina alterniflora) occurring along creek banks. Primary production in the channel is predominantly by phytoplankton, with the pelagic diatom Actinocyclus normanii dominating primary production during summer blooms, and secondarily by microphytobenthos. The most abundant zooplankton are Eurytemora affinis and Acartia tonsa, and common fish include mummichog (Fundulus heteroclitus), white sucker (Catostomus commersoni), white perch (Morone americana), and Atlantic silverside (Menidia menidia). A more complete description of the biota of the oligohaline zone of the Parker River estuary is found in Hughes et al. (2000) and Deegan and Garritt (19 97).

An initial attempt at the experiment began on 10 July 1996 but was aborted after 4 d due to a large storm. The successful [N.sup.15]-[[NO.sub.3].sup.-] addition began 6 August 1996 and continued until 1 September 1996. The isotope was added continuously via peristaltic pump except for an [tilde]24-h interruption (26-27 August) when the pump failed. The isotopic tracer was in the form of [N.sup.15]-enriched [KNO.sub.3] and was added at the rate of 4.8 g [N.sup.15]/d (128 g [N.sup.15] total). Rhodamine WT, a fluorescent dye, was added with the isotope in order to allow comparison of the behavior of relatively conservative (rhodamine) and reactive (nitrate) solutes. The rhodamine/isotope solution was added 2 km seaward from a small dam that defines the head of the estuary (Fig. 1). The volume of water into which the isotope was added, i.e., the "tidal prism" at the isotope addition site, is [tilde]130 000 [m.sup.3]. Sampling began before the isotope addition to establish baseline nitrogen stocks and isotopic va lues and continued for [tilde]1 mo after the termination of the [N.sub.15] addition.

River inputs to the upper estuary were determined using United States Geological Survey (USGS) data from a gauging station on the Parker River [tilde]2 km upstream from the head of the estuary (USGS station number 01101000). We accounted for inputs from the ungauged portion of the watershed between the dam and gauge by assuming a constant area-to-runoff ratio. Nitrate, ammonium, and silica inputs to the upper estuary were assessed by periodically sampling river water flowing over the dam. When ammonium and nitrate concentrations were not measured simultaneously, dissolved inorganic nitrogen (DIN) flux was calculated by interpolating ammonium concentrations between consecutive samples or assuming that ammonium concentration was constant at 1.2 [micro]mol/L after 19 September.

Sampling and analytical procedures

Water-column constituents were sampled at 1-km intervals along a 10-km longitudinal transect beginning at the head of the estuary. Sample locations are designated by distance in kilometers along the river's course, beginning just below the dam (0 km) and increasing in the down-estuary direction. Over the 2-mo period of the study, 15 sampling transects were conducted. Samples for nutrient concentrations, particulate organic nitrogen (PON) concentrations and isotopic enrichment, and chlorophyll a (chla) were collected at each sampling station, whereas other types of samples (phytoplankton, [N.sup.15]-[[NO.sub.3].sup.-], and [N.sup.15]-[[NH.sub.4].sup.+] were collected less frequently. In addition to these samples, temperature and conductivity, and periodically rhodamine concentration, were recorded at each sampling station. Hydrologic residence time in the upper 5 km of the estuary was calculated from river discharge (Vallino and Hopkinson 1998).

Dissolved oxygen (DO), depth, conductivity, and temperature were measured and logged at 15-min intervals at the isotope addition site using an ISCO/YSI autosampler/data logger. These variables were recorded throughout the 2-mo period of the study, except for a 19-h period 2-3 August and a 4-d period 23-27 August when data logging failed.

Transect water samples were collected using a battery-powered peristaltic pump fitted with an inline GF/F filter. Samples were stored on ice until returned to the laboratory. Chlorophyll and PON were collected on glass fiber filters (Whatman GF/F, 47 and 25 mm diameter, respectively). Phytoplankton were collected by towing a net (25-cm diameter, 20-[micro]m mesh opening) for 2 min behind a small boat.

Nutrients were analyzed in a temporary laboratory at Governor Dummer Academy, Byfield, Massachusetts, 1 km from the study site. All nutrient samples were analyzed for nitrate and selected samples were also analyzed for ammonium, DON, and silica. Nitrate was analyzed by chemiluminescence detection after reducing nitrate to NO using vanadium chloride (Garside 1982, Braman and Hendrix 1989). Ammonium was measured manually (Solorzano 1969), DON was determined after UV-oxidation (Walsh 1989), and chla was analyzed colorimetrically (Strickland and Parsons 1972). Dissolved silica was analyzed at the Horn Point Environmental Laboratory of the University of Maryland, using the molybdate blue method (Strickland and Parsons 1972) modified for use with an autoanalyzer. PON concentration was calculated by mass spectrometric determination of nitrogen content on GFIF filters.

Nitrogen isotopic compositions of ammonium and nitrate were measured using modifications of the ammonia diffusion procedure (Sigman et al. 1997, Holmes et al. 1998). To estimate the isotopic composition of DON, we measured [delta][N.sup.15]-TDN and calculated [delta][N.sup.15]-DON using measured values (see next paragraph) and concentrations of nitrate and ammonium. PON isotopic composition was determined by combustion of the entire 25-mm filter. Phytoplankton were separated from detritus prior to isotopic analysis using a combination of procedures including differential filtration, centrifugation, and manual removal of detritus.

Nitrogen isotopic compositions were measured at the Stable Isotope Laboratory of the Marine Biological Laboratory. Isotopic compositions are expressed in standard [delta] notation, where [delta][N.sup.15] = [([R.sub.SA]/[R.sub.ST]) - 1] X [10.sup.3], R = [N.sup.15]/[N.sup.14], and results are expressed as %[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] deviation of the sample (SA) from the standard (ST), [N.sub.2] in atmospheric air ([delta][[N.sup.15].sub.AIR] = 0%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]). In this paper, [delta] values have been standardized by subtracting the natural abundance [delta][N.sup.15] value of similar samples collected prior to the isotope addition, unless otherwise noted. Therefore, the standardized values ([delta][[N.sup.15].sub.t], with the subscripted "t" referring to "tracer") reflect tracer content.

To evaluate the fate of the added tracer, we estimated the standing stocks of tracer-nitrogen in the upper 5 km of the estuary for the major components of the ecosystem that became enriched in [N.sup.15]. Estimates were made for nitrate, PON, zooplankton, fishes (alewife, white sucker, mummichog), grass shrimp, benthic consumers (grouped class including mud crabs, amphipods, oligochaetes, polychaetes, ostracods, and other benthic organisms), and benthic sediments. Estimates of isotopic enrichment of biota were made during the period of maximum enrichment and are reach (0-5 kin) averages. Tracer enrichment in sediments was estimated from surficial sediment scrapes and isotopic enrichment of sediment cores. Surficial sediment samples were obtained by scraping the upper 2-3 mm of sediment from 50-100 [cm.sup.2] of intertidal mud banks or subtidal grab samples, using a spatula. Sediment scrapes were obtained before and at the end of the tracer addition.

The standing stock of tracer-nitrogen in PON was examined in detail to help us discern the fate of tracer once assimilated by phytoplankton. We use the isotopic enrichment of PON as a proxy for phytoplankton (using appropriate scaling functions) because our data set for [[N.sup.15].sub.t]-PON is far more extensive than for [[N.sup.15].sub.t]-phytoplankton. Comparison of the standing stock of tracer N in PON to the cumulative amount of tracer that has entered PON allows us to calculate, by mass balance, the rate of loss of [[N.sup.15].sub.t]-PON from the plankton.

To estimate the advective flux of tracer down-estuary, we calculated the mean concentration and [[N.sup.15].sub.t] of the various components of the water column at the reach boundary (5 km), and multiplied the product of these values by river discharge. We doubled advective flux to estimate total flux (advective plus dispersive). In general, this protocol somewhat overestimates down-estuarine transport of tracer, since at river discharges greater than 0.1 [m.sup.3]/s, advective flux exceeds dispersive flux at the 5 km location of the upper estuary (J. Vallino, personal communication).

Discharge of the freshwater Parker River varied from [less than]0.1 [m.sup.3]/s to [greater than]2.5 [m.sup.3]/s from i August to 1 October 1996 (Fig. 2A). Over the course of the isotope addition (6 August-1 September), river discharge decreased gradually from 0.26 to 0.03 [m.sup.3]/s. Nitrate concentration in the freshwater Parker River was inversely related to river discharge and ranged from [tilde]4 to 16 [micro]mol/L, whereas ammonium concentration was always [less than]4 [micro]mol/L (Fig. 2B). A more extensive, multiyear data set collected as part of the Parker River/Plum Island Sound Land Margin Ecosystem Research (LMER, now LTER) [5] project also found ammonium to be consistently [less than]4 [micro]mol/L. Riverine DIN flux over the 2-mo period ranged from 86 to 1262 mol/d (Fig. 2C). Concentrations of dissolved silica in river water during summer 1996 ranged from 107 to 196 [micro]mol/L (n = 4, mean = 129.5 [micro]mol/L). Assuming a mean DIN concentration of 13.9 [micro]mol/L, the average Si:DIN ratio in ri ver water was 9.3:1.

Residence time, depth, conductivity, and dissolved oxygen at the tracer addition site

Hydrologic residence time in the upper estuary varied between 1 and 16 d during the 2-mo period of the study, and averaged [tilde]12 d during the isotope addition period (Fig. 3A). Hydrologic residence time peaked in late August when river discharge was lowest, but quickly dropped to 1 d after a storm in mid-September.

The record of water-column depth shows that marsh flooding was infrequent during the isotope addition period (Fig. 3B). Water-column depth also illustrates the asymmetry of tides in the upper estuary, both in terms of height and timing of tides. Consecutive tides varied in height, often by as much as 20-30 cm. From 1 August to 1 September 1996, the mean time from high to low tide was 6 h, 52 min whereas flood tides averaged 5 h, 33 min. Over this same time period, mean tidal amplitude was 2.68 m, the range 2.08-3.13 m. The small tidal amplitude of 2.08 m on 18 September was associated with a large storm (Fig. 3B).

River discharge, and tidal stage and amplitude, control salt content (electrical conductivity) of water at the isotope addition site (Fig. 3C). Throughout most of August, water present at the isotope addition site at low tide was essentially freshwater, whereas conductivity at high tide was more variable. At the beginning of August, high tide conductivity reached [tilde]10 mS/cm, but dropped for the next several days as tidal amplitudes decreased and less ocean water was transported in the up-estuary direction. By the end of August, the combination of large tides and low freshwater inputs (long hydrologic residence time) led to the highest conductivity of the study period, [tilde]22 mS/cm (almost 50% seawater). A small storm in the first week of September somewhat shortened residence time and pushed some salt out of the upper estuary, and the large storm of 18 September almost completely replaced the upper estuarine water mass with freshwater.

Dissolved oxygen concentration in water at the isotope addition site varied from [tilde]4 to 11 mg/L (Fig. 3D). During most of the period of the [N.sup.15] addition, there was a single large DO peak each afternoon, corresponding to the period of maximum photosynthetic oxygen production. However, from 1 to 4 August, 27 August until [tilde]7 September, and 18 September until the end of the study, diel DO swings were greatly attenuated and instead of one large midafternoon maximum, there were two smaller DO peaks each day. These peaks were independent of time of day but were instead associated with low tides. The reduced diel DO swings at the beginning of August and end of September were associated with hydrologic residence times in the upper estuary of [less than]6 d. From late August until early September, attenuated diel DO swings were correlated with marsh flooding and the heavy cloud cover of Hurricane Eduardo.

Spatial distributions of nutrients, chlorophyll, and PON in the upper estuary

Transect samples are summarized using data from four representative periods: pre-addition (30 July to 4 August), mid-addition (13 August), late addition (1 September), and post-addition (12 September). Nitrate generally entered the estuary at concentrations of 5-15 [micro]mol/L but declined to [less than]1 [micro]mol/L within the first few kilometers of the estuary (Fig. 4). Farther down the estuary, generally at 5-7 km from the dam, nitrate concentration frequently increased to 3-5 [micro]mol/L, as is apparent on 13 August and 12 September. Over the course of the experiment, nitrate concentration at the isotope addition site was [less than]1 [micro]mol/L (Fig. 5). Ammonium concentration was generally [less than]2 [micro]mol/L in the upper few kilometers of the estuary, but increased down-estuary (Fig. 4). Ammonium data are not available for 1 and 12 September, but data for four additional transects taken during the addition show a pattern similar to what was observed on 13 August.

PON concentration generally was highest (maximum [tilde]50 [micro]mol/L at the upper end of the estuary, and spatially and temporally closely followed chla (Fig. 4). Chlorophyll a generally peaked at 20-100 [micro]g/L and declined to [tilde]5 [micro]g/L at the downstream end of the study reach (Fig. 4).

