IntroductionTop

In the last few decades, the study of N₂O has acquired greater importance due to its contribution to global climate change. N₂O is an important long-lived greenhouse gas in terms of radiative forcing (0.17±0.03 W m^–2) (Myhre et al. 2013Myhre G., Shindell D., Bréon F.M., et al. 2013. Anthropogenic and Natural Radiative Forcing. In: Stocker T.F., Qin D., Plattner G.K., et al. (eds), Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, pp. 659-740.) and represents the major anthropogenic contributor to stratospheric ozone destruction. It has a long atmospheric lifetime of 131±10 years (Prather et al. 2012Prather M.J., Holmes C.D., Hsu J. 2012. Reactive greenhouse gas scenarios: Systematic exploration of uncertainties and the role of atmospheric chemistry. Geophys. Res. Lett. 39: L09803. ) and its global warming potential is 310 times greater than that of carbon dioxide, in a time horizon of 100 years. In 2011 atmospheric N₂O levels (324.2±0.1 ppb) exceeded the pre-industrial levels (270±7 ppb) by about 20% (Myhre et al. 2013Myhre G., Shindell D., Bréon F.M., et al. 2013. Anthropogenic and Natural Radiative Forcing. In: Stocker T.F., Qin D., Plattner G.K., et al. (eds), Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, pp. 659-740.), largely due to increased agricultural activity and industry.

Estuaries have been considered significant N₂O contributors to the atmosphere as a consequence of their high productivity and anthropogenic nitrogen loadings. N₂O is mainly formed during the first step of nitrification, the aerobic oxidation of ammonium (NH₄⁺) to nitrite (NO₂^–), mediated by ammonia-oxidizing bacteria (AOB) and ammonia-oxidizing archaea (AOA), and microbiological denitrification, the biological reduction of nitrate (NO₃^–) to N₂O, and, in turn, nitrogen gas (N₂). Nitrification and denitrification often occur simultaneously in aquatic ecosystems and their relative contribution to total N₂O production is difficult to disentangle. As nitrification is an aerobic process in well-oxygenated estuarine systems, the water column mostly contributes N₂O through nitrification production (de Wilde and de Bie 2000de Wilde H.P.J., de Bie M.J.M. 2000. Nitrous oxide in the Schelde estuary: Production by nitrification and emission to the atmosphere. Mar. Chem. 69: 203-216., Barnes and Upstill-Goddard 2011Barnes J., Upstill-Goddard R.C. 2011. N₂O seasonal distributions and air-sea exchange in UK estuaries: Implications for the tropospheric N₂O source from European coastal waters. J. Geophys. Res. B. 116: G01006.). Denitrification is usually limited to zones under hypoxic conditions (DO <2 ml L^–1), although some denitrification may occur even in relatively oxygenated waters (de Bie et al. 2002de Bie M.J.M., Middelburg J.J. Starink M., et al. 2002. Factors controlling nitrous oxide at the microbial community and estuarine scale. Mar. Ecol. Prog. Ser. 240: 1-9. ).

Both nitrification and denitrification are sensitive to the ongoing environmental changes and any natural or anthropogenic-induced shifts in the N availability have the potential to alter nitrogen cycling in coastal environments and affect N₂O formation and release to the atmosphere (Bange et al. 2010Bange H.W., Freing A., Kock A., et al. 2010. Marine pathways to nitrous oxide. In: Smith K.A. (ed.), Nitrous oxide and climate change. Earthscan, London. U.K. pp. 36-62.). The extent of the denitrification is also strongly controlled by temperature, nitrate concentrations and the availability of organic carbon (e.g. Dong and Nedwell 2006Dong L.F., Nedwell D.B. 2006. Sources of nitrogen used for denitrification and nitrous oxide formation in sediments of the hypernutrified Colne, the nutrified Humber and oligotrophic Conwy Estuaries, United Kingdom. Limnol. Oceanogr. 51(1, part 2): 545-557.). In addition to the supply of oxygen and ammonia, which are the main controls on nitrification, other environmental variables may affect this biological process: temperature (Dai et al. 2008Dai M., Wang L., Guo X., et al. 2008. Nitrification and inorganic nitrogen distribution in a large perturbed river/estuarine system: the Pearl River Estuary, China. Biogeosci. 5: 1227-1244.), salinity (Bollmann and Laanbroek 2002Bollmann A., Laanbroek H.J. 2002. Influence of oxygen partial pressure and salinity on the community composition of ammonia-oxidizing bacteria in the Schelde estuary. Aquat. Microb. Ecol. 28: 239-247. ) and pH (Strauss et al. 2002Strauss E.A., Mitchell N.L., Lamberti G.A. 2002. Factors regulating nitrification in aquatic sediments: effects of organic carbon, nitrogen availability, and pH. Can. J. Fish. Aquat. Sci. 59: 554-563. ).

