Academia.eduAcademia.edu

Figure from:Night and day: short-term variation in nitrogen chemistry and nitrous oxide emissions from streams

See full PDFdownloadDownload figure
Table 4 Environmental predictors of N2O concentration, saturation and flux. Results of stepwise forward multiple regression analyses assessing the effects of environmental variables on NxO. NOy + NO; and NO; concentrations are in uM All data are time-weighted over 24-h period. Group A models have NO; + NO; as a combined predictor and include a larger number of sampling periods than group B models, which were run with NO, and NO, available as predictors. Note that a single outlier was excluded from the analysis.

Table 4 Environmental predictors of N2O concentration, saturation and flux. Results of stepwise forward multiple regression analyses assessing the effects of environmental variables on NxO. NOy + NO; and NO; concentrations are in uM All data are time-weighted over 24-h period. Group A models have NO; + NO; as a combined predictor and include a larger number of sampling periods than group B models, which were run with NO, and NO, available as predictors. Note that a single outlier was excluded from the analysis.

Related Figures (12)

Nitrous oxide is produced via nitrification and denitrifi- cation. Both processes can be affected by variation in oxygen (O,), temperature and pH (Miiller, Sundman & Skujins, 1980; Strauss, Mitchell & Lamberti, 2002), which follow diel cycles in many streams (Fig. 1). Low Oy, Fig. 1 Schematic diagram of potential impacts of diel variation in pH, O2 and temperature on rates of nitrification, denitrification and N2O yields. This diagram is based on the conceptual model of Fire- stone & Davidson (1989), which compares N2O production to water leakage from a pipe with holes. Nitrous oxide is released based on both the rate of N cycling (size of pipe) and the NO yield for N cycling processes (size of hole in the pipe). In systems where coupled nitrification—denitrification is important, changes in nitrification rates will have consequences for denitrification rates. Similarly, diel variation in nitrogen concentration may further affect the rates of N cycling and N,O yields.
Nitrous oxide is produced via nitrification and denitrifi- cation. Both processes can be affected by variation in oxygen (O,), temperature and pH (Miiller, Sundman & Skujins, 1980; Strauss, Mitchell & Lamberti, 2002), which follow diel cycles in many streams (Fig. 1). Low Oy, Fig. 1 Schematic diagram of potential impacts of diel variation in pH, O2 and temperature on rates of nitrification, denitrification and N2O yields. This diagram is based on the conceptual model of Fire- stone & Davidson (1989), which compares N2O production to water leakage from a pipe with holes. Nitrous oxide is released based on both the rate of N cycling (size of pipe) and the NO yield for N cycling processes (size of hole in the pipe). In systems where coupled nitrification—denitrification is important, changes in nitrification rates will have consequences for denitrification rates. Similarly, diel variation in nitrogen concentration may further affect the rates of N cycling and N,O yields.
As noted, changes in rates of N cycling and NoO yields can affect N2O concentrations and fluxes. However, other diel environmental changes may also be important (Fig. 2). Diel variation in groundwater input (Lundquist & Cayan, 2002) could drive changes in N2O concentra- tions, as ground water is frequently supersaturated with N.O (Weymann ef al., 2008; Fig. 2). In the absence of changes in groundwater dynamics or altered biological processing, low temperatures at night will lead to decreases in rates of air—-water gas transfer and increased NO solubility, both of which will contribute to decreased NO fluxes (Fig. 2). In addition to direct effects of temperature on piston velocity (also known as gas transfer velocity), wind speed tends to decline at night, which can contribute to decreased piston velocities (Tobias et al. 2009; Fig. 2). The importance of this change will depend on the stream type. Gas fluxes in deeper, slow-moving streams tend to be driven by wind, while benthic turbulence is the primary driver of gas fluxes in shallow, faster-flowing systems (Schwarzenbach, Gschwend & Imboden, 1993; Raymond & Cole, 2001). Fig. 2 Conceptual diagram of the effects of common diel physical changes on excess N2O and gas transfer coefficient. The size of the excess N»O pool is represented as a cylinder (excess N2O is defined as measured concentration minus saturation concentration). A smaller cylinder relative to the daytime pool represents a smaller excess N»O pool, and a larger cylinder represents the converse. Changes in gas transfer coefficient (K) are reflected by changes in the size and direction of the arrow. This figure represents only direct effects (not longer-term impacts of changes in rates of gas transfer on NO pool sizes). Effects of changes in groundwater inflows on gas transfer coefficient are expected to be system specific.