Isotopic composition of nitrate, ammonium, PON, and phytoplankton

Substantial tracer enrichment was measured in nitrate but not in ammonium. Natural abundance levels were similar for ammonium and nitrate prior to the experimental addition [delta][N.sup.15]-[[NO.sub.3].sup.-] ranged from 0.4 to 4.4%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (26 July, n = 3 samples) and [delta][N.sup.15]-[NH.sub.4].sup.+] ranged from -1.4%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] to 2.1%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4 August, n = 3 samples). To avoid taking an "unmixed" sample of the isotope/rhodamine solution near the point of tracer addition, samples collected at 2 km were always collected on the "up-current" side of the tracer addition point. Consequently, we did not observe extremely high [delta][N.sup.15].sub.t]-[NO.sub.3].sup.-] values at the 2 km station as might be expected if we had sampled immediately downcurrent of the dripper. During ebbing tides (6 and 19 August), high [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] concentration was apparent at the sampling station immediately seaward of the isotope addition site (Fig. 6A). In addition, low-level tracer enrichment of [tilde]6%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] was found at the uppermost sampling station on 19 August. During the flood tide of 13 August, we were only able to analyze the isotopic content of nitrate in samples seaward of the isotope addition site because the remaining samples had nitrate concentrations too low for analysis, and none of those samples that we were able to analyze had substantial [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] enrichment. For the flood tide of 21 August, [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] was enriched at all three locations where nitrate concentrations were sufficient for analysis. [N.sup.15]-[[NH.sub.4].sup.+] samples collected almost 2 wk after the tracer addition began showed little or no enrichment in the upper 4 km of the estuary, but apparent tracer enrichment reached 9.5%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] at 5 km and remained elevated in the lower 5 km of the study site (Fig. 6B). However, we are not certain if this enrichment reflects our tracer addition or if natural processes such as fractionation during nitrification are responsible.

Natural [delta][N.sup.15]-PON abundance prior to the tracer addition (samples collected 4 August) decreased from 5.4%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] at the head of the estuary to 2.3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] at 10 km below the dam. To determine enrichment due to tracer ([delta][[N.sup.15].sub.t]-PON), we assume a natural abundance of 2.3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] for [delta][N.sup.15]-PON. This may lead to a 2-3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] overestimation of [delta][[N.sup.15].sub.t]-PON at the head of the estuary, but this will not affect results and conclusions to any great degree given the relatively large label in PON.

The added [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] was rapidly incorporated into PON (Fig. 7). By 10 h after the start of the addition, [delta] [[N.sup.15].sub.t]-PON had reached 11%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] and increased to 19%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] after another 10 h. Within 54 h of the start of the addition, [delta] [[N.sup.15].sub.t]-PON reached a peak value of 61%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. In general, over the next three weeks, [delta] [[N.sup.15].sub.t]-PON was [greater than]30%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] in the upper 3-4 km of the estuary and dropped below 15%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] enrichment by 5 km. Following the termination of the tracer addition on 1 September, [delta] [[N.sup.15].sub.t]-PON rapidly declined, although some enrichment was evident in the upper estuary until at least 20 September.

Our most extensive set of transect samples was collected on 19 August 1996, and we present these data to facilitate comparison among the various samples that were collected simultaneously. Sampling began at the 10-km station during the final hour of the flood tide and was completed [tilde]2 h into the ebbing tide. Water temperature was relatively constant ([tilde]25[degrees]C) throughout the reach, but conductivity ranged from that of freshwater at the uppermost sampling location to [tilde]75% seawater content at 10 km (Fig. 8A). A bloom (predominantly A. normanii) was present in the upper estuary and chla peaked at [tilde]30 [micro]g/L at the 1-km sampling location (Fig. 8B). Dissolved silica entered the estuary at 196 [micro]mol/L. but its concentration dropped to [less than]2 [micro]mol/L in the first kilometer of the estuary, with the location of the silica minimum corresponding to the chla maximum. The majority of total nitrogen in the upper estuary was DON, with PON generally being the second most abundant form (Fig. 8C). Tot al nitrogen (PON, DON, DIN) concentration declined from [tilde]70 [micro]mol/L in the freshwater Parker River to [less than]25 [micro]mol/L at 10 km. Inorganic nitrogen concentrations were very low in the vicinity of the phytoplankton bloom but increased at the 5- to 10-km stations.

Three spatial patterns of tracer enrichment were apparent, exemplified by (1) nitrate, (2) phytoplankton and PON, and (3) ammonium (Fig. 8). The distribution of rhodamine demonstrates how [delta] [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] would be distributed if no biological uptake or regeneration occurred. [delta] [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] showed a sharp peak at the 3-km sampling location but relatively little or no enrichment elsewhere (Fig. 8D). Comparison to rhodamine indicates that [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] did not behave conservatively (Fig. 8D) a similar conclusion is indicated by the rapid decline in nitrate concentration in the uppermost kilometer of the estuary (Fig. 8C). Tracer enrichment of both phytoplankton and PON peaked at 1 km and declined rapidly in both up- and down-estuary directions (Fig. 8E). On 19 August, purified phytoplankton samples reached a peak tracer enrichment of [tilde]43% greater than that of PON. Ammonium shows little or no enrichment in the upper 4 km of the estuary, although s ome enrichment may have been present in lower reaches of the estuary (Fig. 8D). [delta][N.sup.15]-TDN reached 11.2%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] at 3 km (data not shown), but the calculation of [delta][[N.sup.15].sub.t]-DON accounting for ammonium and nitrate delta values and concentrations showed that [delta][[N.sup.15].sub.t]-DON was [less than]3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

At the end of the isotope addition period, most of the tracer was present in oligohaline sediments (Table 1, Fig. 9). The largest water-column stock of tracer was PON, and the standing stock of tracer in benthic consumers was about the same as contained in PON. Zoo-plankton, shrimp, and fish were relatively unimportant with respect to storage of tracer.

During the isotope addition period, PON concentration was generally 20-30 [micro]mol/L (Fig. 10A) and mean reach-averaged [delta][[N.sup.15].sub.t]-PON (excluding the first 2 d of the addition) was 29.3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (Fig. 10B). For comparison, the mean reach-averaged [delta][[N.sup.15].sub.t]-phytoplankton was 49.3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. On all sampling dates except one, the mass of tracer in PON was less than or equal to the amount of [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] added each day (Fig. 10C). The exception was 8 August, when the mass of tracer in PON reached [greater than]9 g. Except for the first few days of the addition, only a small fraction of the cumulative mass of tracer [N.sup.15] that had been added was present in PON (Fig. 10D). Export of [[N.sup.15].sub.t]-PON down-estuary was [less than]6 g during the isotope addition period, suggesting that PON rapidly sedimented in the upper estuary.

The experimental addition of [N.sup.15]-[[NO.sub.3].sub.-] to the Parker River estuary yielded several new insights about the functioning of the ecosystem. First, we were able to show that the majority of riverine nitrate entering the estuary was initially processed by the planktonic diatoms (mainly Actinocyclus normanii) that form the base of the productive oligohaline food web. Moreover, the magnitude and spatial distribution of isotopically enriched nitrate demonstrated that nitrate turnover was very rapid, even after riverine nitrate had initially been drawn down. Nitrogen demand by phytoplankton greatly exceeded riverine supply, and benthic nitrogen fluxes made up most of the difference. Without the evidence detailed below that the additional nitrogen was not isotopically enriched, we would have overestimated the significance of recycling in the water column. We were also able to investigate the origin of nitrogen exported from the upper estuary as DIN, DON, and PON. Although most riverine nitrate (and the nit rate tracer) was retained in the upper estuary during the summer, the bulk of the tracer that was exported was in the form of PON. The isotope addition experiment aided us in assessing which compartments in the upper estuary were most important for storing riverine nitrate-N, and benthic sediments were the primary storage zone. Finally, our understanding of the trophic structure of the ecosystem was greatly advanced by following the tracer through the food web (Hughes et al. 2000).

Initial processing of the tracer

Although riverine nitrate might follow numerous pathways through the estuarine ecosystem, its initial fate and that of the [N.sup.15] tracer under typical summer low-flow conditions in the upper Parker River estuary was assimilation by phytoplankton. Other pathways, including assimilation by benthic diatoms, macroalgae, bacteria, and marsh vegetation, and loss through denitrification and export to lower reaches of the estuary, were relatively minor over the time scale of this experiment.

The isotopic enrichment of both phytoplankton (Fig. 4) and bulk PON (Fig. 7) indicate that the pelagic diatoms quickly incorporated the nitrate tracer. Three days into the isotope addition, we were able to account for [tilde]75% of the added tracer in PON (Fig. 10D). In contrast, other measured pools had insignificant tracer accumulation this soon after the tracer addition began. Since the turnover time of the diatoms was [tilde]1 d, by the third day of the addition phytoplankton would already have lost a substantial amount of tracer. Therefore, even more than 75% of the [N.sup.15]-[[NO.sub.3].sup.-] tracer must have been initially assimilated by phytoplankton.

The rapid decline in silica concentration (Fig. 8B) indicates a heavy demand for silica (mean silica concentration decline in the upper 1 km of the estuary was 182 [micro]mol/L on 19 August and 27 August). Using conductivity to assess mixing between river and ocean water, we find that only [tilde]10-20 [micro]mol/L of the decrease is due to mixing with low-silica ocean water. Since DIN concentration in the freshwater Parker River was [tilde]16 [micro]mol/L when the silica samples were collected, and assuming the diatoms have a Si:N molar ratio of 1:1 (Redfield et al. 1963), at least 146 [micro]mol/L N in addition to riverine DIN inputs was required by diatoms in order to account for the observed silica depletion. Therefore, only [tilde]10% of planktonic nitrogen demand would have been met by direct utilization of riverine DIN inputs.

Stoichiometric assessment of nitrogen demand by phytoplankton, based on chla concentration, also indicates DIN demand far in excess of river delivery. During the isotope addition, the mean chla concentration in water passing by the addition site with each tide was 34.8 [micro]g/L. Assuming a turnover time of nitrogen in phytoplankton of 1 d (Eppley 1972), a C-to-chla mass ratio of 50:1 (Antia et al. 1963, Eppley 1972), and C:N molar ratio in phytoplankton of 7:1 (Redfield 1958), nitrogen demand by phytoplankton in the 130 000 [m.sup.3] tidal volume was 2693 mol/d. Mean DIN flux over the dam during the isotope addition was 240 mol/d (Fig. 2C), suggesting that only [tilde]9% of phytoplankton N demand was met by direct uptake of watershed-derived DIN.

A final estimate of DIN demand by phytoplankton is based on GPP. During the A. normanii bloom in the upper estuary, GPP averages [tilde]2 g C [m.sup.-2]*[d.sup.-1] (J. Vallino, personal communication). If we assume that net primary production (NPP) was half of GPP (Peterson 1980) and the C:N molar ratio in phytoplankton was 7:1, we calculate DIN demand by phytoplankton to be 2282 mol/d only [tilde]10% of which could be met by direct utilization of riverine DIN.

Although the preceding estimates indicate that direct uptake of riverine DIN inputs meets only [tilde]10% of nitrogen demand by phytoplankton, riverine DIN in-puts during the summer could be sufficient to meet phytoplankton demands if diatom N was rapidly recycled in the water column independently from silica. If the diatoms' siliceous tests dissolved only slowly whereas nitrogen was quickly recycled, silica concentration could be drawn down to a greater extent than would be possible if nitrogen and silica were cycling at the same rates (Schelske and Stoermer 1971, Schelske et al. 1983, Doering et al. 1989, Conley et al. 1993). Many such cycles could be completed within the typical residence time of diatoms in the upper estuary, and silica could be depleted through the rapid recycling of riverine DIN. Without the isotopic tracer, it would have been difficult to refute this hypothesis. However, if this hypothesis were true, we would predict that the isotopic enrichment of ammonium and/or nitrate would closely mimic the enrichment of phytoplankton any DIN present in the vicinity of the diatom bloom would have been recycled from the isotopically enriched diatoms. The isotope data do not support this hypothesis (Figs. 5, and 8D, E). Instead, ammonium showed little or no enrichment in the upper estuary, and nitrate was substantially enriched only at the sampling locations that had recently passed by the isotope addition site. The sharp peak in [delta] [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] indicates that nitrate was being assimilated very quickly and that the nitrate being resupplied was not isotopically enriched. Therefore, rapid recycling of diatom-N and slower dissolution of silica tests does not account for the large silica depletion. Instead, diatom-N along with tracer-N must have been efficiently and rapidly removed from the water column.