Estimates of N₂O release from estuaries to the global inventory reveal wide uncertainties due to the large variability in N₂O data (Bange et al. 2010Bange H.W., Freing A., Kock A., et al. 2010. Marine pathways to nitrous oxide. In: Smith K.A. (ed.), Nitrous oxide and climate change. Earthscan, London. U.K. pp. 36-62., Barnes and Upstill-Goddard 2011Barnes J., Upstill-Goddard R.C. 2011. N₂O seasonal distributions and air-sea exchange in UK estuaries: Implications for the tropospheric N₂O source from European coastal waters. J. Geophys. Res. B. 116: G01006.). However, considerable efforts have been made in the last few decades to better understand the nitrogen cycle, the dynamics of N₂O production and the quantification of the respective emission from European estuaries. More recently Murray et al. (2015)Murray R.H., Erler D.V., Eyre B.D. 2015. Nitrous oxide fluxes in estuarine environments: response to global change. Global Change Biol. 21: 3219-3245. reviewed N₂O global fluxes from estuarine environments and reported a variation of 0.17-0.95 Tg N₂O-N yr^–1.

Studies on the N₂O dynamics and fluxes have been carried out in the Portuguese Tagus, Sado and Douro estuaries (e.g. Gonçalves et al. 2010Gonçalves C., Brogueira M.J., Camões M.F. 2010. Seasonal and tidal influence on the variability of nitrous oxide in the Tagus estuary, Portugal. Sci. Mar. 74S1: 57-66. , 2015Gonçalves C., Brogueira M.J, Nogueira M. 2015. Tidal and spatial variability of nitrous oxide (N₂O) in Sado estuary (Portugal). Est. Coast. Shelf Sci. 167: 466-474, Teixeira et al. 2013Teixeira C., Magalhães C., Joye S.B., et al. 2013. The role of salinity in shaping dissolved inorganic nitrogen and N₂O dynamics in estuarine sediment-water interface. Mar. Pollut. Bull. 66: 225-229. ). However, no data on N₂O levels and fluxes are available for the Minho and Lima estuaries. In this work, we (1) report spatial variability of N₂O concentration in these systems, (2) assess the contribution of different N₂O sources, (3) evaluate the role of environmental properties on the increment of N₂O fluxes, and (4) estimate N₂O emission in a perspective of global N₂O estuarine emissions.

MATERIALS AND METHODSTop

Area description

The Minho and Lima estuaries, both situated in the northern part of Portugal (Fig. 1), differ essentially in terms of river discharge, anthropogenic pressures and morphology (Table 1).

Minho estuary

The Minho estuary is situated in the border region between Portugal (Minho region) and Spain (Galicia region). The estuary has an area of approximately 23 km², a mean depth of 2.6 m, and a maximum width of about 2 km at the confluence of the Coura River, at Caminha. The estuary is mesotidal, with a mean tidal range of 3 m, and its dynamic tidal effects extends 35 km upstream (Bettencourt et al. 2003Bettencourt A., Ramos L., Gomes V., et al. 2003. Estuários Portugueses. Editions INAG - Ministério das Cidades, Ordenamento do Território e Ambiente, Lisboa, 311 pp.).