As noted, changes in rates of N cycling and NoO yields can affect N2O concentrations and fluxes. However, other diel environmental changes may also be important (Fig. 2). Diel variation in groundwater input (Lundquist & Cayan, 2002) could drive changes in N2O concentra- tions, as ground water is frequently supersaturated with N.O (Weymann ef al., 2008; Fig. 2). In the absence of changes in groundwater dynamics or altered biological processing, low temperatures at night will lead to decreases in rates of air—-water gas transfer and increased NO solubility, both of which will contribute to decreased NO fluxes (Fig. 2). In addition to direct effects of temperature on piston velocity (also known as gas transfer velocity), wind speed tends to decline at night, which can contribute to decreased piston velocities (Tobias et al. 2009; Fig. 2). The importance of this change will depend on the stream type. Gas fluxes in deeper, slow-moving streams tend to be driven by wind, while benthic turbulence is the primary driver of gas fluxes in shallow, faster-flowing systems (Schwarzenbach, Gschwend & Imboden, 1993; Raymond & Cole, 2001). Fig. 2 Conceptual diagram of the effects of common diel physical changes on excess N2O and gas transfer coefficient. The size of the excess N»O pool is represented as a cylinder (excess N2O is defined as measured concentration minus saturation concentration). A smaller cylinder relative to the daytime pool represents a smaller excess N»O pool, and a larger cylinder represents the converse. Changes in gas transfer coefficient (K) are reflected by changes in the size and direction of the arrow. This figure represents only direct effects (not longer-term impacts of changes in rates of gas transfer on NO pool sizes). Effects of changes in groundwater inflows on gas transfer coefficient are expected to be system specific.
Fig. 3 Location of study sites. The large map shows Ontario and Quebec, Canada. Inset area shows sites, which are from west to east: A: Boyne River (tributary), B: Beeton Creek, C: Penville Creek, D: Innisfil Creek, E: Schomberg Creek (tributary), F: Black River, G: Stouffville Creek, H: Layton Creek, I: Mariposa Brook (upstream), J: Mariposa Brook (downstream), K: Jackson Creek (upstream), L: Jackson Creek (downstream). We selected 10 streams located north and north-east of Toronto, Canada (Fig. 3). This region is underlain by sedimentary rocks of the Phanerozoic age, and soils are dominated by grey-brown luvisols and melanic brunisols (Baldwin, Desloges & Band, 2000). Study sites were selected to help complement ongoing work on gas and carbon dynamics (Baulch etal., 2011; Wilson & Xenopoulos, 2009). None of the sites is affected by point-source inputs. Although much of this region was historically covered by mixed forests, agriculture is now a large proportion of land use. Annual crops constitute 13- 42% of land cover. Mixed rotations between perennial (hav or pasture) and annual crops cover 5—31%, wooded
Fig. 3 Location of study sites. The large map shows Ontario and Quebec, Canada. Inset area shows sites, which are from west to east: A: Boyne River (tributary), B: Beeton Creek, C: Penville Creek, D: Innisfil Creek, E: Schomberg Creek (tributary), F: Black River, G: Stouffville Creek, H: Layton Creek, I: Mariposa Brook (upstream), J: Mariposa Brook (downstream), K: Jackson Creek (upstream), L: Jackson Creek (downstream). We selected 10 streams located north and north-east of Toronto, Canada (Fig. 3). This region is underlain by sedimentary rocks of the Phanerozoic age, and soils are dominated by grey-brown luvisols and melanic brunisols (Baldwin, Desloges & Band, 2000). Study sites were selected to help complement ongoing work on gas and carbon dynamics (Baulch etal., 2011; Wilson & Xenopoulos, 2009). None of the sites is affected by point-source inputs. Although much of this region was historically covered by mixed forests, agriculture is now a large proportion of land use. Annual crops constitute 13- 42% of land cover. Mixed rotations between perennial (hav or pasture) and annual crops cover 5—31%, wooded