A further indication that planktonic diatoms were the major contributors to whole-ecosystem metabolism and nitrogen uptake in the upper estuary is the record of dissolved oxygen concentration at the isotope addition site, which appears to be related to phytoplankton abundance. When chla was high after 5 August, diel DO variation was pronounced (Fig. 3D). In contrast, prior to 4 August and after 18 September, chla concentrations in the upper estuary were low ([less than]10 [micro]g/L) and diel variation in DO was greatly attenuated. In addition, when the diatom bloom was flushed from the upper estuary earlier in the summer during a high discharge event, diel DO swings and nitrate depletion in the upper estuary were also greatly reduced (R. M. Holmes, unpublished data).

Even with the apparent correlation between phytoplankton abundance and water-column DO, it is possible that primary producers other than phytoplankton played an important role in whole-system metabolism and DIN uptake. For example, benthic diatoms were present on intertidal mudbanks. However, isotopic enrichment of benthic diatoms lagged behind phytoplankton by 2 wk (Hughes et al. 2000). Given this substantial delay, they could not have been important in the initial removal of the [N.sup.15]-[[NO.sub.3].sup.-] tracer from the water column and instead probably acquired their label after remineralization of enriched, deposited organic matter released [[N.sup.15].sub.t]-[[NH.sub.4].sup.+] into surficial sediment porewaters. This conclusion is consistent with the nearly complete uptake of the tracer by PON (phytoplankton) in the water column. Other primary producers in the ecosystem can be discounted because they either were not present in significant quantities (e.g., macroalgae) or were not exposed to the isotope until the marsh flooded at the end of the experiment (e.g., Typha).

The remaining processes that could account for the rapid depletion of riverine nitrate are denitrification and microbial assimilation. Denitrification removes a substantial portion of incoming nitrogen in many estuaries (Seitzinger 1988), with the percentage removed being related to hydrologic residence time (Nixon et al. 1996). However, two factors argue against denitrification being responsible for the rapid initial depletion of watershed-derived nitrate in the upper Parker River estuary. First, nitrate concentration in the overlying water column is generally low, so diffusive flux of nitrate into anoxic sediments, the probable site of denitrification, is also low. Secondly, the small area and short time over which enriched [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] is present in the water column limits the contact period between [[N.sup.15].sub.t]-[[NO.sub.3].sup.-] and sediments. These arguments are consistent with the observed uptake of the majority of the tracer by phytoplankton over the first few days of t he addition. Therefore, we conclude that direct removal of water-column nitrate, and tracer-nitrate, via sediment denitrification was small. Although not relevant to the initial removal of [[N.sup.15].sub.t]-[[NO.sub.3].sup.-], it is still possible that the ultimate fate of much of the label was denitrification, following assimilation by phytoplankton, sedimentation, remineralization, and nitrification.

Maximum observed tracer enrichment in planktonic bacteria was [tilde]15%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (M. Hullar and D. Pakulski, personal communication), much less than for phytoplankton. Based on estimates of bacterial production in the Parker River estuary (Wright et al. 1987), we calculate that nitrate uptake by bacteria was [less than]5% of uptake by phytoplankton. Therefore, bacteria were not a major sink for the nitrate tracer.

Control of the phytoplankton bloom

In the upper Parker River estuary, the tight relationship between river discharge and hydrologic residence time highlights the overall importance of hydrology in controlling phytoplankton dynamics in the oligohaline zone. Over the 2-mo period of the study, the influence of river discharge on phytoplankton is most clearly illustrated by the storm of 18 September. Prior to the storm, river discharge was low ([less than]0.25 [m.sup.3]/s) and hydrologic residence time in the upper estuary exceeded 10 d. Doubling time for planktonic diatoms is typically 1-2 d (Eppley 1972), so there was sufficient time for the diatoms to complete several generations during their residence time in the oligohaline zone, which is consistent with the observed presence of the chlorophyll maximum. The mid-September storm, however, increased river discharge to [tilde]2.5 [m.sup.3]/s and hydrologic residence time declined to [less than]2 d (Figs. 2A and 3A). The DO record (Fig. 3D), and nitrate and chlorophyll concentrations in the upper estuary on 20 September (data not shown), indicate that the diatom bloom was flushed down the estuary. Additional data from earlier in the summer suggest that the diatom bloom only developed when freshwater inputs were less than [tilde]0.5 [m.sup.3]/s, which translates to a hydrologic residence time of [greater than or equal to]5 d. Long-term discharge data show that these conditions are met [tilde]150 d/yr, roughly mid-June until October. Therefore, on average, bloom formation is feasible during summer and early fall but not other seasons, and thus phytoplankton in the upper estuary would not be expected to transform riverine nitrate during the spring runoff season.

What limits the diatom bloom when hydrologic residence time in the upper estuary exceeds 5 d? Possibilities include biotic control by grazers, light limitation, nutrient limitation, and settling of cells to benthic sediments. The most likely possibility seems to be nutrient limitation, by either nitrogen or silica. Phosphorus limitation of phytoplankton growth is another possibility, but dissolved inorganic N:P (molar ratio) in the upper estuary in summer is generally [less than]5 (see the Parker River/Plum Island Sound Land Margin Ecosystem Research [LMER, now LTER] web site), [6] suggesting abundant P relative to N. Both nitrate and silica enter the estuary at relatively high concentrations but are rapidly drawn down to potentially limiting levels in the vicinity of the chlorophyll peak (Figs. 4 and 8). Since silica can only limit diatoms, other forms of phytoplankton might dominate if nitrogen is regenerated more rapidly.

Since DIN in excess of that delivered by the freshwater Parker River is required to account for the large silica decline, regeneration of benthic nitrogen, groundwater inputs, or dispersion from the ocean or lower estuary must provide the additional nitrogen required by phytoplankton. Groundwater inputs, although not well characterized in this system, are probably not substantial, as freshwater inputs over the dam are sufficient to model solute transport in the upper estuary (Vallino and Hopkinson 1998). Benthic sediments are rich in N, and measured fluxes, while variable, are often large (Hopkinson et al., in press). We therefore conclude that regeneration of nitrogen contained in upper estuarine sediments is the major source of additional DIN required by oligohaline diatoms, but that groundwater may make some contribution. The sum of these sources is about an order of magnitude greater that the flux of DIN coming over the dam when the bloom is present. While we view the oligohaline reach as a relatively op en ecosystem, most of the nitrogen demand by primary producers in summer and early fall is met through recycling from benthic sediments.

Export of tracer nitrogen could occur in any of the nitrogen-containing components of the water column or via migration of biota down-estuary. Since the standing stock of tracer in biota was low (Table 1), it is unlikely that migration or advection of tracer in fauna was significant. Moreover, essentially no isotopically enriched nitrate was exported from the upper estuary because nitrate at the 5 km station was not measurably enriched with tracer (Fig. 6). PON and DON were the remaining vectors of tracer export down-estuary. Advective plus dispersive transport of [[N.sup.15].sub.t]-PON out of the upper estuary during the study period was [less than]6 g (out of 128 g [N.sup.15] [[N.sub.3].sup.-] added). Furthermore, since DON concentration was similar to PON but its isotopic enrichment was much less ([less than]3%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] vs. 30%[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]), an insignificant quantity of tracer was exported from the reach as DON. Consequently, it appears that the relatively long hydrologic residence time in the upper estuary during the isotope a ddition resulted in little downsystem transport of tracer in any form, presumably because of fallout of PON (including diatoms) to the sediments, where most tracer was found (Table 1).

Comparison of the standing stock of tracer in PON in the water column to the flux of tracer into PON supports the hypothesis that most [[N.sup.15].sub.t]-PON was transported to sediments. Over the first few days of the addition, we can account for most of the [N.sup.15] tracer in PON (Fig.10D). However, the standing stock of tracer in the tidal volume quickly reaches a plateau and is relatively low, generally [less than]5 g [[N.sup.15].sub.t], (Fig. 10C), indicating that the turnover time of PON is short ([less than]1 d). Since relatively little tracer nitrogen is sequestered in consumers (Table 1) or exported down-estuary, it appears that most nitrogen in PON is transferred to benthic sediments (Fig. 11).

Oligohaline zones of estuaries are regions of rapid change in many physical, chemical, and biological variables (Morris et al. 1978, Anderson 1986, Schuchardt and Schirmer 1991, Fichez et al. 1992, Rehbehn et al. 1993, Schuchardt et al. 1993). Processes occurring at this interface of river and estuary influence the timing, magnitude, and form of material and energy transported down the estuary and to the ocean. In the Parker River estuary during summer, many important transformations of nutrients and organic matter occur in the upper 2-4 km of the estuary. Similar observations have been reported for other estuaries including San Francisco Bay (Alpine and Cloern 1992, Cloern 1996), subestuaries of Chesapeake Bay (Anderson 1986), the North River estuary in Massachusetts (Bowden et al, 1991), and several European systems (Schuchardt and Schirmer 1991, Rehbehn et al. 1993, Schuchardt et al. 1993, Sanders et al. 1997). The tracer addition technique used in this study is complementary to other approaches for studyi ng nitrogen cycling in estuaries and facilitates a detailed understanding of nitrogen cycling in these complex ecosystems. A particular strength of the whole-ecosystem tracer approach is that it allows a simultaneous examination of transport and processing of nitrogen in an intact ecosystem, something that is impossible to achieve in traditional bottle or mesocosm experiments.

Of the 128 g [N.sup.15] added to the estuary over 27 d, only [tilde]5 g was present in phytoplankton at any given time (Fig. 11A). A similar amount of tracer was sequestered in benthic consumers. Other biota contained far less tracer (Table 1), although they may have processed a significant amount of the tracer due to their rapid N turnover via processes such as zooplankton grazing. Both riverine DIN and the nitrate tracer were initially assimilated by phytoplankton, and then transported to sediments as phytoplankton and fecal pellets settled out of the water column (Fig. 11). Since the standing stock of tracer in PON quickly reached a plateau (Fig. 10C), the flux of tracer to sediments was almost as large as the flux into PON (Fig. 11A) because transport down the estuary was minimal. Given the large nitrogen pool and long residence time of nitrogen in benthic sediments, it appears that over the course of the study little tracer escaped sediments following settling of phytoplankton. Although there is uncertai nty in our estimate of the amount of tracer stored in benthic sediments, it is clear that sediments were the primary [[N.sup.15].sub.t] storage zone (Table 1). Some of this nitrogen might have been shunted to the atmosphere as [N.sub.2] via coupled assimilation-mineralization-nitrification-denitrification, but we were unable to quantify this flux because of the high standing stock of [N.sub.2] in the water column and its rapid exchange with the atmosphere, which made [[N.sup.15].sub.t]-enrichment of the [N.sub.2] pool below our limit of detection. Thus, the flux of nitrogen out of the oligohaline zone through denitrification is a major unknown that requires further investigation.

The flow of tracer within the upper estuarine ecosystem (Fig. 11A) illuminated the bulk movement of nitrogen through the system (Fig. 11B). We found that nitrogen demand by phytoplankton during summer far outweighed direct supply from the watershed, and recycling of nitrogen in the Water column was relatively insignificant. Instead, the bulk of DIN required by phytoplankton came from regeneration in benthic sediments. Some phytoplankton-N was transferred to pelagic consumers, but most sedimented to the benthos. During the summer bloom period, the flux of N deposition from the water column to the benthos was roughly balanced by DIN flux from sediments to the water column.

Although riverine DIN supply is much less than demand by phytoplankton during summer, on an annual scale riverine supply greatly exceeds demand by primary producers in the upper estuary. However, most riverine N supply on an annual basis occurs during discharge that prohibits phytoplankton development and thus DIN is transported down the estuary to Plum Island Sound and the Gulf of Maine. Whereas an annual perspective is relevant for some questions such as percentage nitrogen retention on an annual basis, it is the nitrogen that is processed during summer that drives the productive oligohaline food web. The processing of all watershed-derived DIN in the upper estuary during summer highlights the importance of this zone to the nitrogen cycle of the entire estuary. Consequently, a thorough understanding of the biogeochemistry of estuarine ecosystems requires incorporation of processes and transformations occurring in the relatively understudied oligohaline zone.

The research was supported by NSF-DEB-9407829, EPAR824767010, and NSF-OCE-921446l. We thank Susan Oleszko-Szuts and Governor Dummer Academy for use of laboratory facilities at the field site, Bill Morrison for use of his dock, and Gene Stoermer for diatom species identification. We also thank Kris Tholke for running the isotope samples at the Ecosystems Center and for assistance in the field, as well as Joe Vallino, Chuck Hopkinson, Jim McClelland, Hap Garritt, Anne Giblin, Chris Neill, Meredith Hullar, Dean Pakulski, Deana Erdner, Charlie Vorosmarty, Bobbie Sichol, Matt Distler, Raquel Machas, Mike Buchalski, and Claire Peterson for sampling help and important insights during many discussions about the project. Finally, we thank Thermo Environmental Instruments for providing a [NO.sub.x] analyzer for nitrate analysis and two anonymous reviewers for constructive comments on the manuscript.