The annual mean discharge of the Minho River is 300 m³ s^–1 (Table 1), ranging between approximately 100 m³ s^–1 in August and 800 m³ s^–1 in February, and the water residence time is 1.5 days (Sousa et al. 2013Sousa C., Vaz M.N., Alvarez I., et al. 2013. Effect of Minho estuarine plume on Rias Baixas: numerical modeling approach. J. Coast. Res. Spec. 65: 2059-2065.). The estuary has high ecological value, mainly due to the large diversity of habitats and biodiversity and great socio-economic importance from tourism, fishing and agriculture. However, increasing pressure has been detected over recent years, involving in particular the pollution of surface water from both specific and diffuse sources, morphological alterations, changes of land use in the drainage basin and other impacts from human activity, such as aquaculture, textile, rubber and plastic processing industries (Ferreira et al. 2005Ferreira J.G., Nobre A.M., Simas T.C., et al. 2005. Monitoring plan for water quality and ecology for Portuguese traditional and coastal waters: Development of guidelines for the application of the European Union Water Framework Directive. Editions INAG-Instituto da Água IMAR-Institute of Marine Research. Lisbon, Portugal, 142 pp.). The Minho and Coura rivers are the main nitrogen sources to the estuary, contributing 13000 t N yr^–1 and 7000 t N yr^–1, respectively, and effluents from domestic origin contribute 51 t N yr^–1 (Ferreira et al. 2005Ferreira J.G., Nobre A.M., Simas T.C., et al. 2005. Monitoring plan for water quality and ecology for Portuguese traditional and coastal waters: Development of guidelines for the application of the European Union Water Framework Directive. Editions INAG-Instituto da Água IMAR-Institute of Marine Research. Lisbon, Portugal, 142 pp.) (Table 1).

Lima estuary

The Lima estuary, located south of Minho, has an area of approximately 5 km² and a mean depth of 4 m (Table 1). It has a semidiurnal mesotidal regime and the tidal effect extends 20 km upstream (Ramos et al. 2006Ramos S., Cowen R.K., Ré P., et al. 2006. Temporal and spatial distributions of larval fish assemblages in the Lima estuary (Portugal). Est. Coast. Shelf Sci. 66: 303-314. ). The upper estuary is constituted by a narrow channel with some intertidal areas and undisturbed banks. The middle estuary is a shallow salt marsh zone, and the lower estuary consists of a wider and shallow basin that communicates with the sea via a deep, narrow channel (with a typical depth of about 10 m). The estuary mouth is artificially obstructed by a 2-km-long jetty. The Lima River has an annual freshwater discharge of 54 m³ s^–1, and the mean water residence time in the estuary is one day (Table 1).

The lower estuary is a highly urbanized zone and has been subjected to several anthropogenic impacts resulting from harbour activities (Viana do Castelo). However, nutrient loadings also originate from diffuse sources, largely agriculture (0.51 t N yr^–1), although the main N source to the estuary is the Lima River (1077 t N yr^–1) (Ferreira et al. 2005Ferreira J.G., Nobre A.M., Simas T.C., et al. 2005. Monitoring plan for water quality and ecology for Portuguese traditional and coastal waters: Development of guidelines for the application of the European Union Water Framework Directive. Editions INAG-Instituto da Água IMAR-Institute of Marine Research. Lisbon, Portugal, 142 pp.). This N load is, however, approximately 19 times lower than the load to the Minho estuary (Table 1).

Sampling

Water sampling was undertaken in September 2006 during ebb tide (at spring tide), at nine stations located along a main transect of both estuaries, from the upstream limit of tidal influence to the estuary mouth, covering a full range of salinity of 0-30 corresponding to a distance of 26.5 km in Minho and 15 km in Lima estuary (Fig. 1).

Surface water (0.2 m depth) was collected using 2-L Niskin bottles (General Oceanics) for analysis of salinity (S), temperature (T), pH, dissolved inorganic nitrogen (nitrate NO₃^–, nitrite NO₂^– and ammonium NH₄⁺), dissolved oxygen (DO) and nitrous oxide (N₂O). The hydrological characteristics of the Minho and Lima estuaries and the meteorological conditions observed during the sampling period are presented in Table 2.

Analytical procedure

Water temperature (T) was measured in situ with a Seabird SBE19/CTD probe with an accuracy of 0.01°C.

Salinity (S) measurements were carried out using a temperature-controlled Guideline Salinometer (Portasal 8410A), and accuracy was 0.03 salinity. Equipment was calibrated with a certified IAPSO Standard Seawater reference.