NM indicates that the variable was not measured. *Average of two values closest and furthest from solar noon. *DOC values represent the mean of all sampling occasions, with the exception of the streams sampled 9-10 July 2007, where we report daytim only values. Table 1 Chemical and physical characters of study streams during diel sampling
NM indicates that the variable was not measured. *Average of two values closest and furthest from solar noon. *DOC values represent the mean of all sampling occasions, with the exception of the streams sampled 9-10 July 2007, where we report daytim only values. Table 1 Chemical and physical characters of study streams during diel sampling
Weather data were obtained from the National Climate Data and Information Archive, 15 July 2009 http://www.climate.weatheroffice.ec.gc.ca. *45.1 mm was reported at the Barrie station (nearest the sampling sites) on the day preceding diel sampling. Table 2 Rain during sampling and mean cumulative precipitation in the 1- and 2-week periods preceding diel sampling from up to five weather stations (where data were available) near our sampling sites
Weather data were obtained from the National Climate Data and Information Archive, 15 July 2009 http://www.climate.weatheroffice.ec.gc.ca. *45.1 mm was reported at the Barrie station (nearest the sampling sites) on the day preceding diel sampling. Table 2 Rain during sampling and mean cumulative precipitation in the 1- and 2-week periods preceding diel sampling from up to five weather stations (where data were available) near our sampling sites
Fig. 4 Diel variation in Oj, NO and N chemistry at two sites in Jackson Creek: upstream and downstream of a wetland complex. Nitrate, NO;, N2O and Q) are plotted as filled circles, and organic N and NHj as open squares. Saturation concentrations of O2 and N2O are shown as grey lines. Shaded boxes indicate the period between sunset and sunrise. Error bars (on N2O plots) indicate standard deviation.
Fig. 4 Diel variation in Oj, NO and N chemistry at two sites in Jackson Creek: upstream and downstream of a wetland complex. Nitrate, NO;, N2O and Q) are plotted as filled circles, and organic N and NHj as open squares. Saturation concentrations of O2 and N2O are shown as grey lines. Shaded boxes indicate the period between sunset and sunrise. Error bars (on N2O plots) indicate standard deviation.
Fig. 5 Diel variation in Oz, N2O and N chemistry in Mariposa Brook. Nitrate, NO; , N2O and QO, are plotted as filled circles, and organic N and NHj as open squares. For 2006 data, NO; reflects NO; + NO,. Saturation concentrations of O2 and N2O are shown as grey lines. Shaded boxes indicate night-time. Error bars (on NO plots) indicate standard deviation. Note that two sites on Mariposa Brook were sampled in July 2007.
Fig. 5 Diel variation in Oz, N2O and N chemistry in Mariposa Brook. Nitrate, NO; , N2O and QO, are plotted as filled circles, and organic N and NHj as open squares. For 2006 data, NO; reflects NO; + NO,. Saturation concentrations of O2 and N2O are shown as grey lines. Shaded boxes indicate night-time. Error bars (on NO plots) indicate standard deviation. Note that two sites on Mariposa Brook were sampled in July 2007.