(1.) The Ecosystems Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 USA

(2.) Biology Department, Florida International University, Miami, Florida 33199 USA

(4.) Present address: Coastal Ecology Institute, Louisiana State University, Baton Rouge, Louisiana 70803-7503 USA.

Alpine, A. E., and J. E. Cloern. 1992. Trophic interactions and direct physical effects control phytoplankton biomass and production in an estuary. Limnology and Oceanography 37:946-955.

Anderson, G. F 1986. Silica, diatoms and a freshwater productivity maximum in Atlantic coastal plain estuaries, Chesapeake Bay. Estuarine, Coastal and Shelf Science 22:183-197.

Bowden, W. B., C. J. Vorosmarty, J. T. Morris, B. J. Peterson, J. E. Hobbie, P. A. Steudler, and B. Moore. 1991. Transport and processing of nitrogen in a tidal freshwater wetland. Water Resources Research 27:389-408.

Braman, R. S., and S. A. Hendrix. 1989. Nanogram nitrite and nitrate determination in environmental and biological materials by vanadium (III) reduction with chemiluminescence detection. Analytical Chemistry 61:2715-2718.

Breitburg, D. L., N. Steinberg, S. DuBeau, C. Cooksey, and E. D. Houde. 1994. Effects of low dissolved oxygen on predation on estuarine fish larvae. Marine Ecology Progress Series 104:235-246.

Cloern, J. E. 1996. Phytoplankton bloom dynamics in coastal ecosystems: a review with some general lessons from sustained investigation of San Francisco Bay, California. Review of Geophysics 34:127-168.

Conley, D. J., C. L. Schelske, and E. F. Stoermer. 1993. Modification of the biogeochemical cycle of silica with eutrophication. Marine Ecology Progress Series 101:179-192.

Dauer, D. M., A. J. Rodi, and J. A. Ranasinghe. 1992. Effects of low dissolved oxygen events on the macrobenthos of the lower Chesapeake Bay. Estuaries 15:384-391.

Deegan, L. A., and R. H. Garritt. 1997. Evidence for spatial variability in estuarine food webs. Marine Ecology Progress Series 147:31-47.

Doering, P. H., C. A. Oviatt, L. L. Beatty, V. F Banzon, R. Rice, S. P. Kelly, B. K. Sullivan, and J. B. Frithsen. 1989. Structure and function in a model coastal ecosystem: silicon, the benthos and eutrophication. Marine Ecology Progress Series 52:287-299.

Eppley, R. W. 1972. Temperature and phytoplankton growth in the sea. Fishery Bulletin 70:1063-1085.

Fichez, R., T. D. Jickells, and H. M. Edmunds. 1992. Algal blooms in high turbidity, a result of the conflicting consequences of turbulence on nutrient cycling in a shallow water estuary. Estuarine, Coastal and Shelf Science 35:577-592.

Galloway, J. N., W. H. Schlesinger, H. Levy, A. Michaels, and J. L. Schnoor. 1995. Nitrogen fixation: anthropogenic enhancement-environmental response. Global Biogeochemical Cycles 9:235-252.

Garside, C. 1982. A chemiluminescent technique for the determination of nanomolar concentrations of nitrate and nitrite in seawater. Marine Chemistry 11:159-167.

Harding, L. W. 1994. Long-term trends in the distribution of phytoplankton in Chesapeake Bay: roles of light, nutrients and streamflow. Marine Ecology Progress Series 104:267-291.

Holmes, R. M., J. W. McClelland, D. M. Sigman, B. Fry, and B. J. Peterson. 1998. Measuring [N.sup.15]-[[NH.sub.4].sup.4] in marine, estuarine, and fresh waters: an adaptation of the ammonia diffusion method for samples with low ammonium concentrations. Marine Chemistry 60:235-243.

Hopkinson, C. S., A. E. Giblin, J. Tucker, and R. H. Garritt. In press. Benthic metabolism and nutrient cycling along an estuarine salinity gradient. Estuaries.

Hopkinson, C. S., and J. J. Vallino. 1995. The relationship among man's activities in watersheds and estuaries: a model of runoff effects on patterns of estuarine community metabolism. Estuaries 18:598-621.

Howarth, R. W., et al. 1996. Regional nitrogen budgets and riverine N and P fluxes for the drainages to the North Atlantic: natural and human influences. Biogeochemistry 35:75-139.

Hughes, J., L. A. Deegan, B. J. Peterson, and R. M. Holmes, and B. Fry. 2000. Nitrogen flow through the food web in the oligohaline zone of a New England estuary. Ecology 81:433-452.

Jorgensen, S. E. 1979. Handbook of environmental data and ecological parameters. Pergamon, Oxford, UK.

Justic, D., N. N. Rabalais, R. E. Turner, and Q. Dortch. 1995. Changes in nutrient structure of river-dominated coastal waters: stoichiometric nutrient balance and its consequences. Estuarine, Coastal and Shelf Science 40:339-356.

Justic, D., N. N. Rabalais, R. E. Turner, and W. J. Wiseman. 1993. Seasonal coupling between riverborne nutrients, net productivity and hypoxia. Marine Pollution Bulletin 26: 184-189.

Keller, A. A., P. H. Doering, S. P. Kelley, and B. K. Sullivan. 1990. Growth of juvenile Atlantic menhaden, Brevoortia tyrannus (Pisces: Clupeidae) in MERL mesocosms: effects of eutrophication. Limnology and Oceanography 35:109-122.

Lee, V., and S. Olsen. 1985. Eutrophication and management Initiatives for the control of nutrient inputs to Rhode Island coastal lagoons. Estuaries 8: 191-202.

Malone, T. C., L. H. Crocker, S. E. Pike, and B. W. Wendler. 1988. Influence of river flow on the dynamics of phytoplankton production in a partially stratified estuary. Marine Ecology Progress Series 48:235-249.

Morris, A. W., R. F C. Mantoura, A. J. Bale, and R. J. M. Howland. 1978. Very low salinity regions of estuaries: important sites for chemical and biological reactions. Nature 274:678-680.

Nixon, S. W., et al. 1996. The fate of nitrogen and phosphorus at the land--sea margin of the North Atlantic Ocean. Biogeochemistry 35:141-180.

Nixon, S. W., C. A. Oviatt, J. Frithsen, and B. Sullivan. 1986. Nutrients and the productivity of estuarine and coastal marine ecosystems. Journal of the Limnological Society of Southern Africa 12:43-71.

Nixon, S. W., and M. E. Q. Pilson 1983. Nitrogen in estuarine and coastal marine systems. Pages 565-648 in E. J. Carpenter and D. G. Capone, editors. Nitrogen in the marine environment. Academic Press, New York, New York, USA.

Odum, W. E. 1988. Comparative ecology of tidal freshwater and salt marshes. Annual Review of Ecology and Systematics 19:147-176.

Peterson, B. J. 1980. Aquatic primary productivity and the [C.sup.14]-[CO.sub.2] method: a history of the productivity problem. Annual Review of Ecology and Systematics 11:359-385.

Redfield, A. C. 1958. The biological control of chemical factors in the environment. American Scientist 46:205-221.

Redfield, A. C., B. H. Ketchum, and F. A. Richards (1963). The influence of organisms on the composition of seawater. Pages 26-77 in M. N. Hill, editor. The sea. Volume 2. John Wiley and Sons, New York, New York, USA.

Rehbehn, R., B. Schuchardt, M. Schirmer, and G. O. Kirst. 1993. The distribution of Actinocyclus normanii (Bacillariophyceae) in estuaries: field observations and laboratory investigations. Netherlands Journal of Aquatic Ecology 27: 205-214.

Sanders, R., C. Klein, and T. Jickells. 1997. Biogeochemical nutrient cycling in the upper Great Ouse Estuary, Norfolk, U.K. Estuarine, Coastal and Shelf Science 44:543-555.

Schelske, C. L., and E. F. Stoermer. 1971. Eutrophication, silica depletion, and predicted changes in algal quality in Lake Michigan. Science 173:423-424.

Schelske, C. L., E. F. Stoermer, D. J. Conley, J. A. Robbins, and R. M. Glover. 1983. Early eutrophication in the lower Great Lakes: new evidence from biogenic silica in sediments. Science 222:320-322.

Schuchardt, B., U. Haesloop, and M. Schirmer. 1993. The tidal freshwater reach of the Weser Estuary: riverine or estuarine? Netherlands Journal of Aquatic Ecology 27:215-226.

Schuchardt, B., and M. Schirmer. 1991. Phytoplankton maxima in the tidal freshwater reaches of two coastal plain estuaries. Estuarine, Coastal and Shelf Science 32:187-206.

Seitzinger, S. P. 1988. Denitrification in freshwater and coastal marine ecosystems: ecological and geochemical significance. Limnology and Oceanography 33:702-724.

Sigman, D. M., M. A. Altabet, R. Michener, D. C. McCorkle, B. Fry, and R. M. Holmes. 1997. Natural abundance-level measurement of the nitrogen isotopic composition of oceanic nitrate: an adaptation of the ammonia diffusion method. Marine Chemistry 57:227-242.

Solorzano, L. 1969. Determination of ammonium in natural waters by the phenolhypochlorite method. Limnology and Oceanography 14:799-801.

Strickland, J. D. H., and T. R. Parsons (1972). A practical handbook of seawater analysis. Ottawa, Canada, Fisheries Research Board of Canada.

UNESCO. 1985. The international system of units (SI) in oceanography. UNESCO Technical Papers Number 45. IAPSO Publication Scientifique Number 32, Paris, France.

Valiela, I., G. Collins, J. Kremer, K. Lajtha, M. Geist, B. Seely, J. Brawley, and C. H. Shaw. 1997. Nitrogen loading from coastal watersheds to receiving estuaries: new method and application. Ecological Applications 7:358-380.

Vallino, J. J., and C. S. Hopkinson. 1998. Estimation of dispersion and characteristic mixing times in Plum Island Sound estuary. Estuarine, Coastal and Shelf Science 46:333-350.

Walsh, T. W. 1989. Total dissolved nitrogen in seawater: a new-high-temperature combustion method and a comparison with photo-oxidation. Marine Chemistry 26:295-311.

Wright, R. T., R. B. Coffin, and M. E. Lebo. 1987. Dynamics of planktonic bacteria and heterotrophic microflagellates in the Parker Estuary, northern Massachusetts. Continental Shelf Research 7:1383-1397.


Introduction

Root development in plants is a complex process involving a high degree of morphological plasticity, which reflects inherent adaptive mechanisms to highly variable environmental conditions. Although the molecular determinants of root morphology and functioning are only starting to be understood, classical physiological experiments have clearly implicated both local and systemic regulatory circuits in the determination of root plasticity. In addition to its role in selective exploitation of specific soil domains for available nutrient and water sources, an important role of root developmental plasticity is to provide plants with the ability to recognize and respond to diverse biotic signals from soil microorganisms. Discriminative recognition of and appropriate responses to these biotic cues are essential for plant survival, and in the case of symbiotic plant–microbe interactions provides an additional means for selective exploitation of otherwise inaccessible nutrient sources. Nitrogen-fixing symbioses of legume plants provide an interesting example of the latter phenomenon ( Vance, 1998 ). Upon infection with specific strains of rhizobia, root cortical cells of legume plants undergo dedifferention, initiate cell divisions, and redirect their developmental fate towards the formation of nodule primordia. Thereafter, through a highly organized and controlled series of events, the nodule primordia develop into fully functional, nitrogen-fixing organs, root nodules (for recent reviews see Hadri et al. 1998 Hirsh, 1992 ). Nodule organogenesis is activated in response to specific lipo-chitooligosacharide signal molecules (Nod factors) synthesized by compatible strains of rhizobia ( Hadri & Bisseling, 1998 Spaink, 1996 ). Structural and functional adaptations of the root to Nod factors and rhizobial infection is controlled by the host plant and has been shown to be modulated by various environmental factors, including the availability of combined nitrogen, as well as developmental cues associated with plant growth ( Caetano-Anolles & Gresshoff, 1991 Francisco & Harper, 1995 Nutman, 1952 Parsons et al. 1993 Streeter, 1988 ). One important aspect of this control process is the plant-mediated regulation of the extent of nodulation in response to rhizobial infection. The plant host actively controls the number of successful nodulation events on at least two different levels. One level involves a premature arrest of the majority of rhizobial infections, such that only a restricted number of nodules are formed within a highly specific susceptible zone, located just behind the growing root tip ( Vasse et al. 1993 ). In Medicago truncatula, the sickle mutation has been shown to cause a dramatic increase in the number of persistent rhizobial infection, resulting in hypernodulation of the susceptible zone of the mutant root ( Penmetsa & Cook, 1997 ). The latter phenotype was attributed to a second effect of the same mutation, namely overall insensitivity of the mutant plant to the hormone ethylene ( Penmetsa & Cook, 1997 ). These results suggest that in addition to its well-characterized functions in plant development, ethylene is also involved in the signaling pathway controlling rhizobial nodulation of legumes. A role of the plant hormone ethylene in several other aspects of symbiotic development has been well documented for at least some legume plant species ( Fernandez-Lopez et al. 1998 Grobbelaar et al. 1970 Heidstra et al. 1997 for a recent review see Hirsch & Fang, 1994 ). However, ethylene-insensitive mutants of soybean were shown to display a wild-type nodulation pattern ( Schmidt et al. 1999 ). It is unclear at present whether the difference between the effects of ethylene and/or ethylene-sensitivity on nodulation in soybean and other legume plant species reflects a differential role of this hormone in regulating nodule development.