Meteorological parameters (air temperature, pressure, and wind speed and direction) were determined using a portable meteorological station (Campbell Scientific CR510). Measurements represent the average of physical parameters taken using a sampling time of 5 seconds and a storage time of 1 minute. Wind speed 10-minute average was determined for each sampling station and converted to wind speed values at 10 m height (u₁₀) using a logarithmic correction (Hartman and Hammond 1985Hartman B., Hammond D. 1985. Gas exchange in San Francisco Bay. Hydrobiology 129: 59-68.).

The Minho and Lima River discharges were calculated as an average of the flow 10 days before sampling, at the hydrometric stations of Foz de Mouro and Ponte da Barca, respectively (SNIRH 2013SNIRH (Sistema Nacional de Informação de Recursos Hídricos). 2013. Instituto da Água IP.).

Dissolved oxygen (DO, μmol L^–1) was measured using whole-bottle Winkler’s titration method (Aminot and Chaussepied 1983Aminot A., Chaussepied M. 1983. Manuel des analyses chimiques en milieu marin. Centre National pour l’Exploitation des Océans. CNEXO, Brest, France, 395 pp.). A Methrom titrator was used to dispense small amounts of thiosulphate, and starch endpoint was detected visually. Precision of the method was in the range of 0.08% to 0.25%. DO saturation, expressed in percentage (%), was determined as the ratio of the oxygen concentration determined and the equilibrium values of DO calculated with the Weiss (1970)Weiss R.F. 1970. The solubility of nitrogen, oxygen, and argon in water and seawater. Deep-Sea Res. Oceanogr. Abstr. 17: 721-735. equation. Apparent oxygen utilization (AOU, μmol L^–1) was calculated as the difference between the saturation oxygen concentration and the dissolved oxygen concentration measured in the sample.

Water samples for determination of dissolved inorganic nitrogen were filtered through acetate cellulose filters (pore size 0.45 μm) and stored at –20ºC until analysis. Analyses were carried out using a Traacs autoanalyser following colorimetric techniques outlined by the manufacturer. Estimated precision was ±0.8% for nitrate and nitrite (NO₃^– and NO₂^–) and ±2.0% for ammonium (NH₄⁺), at mid-scale concentrations. Accuracy of nutrient measurements was maintained by using CSK Standards (WAKO, Japan).

pH measurement was carried out immediately after collecting water samples using a Metrohm 704 pH-meter and a combined electrode (Metrohm), standardized against NBS buffers (6.865 and 9.180 pH). Precision of pH measurements was ±0.01.

Water samples for determination of N₂O were collected in triplicate in 20-mL glass headspace vials and poisoned with saturated aqueous mercury chloride (HgCl₂) to stop biological activity. The vials were stored upside down, in the dark, at 4°C in the refrigerator until analysis, performed within 10 days. Dissolved N₂O was determined by a headspace equilibration technique coupled with gas chromatographic analysis (GC-3800, Varian). Briefly, 20 mL of sample was equilibrated with 5 mL of highly purified helium (purity =99.9999%) in a headspace CombiPAL autosampler. Gas chromatographic separation was carried out using a stainless steel column packed with 80/100 (mesh) Porapak. Oven and detector temperature was set at 50°C and 320°C, respectively, and high purity nitrogen (99.9999%) was used as the carrier gas (flow rate 30 mL min^–1). To remove water vapour and carbon dioxide, absorbent columns packed respectively with Mg(ClO₄)₂ and Carbosorb were located in the carrier gas line between the sample loop and the separation column. N₂O peak was detected with a ⁶³Ni electron capture detector (ECD). Calibration of ECD response was performed using standard gas mixtures with 400, 780 and 1980 ppb N₂O in synthetic air (Air Liquide), and method precision was 2.6% (30 replicate measurements using samples containing 10 nmol L^–1 of N₂O). In situ concentration of N₂O (C, nmol L^–1) was calculated from the concentrations measured in the headspace according to the solubility equation of Weiss and Price (1980)Weiss R.F., Price B.A. 1980. Nitrous oxide in water and seawater. Mar. Chem. 8: 347-359. :

C = β (TS) x′P

where x′ is the measured N₂O dry mole fraction, P is the atmospheric pressure, and β is the solubility coefficient, which is a function of the water temperature (T) and salinity (S). N₂O equilibrium concentrations were calculated assuming an atmospheric N₂O mixing ratio of 320.1 ppb (WMO 2006WMO (World Meteorological Organization). 2006. World Meteorological Organization greenhouse gas bulletin: The State of Greenhouse Gases in the Atmosphere Using Global Observations through 2006. Bulletin No. 3. Global Atmosphere Watch, Geneva, Switzerland, 4 pp.).