Data with P < 0.10, 0.05 and 0.01 are indicated with *, ** and ***. For NxO-N/NO; -N, we also report values for net N2O-emitting source streams only (indicated in brackets). Table 3 Differences in water chemistry, temperature, N2O and modelled nitrification factor between solar noon (average of two data point: closest to solar noon) and night (average of two points furthest from solar noon)
Data with P < 0.10, 0.05 and 0.01 are indicated with *, ** and ***. For NxO-N/NO; -N, we also report values for net N2O-emitting source streams only (indicated in brackets). Table 3 Differences in water chemistry, temperature, N2O and modelled nitrification factor between solar noon (average of two data point: closest to solar noon) and night (average of two points furthest from solar noon)
Fig. 6 Diel variation in Oz, N.O and N chemistry in the Black River. Nitrate, NO;, N2O and O, are plotted as filled circles, and organic N and NHj as open squares. Saturation concentrations of O2 and N2O are shown as grey lines. Shaded boxes indicate night-time. Error bars (on N2O plots) indicate standard deviation.
Fig. 6 Diel variation in Oz, N.O and N chemistry in the Black River. Nitrate, NO;, N2O and O, are plotted as filled circles, and organic N and NHj as open squares. Saturation concentrations of O2 and N2O are shown as grey lines. Shaded boxes indicate night-time. Error bars (on N2O plots) indicate standard deviation.
Fig. 7 Diel variation in NO flux across all streams. Hours are indicated on the x-axis, and night-time is indicated by shaded boxes. Nitrous oxide fluxes did not vary systematically over diel periods (Fig. 7; Table 3), and diel patterns differed even within a single stream across either multiple expe- ditions (Mariposa Brook, Black River) or sites (Mariposa Brook, Jackson Creek; Fig. 7). It should be noted that mean fluxes (Table 3) were strongly influenced by high fluxes at a single site (Beeton Creek) and data from four other sites may have been affected by changes in hydrology following a storm event (9-10 July sampling; see Wilson, 2009 for details). When data were reanalysed having excluded potentially affected sites (Beeton, Boyne, Schom- berg, Penville and Innisfil), the restricted suite of study streams showed higher nocturnal NO fluxes than by day Nitrous oxide was supersaturated in the majority of streams and typically remained so throughout the diel cycle (Figs 5, 6 and S1-S3). Beeton Creek had the highest NO, +NO, concentrations of all of the study streams and the highest concentrations of dissolved NO. Nitrite concentrations were insignificant in comparison with
Fig. 7 Diel variation in NO flux across all streams. Hours are indicated on the x-axis, and night-time is indicated by shaded boxes. Nitrous oxide fluxes did not vary systematically over diel periods (Fig. 7; Table 3), and diel patterns differed even within a single stream across either multiple expe- ditions (Mariposa Brook, Black River) or sites (Mariposa Brook, Jackson Creek; Fig. 7). It should be noted that mean fluxes (Table 3) were strongly influenced by high fluxes at a single site (Beeton Creek) and data from four other sites may have been affected by changes in hydrology following a storm event (9-10 July sampling; see Wilson, 2009 for details). When data were reanalysed having excluded potentially affected sites (Beeton, Boyne, Schom- berg, Penville and Innisfil), the restricted suite of study streams showed higher nocturnal NO fluxes than by day Nitrous oxide was supersaturated in the majority of streams and typically remained so throughout the diel cycle (Figs 5, 6 and S1-S3). Beeton Creek had the highest NO, +NO, concentrations of all of the study streams and the highest concentrations of dissolved NO. Nitrite concentrations were insignificant in comparison with
Fig. 8 Relationship between mean daily NO; + NO; concentration and mean daily N2O flux. Error bars show minimum and maximum flux and concentrations. Grey circles show point measurements. Data from Beeton Creek are not shown on this plot, owing to high fluxes.
Fig. 8 Relationship between mean daily NO; + NO; concentration and mean daily N2O flux. Error bars show minimum and maximum flux and concentrations. Grey circles show point measurements. Data from Beeton Creek are not shown on this plot, owing to high fluxes.
Fig. 9 Ratios of N2O-N concentrations to NO; + NO, concentrations in study streams over day-night sampling periods. Night-time is indicated by shaded boxes.
Fig. 9 Ratios of N2O-N concentrations to NO; + NO, concentrations in study streams over day-night sampling periods. Night-time is indicated by shaded boxes.