In addition to limiting the number of persistent rhizobial infections within the root susceptible zone, the plant also exerts a spatial and temporal control of root susceptibility to nodulation. This mechanism is referred to as autoregulation or feedback regulation of nodulation, and involves inhibition of nodule formation on younger root tissues by prior nodulation events in older root regions ( Kosslak & Bohlool, 1984 Nutman, 1952 Pierce & Bauer, 1983 ). Autoregulation renders the root cells only transiently susceptible to rhizobial infection, resulting in a narrow zone of infection and nodule differentiation (susceptibility zone Bhuvaneswari et al. 1981 ). Plants defective in this mechanism continue to nodulate on newly developing roots and form a large number of nodules over the entire root system (hypernodulation or supernodulation phenotype). Based on experiments involving split root system and grafting between wild-type and supernodulation mutant plants, an interplay between local and systemic signaling events in establishing the autoregulatory control of nodulation has been postulated ( Caetano-Anolles et al. 1991 Sheng & Harper, 1997 ). Similar experiments have identified leaf tissues as the major source of the systemic signal(s), implicating long distance communication between the root and shoot in autoregulation of nodule number. A local, non-systemic control exerted by fully mature nodules over the outgrowth of younger nodulation events has also been postulated ( Caetano-Anolles et al. 1991 Nutman, 1952 ). The exact nature of the mechanisms involved in autoregulation is not understood and the identity of the postulated systemic and local signaling compounds remains unknown. However, it is tempting to speculate that the autoregulatory response relies, at least in part, on the mechanism of sensing and regulating cell divisions, and thus may constitute a part of a more general mechanism regulating plant growth. In this context it is interesting to note that plant developmental processes other than nodulation, such as those associated with the generation of apical root meristems, have been shown to influence nodule formation via a mechanism that resembles autoregulation ( Gresshoff et al. 1989 Nutman, 1952 ). Conversely, mutations that impair the autoregulatory response have been found to exert various pleiotropic effects on plant growth and, almost invariably, lead to a high nitrate-tolerant symbiosis (efficient nodulation in the presence of high nitrate nts phenotype). Nitrate (NO3 – ) modulates plant growth and exerts complex effects on root development, symbiont recognition and nodulation ( Dazzo & Brill, 1978 Gresshoff, 1993 Zhang et al. 1999 ). Common factors may be involved in the mechanisms regulating the extent of nodulation, nitrate inhibition and other related plant growth responses (e.g. lateral root formation). Alternatively, close interactions between specific regulatory pathways (e.g. autoregulatory and nitrate inhibition pathways) may be sufficient to account for the pleiotropic effects of a single mutation in one of the controlling elements. Understanding the nature of the regulatory processes controlling nodule differentiation and nodule number, and integrating them in the overall mechanisms governing root growth and development constitute important elements of our quest to understand symbiotic nitrogen fixation in legumes.

We have previously identified plant mutants from the diploid legume Lotus japonicus that define a locus controlling normal root development. The same mutations have been found to confer an aberrant response of the mutant plant to the challenge by symbiotic rhizobia, resulting in hypernodulation and abnormal plant growth phenotypes ( de Bruijn et al. 1998 Schauser et al. 1998 Szczyglowski et al. 1998a Szczyglowski et al. 1998b ). Here we present a detailed characterization of these mutant lines and show that the underlying mutations affect plant development by changing the position and duration of root cell growth.


Results

Isolation and genetic analysis of hypernodulated aberrant root formation (Har) mutants of L. japonicus

We have previously described the isolation of two allelic EMS-induced mutant lines of L. japonicus ecotype Gifu (Ljsym34-1 and Ljsym34-2), which display both unusual symbiotic (hypernodulated) as well as drastically altered root developmental (aberrant root) phenotypes ( de Bruijn et al. 1998 Szczyglowski et al. 1998a see Fig. 1). In addition, we identified an independent mutant line from a T-DNA mutagenesis experiment (sym16) with a highly similar phenotype, but the mutant phenotype was found to be genetically unlinked to the T-DNA insertion ( Schauser et al. 1998 ). The allelic Ljsym34-1 and Ljsym34-2 mutations were found to be monogenic recessive with regard to the aberrant root phenotype and incomplete dominant in terms of the symbiotic hypernodulation phenotype ( Szczyglowski et al. 1998a Szczyglowski et al. 1998b , data not shown). The sym16 mutation was monogenic recessive for all phenotypes ( Schauser et al. 1998 ). Reciprocal crosses revealed that the Ljsym34-1/2 and sym16 mutants belong to the same complementation group (data not shown). In accordance with the recently proposed guidelines for genetic nomenclature for L. japonicus ( Stougaard et al. 1999 ), the corresponding alleles were re-named har1-1 (formerly Ljsym34-1), har1-2 (formerly Ljsym34-2), and har1-3 (formerly sym16), and the corresponding wild-type gene was named Har1. Based on its slightly stronger mutant phenotype, the har1-1 allele was chosen for further detailed analysis.

Root and nodulation phenotypes of wild-type and har1 mutant L. japonicus plants.

Plants were grown for 21 days in the presence (a wild-type and b har1-1) or absence (c har1-1) of Mezorhizobium loti NZP2235, with 0.5 m m KNO3 in the watering solution. Panels (d) and (e) show a close-up of the nodulated roots shown in (a) and (b), respectively. Panel (f) shows har1-3 grown for 3 weeks in the presence of rhizobia and for an 8 additional weeks in nitrogen-rich Hornum growth medium containing 8 m m KNO3 and 5 m m NH4 + ( Thykjaer et al. 1998 ).

The symbiotic (hypernodulated) phenotype of the har1-1 mutant

Inoculation of L. japonicus har1-1 plants with Mesorhizobium loti strain NZP2235 resulted in an almost total inhibition of plant growth and the unusual hypernodulation phenotype previously described ( Szczyglowski et al. 1998a see Fig. 1). Nodule-like structures covering nearly the entire short root system developed concomitantly with inhibition of plant growth and deterioration of overall plant vitality (hypernodulation response, HNR, phenotype Szczyglowski et al. 1998a Szczyglowski et al. 1998b ). To further examine this novel hypernodulation phenotype, a derivative of M. loti strain NZP2235, carrying a constitutively expressed hemA::lacZ reporter gene fusion, was used to analyze early events during the infection of wild-type and har1-1 mutant plants. Microscopical analyses revealed that the mode of rhizobial primary entry into har1-1 mutant roots was through infection threads initiated within deformed root hairs, as in wild-type plants. Other early infection events, such as root hair deformation (Had), hair curling (Hac) and infection thread formation (Inf) were also similar in har1-1 mutant and wild-type plants (data not shown). However, subsequent stages of symbiotic development were found to differ significantly. In wild-type plants, the majority of primary infection events were found to be arrested early during symbiotic development, without advancing beyond the stage of a few cortical cell divisions, resulting in 9–15 nitrogen-fixing nodules on the upper portion of fully elongated 21-day-old L. japonicus wild-type roots (see Fig. 1a,d). In contrast, roots of har1-1 mutant plants, at 11 days after inoculation with rhizobia, had much more abundant foci of cortical cell divisions than wild-type roots spanning almost the entire length of the root ( Figs 2 and 3a,b). These initial cortical cell divisions gave rise to nodule primordia, and subsequently to a mass of nodules covering almost the entire root ( Figs 1e,f and 3c,d). In addition, nodule morphogenesis on har1-1 mutant plants was found to be insensitive to normally inhibitory concentrations of combined nitrogen (5–15 m m NO3 – ). Six weeks after inoculation with rhizobia, har1-1 mutant plants developed approximately 40–60 nodules, in the presence of high concentrations of nitrate (5–15 m m ), or ammonia (1–3 m m ). In contrast, nodule development in control wild-type L. japonicus plants was found to be highly inhibited by combined nitrogen sources. For example, in the presence of 15 m m KNO3 only a few small nodule-like structures (bumps) were observed. When grown in the presence of a low concentration of KNO3 (0.5 m m ), 21-day-old nodules on har1-1 mutant plants were significantly smaller than wild-type nodules of the same age (data not shown). In spite of the size difference, light and transmission electron microscopic examination of the infected zone of nodules formed on har1-1 plants revealed a normal (wild-type like) cytology and histology ( Fig. 3e,f). Moreover, nodules formed on har1-1 mutant roots had the capacity to fix nitrogen (reduce acetylene) at levels comparable to wild-type nodules when calculated on a per plant basis (data not shown).

Infection and nodulation events in wild-type and har1-1 plants upon inoculation with M. loti strain NZP2235 carrying a hemA:lacZ reporter gene fusion.

Roots were stained for β-galactosidase activity and examined using brightfield microscopy. Open bars denote the number of root hairs with visible infection threads solid bars indicate the number of nodules and nodule primordia. Each value represents the mean of measurements from 7 to 15 plants. Error bars represent 95% confidence intervals.

Microscopic analysis of symbiotic development in wild-type and har1-1 plants.

(a,b) Brightfield micrographs of cleared wild-type and har1-1 mutant roots 11 days after infection with M. loti NZP2235. Nodule primordia are indicated by arrows. (c,d) Montages of the nodulation phenotypes of wild-type and har1-1 mutant plants 14 days after inoculation with M. loti NZP2235 carrying a hemA:lacZ reporter gene fusion. Roots were stained for β-galactosidase activity and examined using brightfield microscopy. (e,f) Transmission electron micrographs of the central zone of wild-type and har1-1 mutant nodules showing infected host cells filled with bacterial endosymbionts, and a portion of adjacent highly vacuolated uninfected cells.

The non-symbiotic (aberrant root formation) phenotype of the har1-1 mutant

Uninoculated L. japonicus har1-1 mutant plants develop a significantly shortened root system and an enhanced number of lateral roots as compared to wild-type plants ( Szczyglowski et al. 1998a ). To further analyze this phenotype, we examined lateral root formation in uninoculated har1-1 and wild-type primary roots and found no significant differences in their position relative to the root tip, as almost all LRPs were found in a region located 0.75–4.5 cm from the root apex. However, the density of lateral root primordia and emerged lateral roots (number per unit length of root) were at least three times higher in the har1-1 mutant than in wild-type plants (data not shown). Detailed examination of median longitudinal sections of uninoculated har1-1 root samples approximately 2 cm above the root tip revealed a significantly higher level of mitotic activity in the root pericycle layer of har1-1 versus wild-type roots ( Fig. 4a,b). In 9-day-old roots of the har1-1 mutant, periclinal cell divisions in the pericycle, as well as the development of two or three new cell layers, were detected in all sections examined. Abundant anticlinal cell divisions in the pericycle, as well as in the neighboring cortical cell layers, could also be detected in sections of har1-1 mutant roots, and in several cases well-developed lateral root primordia were observed. In contrast, wild-type roots of a similar age exhibited only limited mitotic activity in the pericycle, which was mostly composed of a single cell layer ( Fig. 4a,b). Lateral root primordia were only rarely observed in sections of wild-type roots.

Mitotic activity of the root pericycle in wild-type and har1-1 mutant roots.

(a,b) Median longitudinal section of segments of 9-day-old uninoculated roots of wild-type (a) and har1-1 (b) mutant plants approximately 2 cm above the root tip. Arrows indicate the pericycle cell layer.

Root and shoot phenotype of the har1-1 mutant

The short root phenotype of L. japonicus har1-1 plants has been documented previously ( de Bruijn et al. 1998 Szczyglowski et al. 1998a Szczyglowski et al. 1998b ). To further assess this phenotype quantitatively, the longitudinal growth of uninoculated and inoculated har1-1 and wild-type roots was measured. An average root length of 61.3 ± 0.2 mm was observed for uninoculated har1-1 plants 21 days after sowing versus 131.9 ± 1.0 mm for wild-type uninoculated control plants ( Fig. 5a). In addition, the root mass of uninoculated har1-1 plants at 21 days after sowing (26 ± 2 mg fresh weight) was significantly smaller than that of the wild-type plants (45 ± 6 mg). The aberrant har1-1 root growth phenotype was even more extreme when plants were inoculated with M. loti strain NZP2235. Root growth of har1-1 plants ceased entirely within the first days after inoculation, and the roots did not exceed an average length of 20 ± 1.5 mm ( Fig. 5b). The shoot mass of inoculated har1-1 plants was also significantly reduced in infected plants, while it was comparable in uninoculated har1-1 versus wild-type shoots ( Fig. 5c,d).

Growth kinetics of roots and shoots of wild-type and har1-1 mutant plants in the presence or absence of M. loti.

(a,b) Root growth kinetics of uninoculated (a) and inoculated (b) wild-type and har1-1 mutant plants. (c,d) Shoot growth kinetics of uninoculated (c) and inoculated (d) wild-type and har1-1 mutant plants. All plants were grown in the presents of 0.5 m m KNO3. Each value represents the mean of measurements of at least 30 plants. Error bars represent 95% confidence intervals.

Cytology of har1-1 mutant roots

Having established that several root developmental parameters (root length/elongation, lateral root formation and overall root mass accumulation) were significantly altered in the har1-1 mutant line, the phenotype of the har1-1 mutant roots was further analyzed. First, the cellular organization of wild-type L. japonicus roots was examined. Primary L. japonicus roots were found to contain an outer single layer of epidermal cells, 3–5 irregularly shaped cortical cell layers surrounding a single layer of endodermal cells, an innermost region consisting of a single layer of pericycle cells enclosing the vascular cylinder, and a distally located area of root cap cells. The same general organization of root cell layers was found in har1-1 mutant roots ( Fig. 6). However, several significant differences were observed. Vacuolation, which typically accompanies cell expansion, occurred closer to the root tip in har1-1 versus wild-type roots ( Fig. 6a,b). In addition, the diameter of mutant har1-1 roots (0.23 ± 0.03 mm), measured using digitized micrographs of living roots at 3–6 mm from the root tip, was significantly smaller than the average diameter of the corresponding region of wild-type L. japonicus roots (0.31 ± 0.02 mm see also Fig. 6c,d). Based on these resultes, the following two hypotheses were formulated and tested: (1) the har1-1 mutation affects the radial organization of the L. japonicus root, and/or (2) the decreased root length and diameter is a result of abnormal (reduced) cell expansion.

Histology of non-symbiotic root development.

(a,b) Median longitudinal sections of the root tip regions of 9-day-old wild-type and har1-1 mutant plants showing the overall anatomy of the meristematic region and the position of the cell elongation/vacuolation zone. (c,d) Cross-sections of roots of 22-day-old plants at approximately 600 μm above the root tip, showing differences in the extent of cell vacuolation and root diameter. (e,f) Cross-sections of mature root regions of 6-day-old plants approximately 1 cm above the root tip.

The har1-1 mutation results in a diminished radial expansion of root cells

To test the first hypothesis (a defect in the radial organization of roots), a large number of root sections were examined microscopically. No clear evidence for one or more missing cell layers or a diminished number of root cells in har1-1 mutant roots was found ( Fig. 6). Therefore, we tested the second hypothesis (changes in root cell expansion), by analyzing sections of the fully differentiated region of primary roots ( Fig. 6e,f). The average total cross-sectional area of har1-1 mutant roots was almost two times smaller than that of the wild-type root ( Table 1). Individual cell layers (epidermis, cortex, endodermis and root stele) of har1-1 mutant roots showed a similar reduction in size, contributing nearly equally to the overall decrease in root diameter ( Table 1). Subsequently, the projected cross-sectional area of individual root cells was analyzed to determine whether cell radial expansion was altered in har1-1 mutant roots. The radial surface areas of individual cells of the epidermis, cortex and endodermis of roots had a frequency distribution that was confined to a much smaller size range for the har1-1 mutant ( Fig. 7), suggesting that the har1-1 mutation limits the ability of root cells to expand.

Tissue Wild type (a) har1-1 (b) b/a ratio
Total section 62357 ± 14772 33786 ± 4845 0.54
Epidermis 10073 ± 2250 5626 ± 760 0.56
Cortex 45228 ± 11283 24433 ± 3820 0.54
Endodermis 2097 ± 373 1387 ± 175 0.66
Stele 5070 ± 1033 2655 ± 513 0.52

Root cell radial expansion in wild-type and har1-1 mutant plants.

The frequency distribution of the cross-sectional area of root epidermal (a), cortical (b) and endodermal (c) cells of 6-day-old plants measured at the location of approximately 1 cm above the root tip is shown. Ten wild-type and mutant plants were analyzed and each point represents a frequency value for cell size range increments of 35 μm (epidermis), 200 μm (cortex) and 20 μm (endodermis) n, represents the number of cells measured.

The length of the meristematic region is shortened in har1-1 mutant roots

To examine whether the short root phenotype of har1-1 mutant plants was caused by alteration in the primary direction of cell expansion along the apical–basal axis, the length of har1-1 root epidermal cells was measured. The average length of fully expanded epidermal cells of har1-1 mutant roots (138 ± 30 μm) was nearly equal to that of wild-type roots (132 ± 30 μm), indicating that the har1-1 mutation did not affect longitudinal cell expansion of the epidermis ( Fig. 8a). These results suggest that the short root phenotype of the mutant plant is unlikely to be due to an abnormal longitudinal expansion of root cells. However, consistent with our earlier observations (see above), epidermal cells of the har1-1 mutant roots showed evidence of elongation along the longitudinal axis significantly closer to the root tip (hence earlier in development) than wild-type roots ( Fig. 8a). Transmitted brightfield microscopy and laser scanning microscopy of whole cleared roots stained with acetocarmine revealed an area of densely cytoplasmic cells constituting the root meristematic region whose borders could be defined by interactive threshholding techniques using digital image processing. In independent experiments using both types of microscopy, an approximately 2.6-fold reduction in the projected area of the har1-1 mutant versus wild-type root meristematic regions was detected (45082 ± 4373 μm 2 n = 11 versus 121635 ± 12545 μm 2 , n = 19 see also Fig. 8b). This reduction in the size of the root meristematic region was invariably associated with an acropetal displacement of the root cell elongation/vacuolation and vascularization zones ( Fig. 8b). har1-1 mutant roots also showed an inferior root cap structure in comparison with the wild type, but remained gravitropic (data not shown). In addition, the mitotic index was measured in order to estimate the proportion of mitotic cells in the root meristematic regions of har1-1 versus wild-type plants of an equal age, but no significant differences were found (har1-1 mutant MI = 3.8 ± 0.8 versus wild-type MI = 3.96 ± 0.7).

Root cell elongation and size of meristematic region in wild-type and har1-1 mutant plants.

(a) Epidermal cell length along the root axis. Each point represents the mean cell length value for range increments of 250 μm (for the first 4 mm from the root tip), and 500 μm (between 4 and 11 mm from the root tip), along the root axis. Single and double arrows indicate significant differences between the mean epidermal cell length at two consecutive and equivalent positions in wild-type and har1-1 mutant roots. (b) Intact roots of 14-day-old plants stained with acetocarmine. The red-stained meristematic regions of the wild-type and har1-1 mutant roots appear as dark areas, and their extent is indicated by the brackets.

Effect of hormones on har1-1 mutant root elongation

The gross changes in har1-1 root morphology suggested that the hormonal regulation of the mutant root development could be disturbed. In order to test this hypothesis the effects of exogenous hormone applications on root elongation of wild-type and har1-1 mutant roots were investigated using a plate bioassay specifically developed for this purpose. Sucrose was found to be required at a relativey high concentration (4.5%) to support the maximal and uniform root elongation of both wild-type and har1-1 mutant roots. The roots of wild-type plants elongated more rapidly in the dark than in the light especially during the first 2–3 days of growth, after which time the difference in growth rate disappeared ( Fig. 9a). har1-1 mutant plants formed short roots when grown in the dark under conditions that provided maximal wild-type root growth. A dramatic inhibition or retardation of root elongation was observed after 2 days of incubation in the dark. Interestingly, the short root phenotype was partially suppressed when har1-1 mutant plants were grown in the light ( Fig. 9b).

Effect of light on the elongation rates of wild-type and har1-1 mutant roots.

(a) The rates of wild-type root elongation. (b) The rates of har1-1 mutant root elongation. Each value is the mean of measurements on 20 plants. Error bars represent 95% confidence intervals.

Since har1-1 roots exhibited an abnormal pattern of radial expansion of root cells, as well as a hypernodulation response (see Figs 1, 3d and 6), and since the hypernodulation phenotype of a Medicago truncatula mutant (sickle) had been correlated with a change in ethylene sensitivity ( Penmetsa & Cook, 1997 ), the sensitivity of wild-type and har1-1 seedlings to the ethylene precursor 1-aminocyclopropane1-carboxylic acid (ACC) was examined. When grown vertically in the dark on agar plates containing increasing concentrations of ACC, both wild-type and har1-1 mutant seedlings showed the same overall sensitivity pattern to ethylene inhibition of root growth ( Fig. 10a). However, in two independent experiments, har1-1 mutant roots displayed a slightly increased resistance to certain concentrations of ACC as compared with wild-type roots (e.g. 1 × 10 −7 to 1 × 10 −6 M ACC in Fig. 10a). To further evaluate the observed decrease in sensitivity of the har1-1 mutant line to ACC, we examined the responses of entire seedlings to exogenously applied ACC. When grown in the dark in the presence of ACC, both wild-type and har1-1 seedlings showed a typical triple response ( Guzman & Ecker, 1990 ), consisting of a shortening of the hypocotyl, inhibition of root elongation, and exaggeration of apical hook curvature ( Fig. 11). Both genotypes exhibited similar levels of sensitivity to ACC in terms of hypocotyl length (data not shown).

Wild-type and har1-1 root growth in the presence of increasing concentrations of exogenously applied plant hormones.

The relative elongation of wild-type and har1-1 mutant roots in the presence of (a) 1-aminocyclopropane1-carboxilic acid (ACC) (b) α-naphthalene-acetic acid (NAA) or 6-benzylaminopurine (BA) is shown. Each value is the mean of measurements on 20 plants. Error bars represent 95% confidence intervals. The mean value of 100% root growth in (a) wild-type, 44.5 ± 3.4 mm har1-1 21.8 ± 1.2 mm in (b) wild-type, 56.9 ± 2.4 mm har1-1, 25.8 ± 1.6 mm in (c) wild-type 37.1 ± 1.9 har1-1, 21.6 ± 1.1 mm.

Triple response to ACC of wild-type and har1-1 mutant plants.

Wild-type and har1-1 mutant seeds were germinated on MS medium in the dark at 28°C in the absence (0) or presence of increasing concentations (1–100 m m ) of ACC. The photograph was taken 6 days after incubation in the dark. The triple response is characterized by shortened hypocotyls, roots and exaggerated apical hook formation.

Subsequently, the effect of the auxin α-naphthalene-acetic acid (NAA) on root growth was examined. Wild-type and har1-1 mutant roots were found to display a similar overall NAA sensitivity pattern but, as had been observed with ACC, roots of har1-1 mutant plants showed a slight NAA-insensitive phenotype at higher concentrations of NAA ( Fig. 10b).

In Arabidopsis, cytokinin inhibits root elongation in light- and dark-grown seedlings due to the stimulation of endogenous ethylene production ( Cary et al. 1995 ). Therefore, the sensitivity of wild-type and har1-1 mutant roots to exogenously applied cytokinin (thus endogenously produced ethylene) was also examined. The presence of even very low concentrations (10–50 n m ) of the synthetic cytokinin, 6-benzylaminopurine (BA), significantly reduced root growth in both genotypes. However, again the har1-1 mutant roots exhibited a moderately higher resistance to a wide range of BAP concentrations then wild-type roots ( Fig. 10c). This slightly elevated resistance of har1-1 mutant root could be due to either an altered ethylene-independent response to cytokinin, diminished cytokinin-stimulated ethylene production, and/or an attenuated response to endogenously produced ethylene. To distinguish between these possibilities, the influence of exogenously added cytokinin on inhibition of root growth was examined in the presence silver ions, applied as silver thiosulfate, to inhibit ethylene binding ( Beyer, 1979 ), or aminoetoxyvinylglycine (AVG, Yang & Hoffman, 1984 ) to inhibit ethylene biosynthesis. Since 50 nanomolar BA almost maximaly inhibited root elongation in both genotypes, we used this concentration of cytokinin in combination with variable concentrations of inhibitors.

In the absence of BA, Ag + only had a limited stimulatory effect on the growth/elongation of wild-type and har1-1 mutant roots ( Fig. 12a). Five μM Ag + was sufficient to overcome all the inhibitory effects of cytokinin on har1-1 mutant roots, whereas 5 μ m Ag + restored wild-type root growth to a level of about 60% of that of untreated control roots ( Fig. 12b).

Effect of BA on elongation of the wild-type and har1-1 mutant roots in the presence of ethylene perception/synthesis inhibitors.

The relative elongation of wild-type and har1-1 mutant roots in the presence of (a) Ag + (b) BA plus Ag + (c) AVG and (d) BA plus AVG is shown. Mean value of 100% root growth in (a) and (b) wild-type, 41.6 ± 3.0 mm har1-1, 23.3 ± 1.2 mm in (c) and (d) wild-type, 47.5 ± 4.0 mm har1-1, 23.2 ± 1.6 mm. For further details see legend to Fig. 10.

A similar set of experiments was conducted with an inhibitor of ethylene synthesis, AVG. In control experiments, an AVG concentration equal to or lower than 0.1 μ m had no measurable effect on root growth in both genotypes ( Fig. 12c), whereas higher concentrations were strongly inhibitory (data not shown). In contrast to the silver-mediated phenotypic suppression, a low concentration of AVG (0.1 μ m ) was found to relieve all of the inhibition caused by cytokinin, and to restore a normal long-root phenotype to wild-type plants ( Fig. 12d). However, the har1-1 mutant plants responded differently to the same treatment, not only by recovering a short root phenotype, but also by an additional stimulation of root growth. The latter effect resulted in the mutant plants developing long roots that elongated at the same rate as wild-type roots. However, a prolonged incubation (more then 10 days) of both wild-type and mutant har1-1 plants in the presence of cytokinin and AVG resulted in the total arrest of their root growth. This effect was found to be associated with a terminal differentiation of the root meristem, thus precluding experiments with longer treatment times (data not shown).


5.15: Nitrogen Fixation - Biology

Advanced A/AS Level Chemistry: More on shapes of inorganic molecules & ions

Doc Brown's Advanced A Level Chemistry

Theoretical Physical Chemistry Revision Notes

The s hapes of Molecules and Ions and bond angles related to their Electronic Structure

Part 2 Some other molecules and ions of carbon, nitrogen, sulphur and chlorine

A description, explanation, shapes and bond angles of the carbonate ion, nitrate(III) ion (nitrite ion), nitrate(V) ion (nitrate ion), nitrogen(IV) oxide (nitrogen dioxide), nitronium ion, sulfur(IV) oxide (sulphur dioxide), sulfur(VI) oxide (sulphur dioxide), sulfate(IV) ion (sulphate ion), sulfur(VI) ion (sulphate ion), chlorate(III) ion, ClO2 - (chlorite ion), chlorate(V) ion, ClO3 - (chlorate ion) and the chlorate(VII) ion, ClO4 - (perchlorate ion) are all described and discussed. All is described and explained!

Molecule shapes, dot and cross diagrams, bond angles for selected molecules and ions of nitrogen, sulfur and chlorine using the valence-shell electron-pair repulsion model (VSEPR) and the dot and cross (ox) diagrams are presented in 'Lewis style' The 'scribbles' will be replaced by neat diagrams eventually!

  • Carbonate ion, CO3 2- is trigonal planar in shape with a O-C-O bond angle of 120 o because of three groups of bonding electrons and no lone pairs of electrons.
  • The shape is deduced below using dot and cross diagrams and VSEPR theory and illustrated below.
  • Note that all the C-O bonds are identical due to delocalisation of some of the electrons ( σ sigma and π pi bonding)

valence bond dot and cross diagram for the carbonate ion

    Nitrogen(IV) oxide, NO2 (nitrogen dioxide) is bent shaped (angular), O-N-O bond angle

valence bond dot and cross diagrams for nitrogen oxides and nitrogen oxyanions

  • Note that all the N-O bonds within the molecule or ion are identical due to delocalisation of some of the electrons ( σ sigma and π pi bonding)

    Sulfur(IV) oxide, SO2 (sulphur dioxide/sulfur dioxide) molecule is a bent shape (angular), O-S-O bond angle

valence bond dot and cross diagrams for sulfur oxides and sulfur oxyanions

  • Note that all the S-O bonds within the molecule or ion are identical due to delocalisation of some of the electrons ( σ sigma and π pi bonding)

    The chlorate(III) ion, ClO2 - (chlorite ion) , is bent shaped (angular) , O-Cl-O bond angle

valence bond dot and cross diagrams for selected chlorate ions

  • Note that all the Cl-O bonds within the molecule or ion are identical due to delocalisation of some of the electrons ( σ sigma and π pi bonding)

Revision notes for GCE Advanced Subsidiary Level AS Advanced Level A2 IB Revise AQA GCE Chemistry OCR GCE Chemistry Edexcel GCE Chemistry Salters Chemistry CIE Chemistry, WJEC GCE AS A2 Chemistry, CCEA/CEA GCE AS A2 Chemistry revising courses for pre-university students (equal to US grade 11 and grade 12 and AP Honours/honors level courses)


1. Introduction

Reactive nitrogen (Nr), such as nitrate, nitrite and ammonium, is essential for the functions, processes and dynamics of ecosystems (Vitousek and Howarth 1991). Together with the advent of unlimited industrial nitrogen fixation at low costs by the Haber–Bosch process, anthropogenic activities have at least doubled annual global Nr inputs to ecosystems as compared with Nr inputs during pre-industrial times (Galloway et al 2004). Increased Nr supports the food and fuel needs of a growing human population, but it also causes numerous adverse impacts on human health and environmental sustainability, including eutrophication of aquatic ecosystems and increased N2O emissions—a potent greenhouse and ozone-depleting gas (Vitousek et al 1997, Galloway et al 2008, Sobota et al 2013). However, increased Nr is not evenly distributed at spatial scales. In Africa—a region with too little Nr—the agricultural sector has not been able to produce sufficient food for the rapidly growing population and insufficient N inputs can lead to mining of soil organic N stocks (Davidson 2009). Thus, compensating for the negative impacts associated with anthropogenic Nr inputs represents an important challenge faced by land and water managers worldwide.

Better understanding of Nr inputs and sources is critical to improve the balance between their positive and negative impacts (Bouwman et al 2009, Hong et al 2011, Swaney et al 2012). So far, regional anthropogenic Nr assessments have been made for the European Union (Sutton et al 2011), North America (Sobota et al 2013) and China (Ti et al 2012). The only existing synthesis of Nr in Africa was completed for West Africa three decades ago (Robertson and Rosswall 1986). We believe that anthropogenic Nr in Africa should be included in global assessments of human-mediated Nr.

Lake Victoria in East Africa is the second largest fresh water lake in the world and the watershed is one of the most densely populated regions in Africa. The rapidly expanding population and economy within the basin (Muyodi et al 2010), has resulted in notable changes to the physical, chemical and biological regime of the lake over the last 50 years (Juma et al 2014), including enrichment of Nr (Lung'ayia et al 2001). Previous studies in the Lake Victoria Basin were mainly focused on either water N concentrations (Gikuma-Njuru and Hecky 2005) or estimation of loading at a relatively small scale (Lindenschmidt et al 1998). However, there are no studies on the regional nitrogen budget of the basin, which is critical for improving regional Nr management and balancing its negative and positive impacts.

Here, we synthesize the existing data to develop a regional Nr budget in the Lake Victoria Basin using the net anthropogenic nitrogen input (NANI) approach. The NANI approach is an effective method to assess human-induced Nr inputs to the landscape and to evaluate their potential impacts on riverine export from large basins (Hong et al 2013). The objectives of this paper are to (1) evaluate the regional Nr budget, highlighting its underlying uncertainties, and (2) identify research gaps and suggest ways to improve future estimates.


Analytical Profiles of Drug Substances and Excipients

4. Methods of Analysis

4.1 Identification

The identity of zileuton can be confirmed by comparing the infrared absorption spectrum of the sample to that reported in Figure 9 .

4.2 Elemental Analysis

A typical elemental analysis of a sample of zileuton is as follows:

4.3 Chromatographic Methods of Analysis

4.3.1 Thin Layer Chromatography

Several thin-layer chromatographic systems have been investigated to evaluate the purity of zileuton. Two of these have been found to give the best separation of zileuton, its impurities, and degradation products.

System I uses ethyl acetate to effect the analytical separation on silica gel 60 F254, with the detection being made with short-wave UV irradiation. N-(1-benzo[b]thien-2-ylethyl)-urea (Rf, 0.12), (Z)-1-benzo[b]thien-2-ylethanone oxime (Rf, 0.59), (E)-1-benzo[b]thien-2ylethanone oxime (Rf, 0.65), and N-1benzo[b]thien-2-ylethyl) hydroxylamine can be separated from zileuton (Rf, 0.21).

System II uses a mobile phase consisting of 50:50 1 chloroform / methylene chloride / ammonium hydroxide, the separation being performed on silica gel 60 F254, and the detection is made with short-wave UV irradiation. This system can be used to separate (Z)-1-benzo[b]thien-2-ylethanone oxime (Rf, 0.06) (E)-1-benzo[b]thien-2-ylethanone oxime (Rf, 0.17), 1-benzo[b]thien-2-ylethanone (Rf, 0.56), 1-benzo[b]thien-2-ylethyamine (Rf, 0.07), and 0-(1-benzo[b]thien-2-ylethyl)-1-benzo[b]thien-2-ylethanone oxime (Rf, 0.33) from zileuton (Rf, origin).

4.3.2 High Performance Liquid Chromatography

Several HPLC methods have been developed to evaluate the quality of zileuton drug substance. System I is used to quantitate the bulk drug substance potency, while systems II and III are used to quantitate the impurities and degradation products in bulk substance. The characteristics of Systems II and III are such that they cover the range of polarities associated with the various impurities in zileuton. System II is used to monitor the degradation products.

System I
Mobile Phase:0.1 M ammonium acetate solution containing 0.025 % acetohydroxamic acid (adjust the solution with perchloric acid to pH 2.0)/acetonitrile (72:28)
Column:30 cm × 1/4″ (o.d.) × 4.6 mm (i.d.) packed with Spherisorb S10 ODS
Flow Rate:approximately 1.5 mL/minute
Detector:260 nm, 0.1 AUFS
System II
Mobile Phase:0.1 M ammonium acetate solution containing 0.025 % acetohydroxamic acid (adjust the solution with perchloric acid to pH 2.0)/acetonitrile (82:18)
Column:30 cm × 1/4″ (o.d.) × 4.6 mm (i.d.) packed with Spherisorb S10 ODS
Flow Rate:2.2 mL/minute
Detector:260 nm, 0.01 AUFS
System III
Mobile Phase:1:1 acetonitrile/0.5% perchloric acid
Column:30 cm × 1/4″ (o.d.) × 4.6 mm (i.d.) packed with Spherisorb S10 ODS
Flow Rate:approximately 1.5 mL/minute
Detector:260 nm, 0.01 AUFS

In addition, a direct chiral chromatography method for the separation of zileuton enantiomers has been developed, which uses the following conditions:

Mobile Phase:92:8:0.1 hexane / 2-propanol / trifluwoacetic acid
Column:Daicel Chiralpak AD, 250 × 4.6 mm (i.d.) (Regis) - operated at 25°C
Injection volume:10 μL (0.1 mg/mL)
Flow Rate:1.0 mL/minute
Detector:260 nm, 0.02 AUFS

4.4 Determination in Pharmaceutical Dosage Forms

The potency and primary degradation products in zileuton tablet formulations can be analyzed by a high performance liquid chromatography procedure using methyl 4-hydroxybenzoate (methylparaben) as the internal standard. The method uses a Spherisorb S10 ODS, 10 μm column, a mobile phase consisting of 72 parts of 0.1 M ammonium acetate solution containing 0.025 % acetohydroxamic acid (adjusted with perchloric acid to pH 2.0) and 28n parts acetonitrile.

4.5 Determination in Body Fluids

Simultaneous determination of zileuton and its N-dehydroxylated metabolite in untreated rat urine by micellar liquid chromatography was developed by Thomas and Albazi [ 23 ]. Separation of these compounds is achieved using sodium dodecyl sulfate (SDS) as the mobile phase, a CN-bonded silica column, and UV detection at 262 nm. Because of the solubilizing power of the micellar mobile phase, urine samples were injected into the system without any time-consuming protein precipitation and/or drug extraction steps.

A HPLC method was also developed for the determination of zileuton and its inactive N-dehydroxylated metabolite in plasma [ 24 ].


DISCUSSION

Leaf Chlorophyll Index and Maize Leaf Nitrogen Concentration

The N rates applied were absorbed by maize as evidenced by increased N leaf concentration and LCI once N is one of the components of the chlorophyll molecule ( Galindo et al., 2016 ). There have been several studies that reported the positive linear correlation with LCI and increasing N rates in maize crops. Kappes et al. (2013a) applied up to 90 kg ha −1 of N as urea, and Kappes et al. (2014) used 0, 50, 100, and 150 kg ha −1 of N as urea and both reported a linear relationship between N rate and LCI measurements. Although the LCI values are relatively high in our study, even in the control crops (ranging from 61 to almost 71, respectively), the results are comparable to those reported in the literature. Kappes et al. (2013a) reported LCI values ranging from 45.8 to 62.9, and Kappes et al. (2014) reported LCI values ranging from 55.3 to 62.1.

With respect to N leaf concentration, similar results were obtained by Costa et al. (2012) , who observed a linear and positive effect of N rates on N concentration in leaf tissue. It is worth noting that, the N leaf concentration was in the range considered adequate (27–35 g kg −1 ) ( Cantarella et al., 1997 ), even in the control crops (27.21 g kg −1 ). However, it is noteworthy the higher N requirement of maize hybrids of earlier cycle and greater production potential.

The similar result between the N sources for LCI can be attributed to the similar N concentrations of leaves obtained with urea and urea with urease inhibitor NBPT. This could be a result of poor efficiency of NBPT at neutralizing urease enzymes in the soil. Some of the reasons for the inefficiency of NBPT to control urease activity could be because some of the straw from the previous year’s wheat crop remained in the soil surface, and the year was exceptionally hot (Fig. 1). Similarly, other studies have reported no significant differences in yield between urea and improved N sources with slow-release polymers in maize ( Queiroz et al., 2011 , Valderrama et al., 2011 , Galindo et al., 2016 ).

Regarding the positive results of inoculation in LCI, similarly, Müller et al. (2016) and Galindo et al. (2016) found that maize plants that were inoculated with A. brasilense had higher LCI than plants not inoculated. Positive responses to inoculation with this diazotrophic bacteria have been obtained even when the crops are grown under condition that provided adequate amounts of N for optimum growth ( Galindo et al., 2017c ). This suggests that the positive plant responses to inoculation occurs not exclusive because of BNF in grasses but also because of the production of plant growth hormones, including indoleacetic acid, gibberellin, and cytokinin ( Galindo et al., 2017c ), that can play an essential role in the promotion of plant growth ( Bashan and de-Bashan, 2010 ). According to Pankievicz et al. (2015) , the inoculation with A. brasilense can improve development and root system growth in Setaria viridis grass due to greater CO2 fixation and lower accumulation of photoassimilated carbon in the leaves, which led to greater above ground growth, higher water content in tissue, and less stress. In addition, increased indoleacetic acid production may improve the uptake of nutrients by the higher growth of root system ( Hungria et al., 2010 ).

Maize Biometric Evaluations and Productive Components

Castro et al. (2008) reported that plant height is influenced by the availability of N in the soil since this nutrient participates directly in photosynthetic process and cell division and expansion. Gross et al. (2006) recommend that N should be applied in one or two application during the season only, due to the positive effects on plant height and maize grain yield. However, it is worth noting that plant height does not always correlate with productivity since modern hybrids with high productive potential are mostly of lower height ( Cruz et al., 2008 ). In our study it was observed that increasing N rates led to increased N availability, which possibly increased the N grain accumulation and subsequently increased the plant height, ear diameter, number of grains per spike, mass of 100 grains, and maize grain yield.

With respect to the mass of 100 grains, grain mass is a characteristic influenced by genotype, climatic conditions, and nutrient availability during grain filling stages ( Chen et al., 2012 ). Additionally, for Mello et al. (2017) , this productive component is very dependent on N uptake by maize, which reaches the uptake peak during the period between the beginning of flowering and the beginning of grain formation. Nitrogen deficiency in this period may contribute to the formation of grains with lower specific mass due to the non-translocation of adequate amounts of nutrients, which justifies the increased mass of 100 grains observed in the present study with the increase in N rates applied.

On the subject of seed inoculation with A. brasilense, the increase in stem diameter with inoculation is interesting since this morphological characteristic is one that has been more related to the percentage of lodging or plant breakage in maize. In addition, stem diameter has been reported to be an important driver for high yield because the larger the diameter the greater the capacity for plant to store photoassimilates which contribute to grain filling ( Cruz et al., 2008 Lana et al., 2009 ), which also justifies the increase in LCI, ear length and grain yield verified in the present study as a function of the inoculation with A. brasilense. Inoculation with A. brasilense also increased ear length compared with uninoculated treatments. It is possible that inoculation with A. brasilense favored the development of an improved root system, leading to a higher uptake of water and nutrients, positively influencing the nutritional status of the plant. The amount of water and nutrients to be sent to the ear is directly related to the nutritional status of the plant, the better-nourished maize plant tends to better develop its spike, which is demonstrated in the length.

It is possible that the lack of response to N sources is due to the fact that when irrigating the area, a substantial reduction in N-NH3 volatilization probably occurred as a result of increased contact between fertilizer and soil particles leading to higher NH4 + adsorption by soil ( Silva et al., 1995 ), and the effect of NBPT in reducing volatilization losses were reduced under high N rates ( Silva et al., 2017 ). In addition, the magnitude of the positive effects associated with the use of urea with NBPT has varied greatly with soil characteristics, crop management, and climatic conditions that alter NH3 volatilization at the time of fertilizer application and in the early days subsequent to this practice ( Cantarella et al., 2008 Tasca et al., 2011 ). Several studies report that the addition of urease inhibitors to urea retards the NH3 volatilization peak which, for conventional urea, is concentrated in the first week after the fertilizer application to the soil surface ( Cantarella et al., 2008 Rochette et al., 2009 Tasca et al., 2011 ). In this way, irrigation of the experimental area soon after nitrogen fertilization associated to the rainfall that occurred in the week in which the fertilizers was applied in the first year (45 mm of rainfall between 11 to 16 Jan. 2014, 3 d after the application of nitrogen fertilizers Fig. 1A) and in the second year (19 mm of rainfall on 14 Jan. 2015, and 21 mm on 21 Jan. 2015, 10 and 17 d after the application of nitrogen fertilizers, respectively Fig. 1B) effectively contributed to minimize the volatilization losses of urea, providing a similar effect to urea with NBPT. Therefore, the use of urea with NBPT in hot days and in weeks with dry conditions, a common climatic condition between April and September in the Brazilian Savannah could be very advantageous, elucidating new studies.

Studies involving the use of polymer-coated urea compared to conventional urea have shown similar effect ( Queiroz et al., 2011 Mello et al., 2017 ). Additionally, Valderrama et al. (2011) , comparing the effect of conventional urea and soluble polymer coated urea did not find advantages with the encapsulation of urea with polymers for maize grown in the Brazilian Cerrado. As the sources did not influence the main evaluations performed, urea becomes more advantageous due to its better cost-benefit ratio, in agreement with Queiroz et al. (2011) and Maestrelo et al. (2014) .

On the other hand, unlike in the present study, Abalos et al. (2014) support the hypothesis that the use of the urea inhibitor NBPT is the most appropriate option if losses through NH3 volatilization are expected to be high. According to Abalos et al. (2014) , under conditions where high inputs of N fertilizer are applied and that favors high drainage, with irrigated systems the efficiency of the urea with inhibitor NBPT can be higher. However, the author concluded that new studies are needed to improve our understanding of the conditions under which the improved efficiency fertilizers are economically viable and to compare their efficiency with that of other options, such as improved water and N fertilizer management.

Nitrogen Use Efficiency, Maize Grain Yield, and Economical Analysis

The reduction in NUE as a function of N rates increasing can be attributed to the loss of N, as clearly described in the literature. Greater N rates result in greater losses and less utilization by the crops since plant nutritional demand is limited ( Galindo et al., 2016 ). Plants are able to absorb a certain quantity of nutrients in a certain time the N that is applied and is not absorbed can be lost, decreasing the efficiency of fertilization with higher N rates, as stated in the literature as the law of diminishing returns.

The increase in NUE due to the inoculation with A. brasilense was accentuated. On average, the NUE provided by the inoculation was 3.5 times higher than the NUE of the non-inoculated. According to Cormier et al. (2013) , two strategies may be devised for NUE improvement: maintaining high yield when reducing N supply and/or increasing yield at a constant N supply. Based on the results obtained, the inoculation with Azospirillum brasilense is a very interesting strategy in N efficiency improvement and can be used to decrease N fertilizer rates applied, maintaining the same NUE and without affect negatively grain yield.

Concerning grain yield, in several studies increases in maize grain yield have been reported with the application of increasing N rates ( Kitchen et al., 2009 Holland and Schepers, 2010 Venterea et al., 2011 Kappes et al., 2014 Galindo et al., 2016 ), supporting the results observed in the current study. Also, Galindo et al. (2016) verified that the highest maize grain yield was obtained when N was supplied in larger rates at the topdressing time, and that the NH3 volatilization losses of urea that occur in irrigated maize crop did not reduce the grain yield. According to the authors, there was N available in the soil solution in the period in which the plant requires higher amounts of the nutrient. An explanation would probably be that the N applied at sowing is already in the soil solution and, when N is added, the plant has a larger amount of the nutrient to be absorbed.

It was verified an increase in maize grain yield as a function of inoculation compared to the non-inoculated treatments verified in the present study was 1012.05 kg ha −1 (equivalent to 14.3%). In addition, based on the results obtained, even with the application of high N rates associated with inoculation with A. brasilense maize grain yield was not negatively affected, indicating that high N rates do not eliminate the benefits of inoculation with A. brasilense. Kappes et al. (2013b) reported that maize inoculated with A. brasilense had a 9.4% increase in yield. Cavallet et al. (2000) reported a 17% increase in maize yield when the seeds were inoculated with Azospirillum spp. Galindo et al. (2018) verified a 5.7% increase in maize grain yield by seed inoculation with A. brasilense, using five N rates in topdressing, with an average grain yield above 9826 kg ha −1 . Similar results were obtained by Müller et al. (2016) , where maize yields were 3.8% higher with seed inoculation of A. brasilense when compared to the control, with an average grain yield above 11,000 kg ha −1 . Hungria et al. (2010) also obtained increases in maize yield on the order of 27%, corresponding to 743 kg ha −1 .

According to Hungria (2011) , the effects of maize seed inoculation on grain yield depend on the genetic characteristics of plants and strains (bacteria) in addition to the environmental conditions. In this sense, the interaction between genotypes with efficient strains of bacteria is the key fact for the success of biological N fixation on grasses ( Lana et al., 2012 ), and can explain the variation in the increase of maize grain yield verified in the literature.

The maize grain yield increased as a result of inoculation with A. brasilense have been commonly attributed to multiple mechanisms, including, but not restricted to, the synthesis of phytohormones (for example auxin, cytokinin and gibberellin), improvement in N nutrition and use, enhancement in leaf photosynthetic parameters, attenuation/minimization of stress, and biological control of some pathogenic agents ( Bashan and de-Bashan, 2010 ). In the present study, the principal evaluations that were positive influenced by inoculation was LCI, stem diameter, ear length, and NUE, which reflected positively in the increase of maize grain yield, and consequently increasing profitability with maize production when inoculated with A. brasilense.

On the subject of the economic analysis, the highest TOCs of the treatments with urea application with NBPT and inoculation with A. brasilense are due to the cost of these agricultural inputs. The average price paid by the farmers was USD $599.33 and $673.40 per ton for urea and urea with NBPT, respectively. For the inoculation with A. brasilense, the expenditure was around USD $3.37 per dose, and two doses were used per hectare in both maize crops, totalizing USD $6.74. As urea with NBPT did not lead to an increase in grain yield, the OP was not favored either however, inoculation with A. brasilense increased grain yield by 15.7% in the average of the 2 yr of cultivation, and due to the low acquisition and application cost (only 0.71% of TOC USD $6.74 per ha), resulted in an increase in OP with maize production, regardless N source and rate applied.

Taking into account the cost of N rates applied and the OP provided by them, the application of 100 kg ha −1 of N as urea source associated with A. brasilense provided higher profitability in maize production (US$ 360.84), while in the absence of inoculation, the highest profitability was obtained without nitrogen fertilization (US$ 174.88), difference of 106.34% in the OP, reiterating the importance of inoculation with A. brasilense increasing NUE, grain yield, and profitability with maize crop.

Brazil is the third largest producer and second largest exporter of maize in the world, with about 16.5 million ha cultivated ( CONAB, 2018 ). Thus, based on the increase in profitability obtained by inoculation with A. brasilense in maize crop, the adoption of this technology by the farmers and due to the large volume and area of production, it is possible to increase the profits obtained with this activity in the order of millions of dollars per year, positively impacting the Brazilian agricultural production system. This can be extrapolated to tropical conditions in the future and spread in several countries, benefiting the world agriculture.

The results obtained demonstrate a benefit in maize grain yield as a function of seed inoculation with A. brasilense. In function of low economic cost, ease of application, non-toxic to the environment, and with a high potential of response from the maize crop, even with the application of N rates considered high for BNF, the inoculation with A. brasilense likely to be a technology increasingly used by farmers.