N₂O saturation, expressed in percentage (%), was determined as the ratio between the measured dissolved N₂O concentration and the equilibrium concentration. The N₂O water-air flux (F_N2O) was estimated according to the following equation:

F_N2O = k_N2O ΔN₂O

where ΔN₂O, the excess of N₂O, is the difference between the measured concentration and the equilibrium concentration with the atmosphere in the estuarine water at the local temperature and salinity; and k_N2O (cm h^–1) is the N₂O transfer velocity, which is expressed as a function of the wind speed and the Schmidt number (Sc). Since no direct measurements of k_N2O were made in the Minho and Lima estuaries, both the k-wind relationships of Carini et al. (1996)Carini S., Weston N., Hopkinson C., et al. 1996. Gas exchange rates in the Parker River estuary, Massachusetts. Biol. Bull. 191(2): 333-334. (hereinafter referred to as C96) and Raymond and Cole (2001)Raymond P.A., Cole J.J. 2001. Gas exchange in rivers and estuaries: choosing a gas transfer velocity. Estuaries 24: 312-317. (hereinafter referred to as RC01) were, respectively, used to compute k:

k_C96 = 0.045 + 2.0277u₁₀

k_RC01 = 1.91 e^0.35u10

The gas transfer velocities and air-sea fluxes were estimated using in situ wind speeds normalized to 10 m height (u₁₀). The k coefficients were corrected for in situ temperature using the following relationship:

k_N2O /k₆₀₀ = (Sc_N2O /600)^–0.5

where Sc_N2O is the Schmidt number for N₂O calculated according to the equation of Wanninkhof (1992)Wanninkhof R. 1992. Relationship between wind speed and gas exchange over the ocean. J. Geophys. Res. C 97(C5): 7373-7382.:

Sc = 2301.1 – 151.1 T + 4.7364 T² + 0.057431 T³

where T is the temperature (ºC).

Statistical analysis

An unpaired t-test was used to identify statistical differences in levels of N₂O and other environmental variables between and along estuaries.

Pearson’s correlation analyses were performed to evaluate the existence of relationships between ΔN₂O and the variables NH₄⁺, AOU, NO₂^– and pH assumed to be connected with production pathways of this biogas.

The multivariate techniques principal component analysis (PCA) and cluster analysis were applied to environmental data in order to identify and compare inter-relationships between these variables in both estuaries. Data were log(x+1) transformed and handled using correlation-based PCA on the basis of standard Euclidean distance between samples to define their dissimilarity. PRIMER (version 6) was employed for the multivariate analysis.

Figures were created in Golden Software Grapher program (version 9.6.1001).

RESULTSTop

N₂O levels and fluxes

Concentrations of N₂O and the studied environmental parameters plotted against salinity along both Minho and Lima estuaries are shown in Figure 2.

Distribution of N₂O exhibits a declining tendency towards the mouths of both estuaries. In general, concentrations were higher in Lima than in Minho (Fig. 2A). In Lima values ranged from 10.1 to 20.0 nmol L^–1 and were below the conservative mixing line. The maximum value was reached in the upper estuary, suggesting that the Lima River was the main source of N₂O to the estuarine system. From salinity 0 to 4 a sharp decrease was observed, suggesting a greater N₂O loss in the zone, presumably through water-air gas exchange. Downstream of salinity 4, N₂O decreased more slightly and the mixing with the N₂O-poorer seawater is well perceptible. In the Minho estuary, concentrations varied from 8.6 to 14.4 nmol L^–1 and by contrast with those observed in Lima were above the conservative mixing line, pointing to the existence of N₂O sources within the estuary. A major concentration increase from 10.5 to 14.4 nmol L^–1 was detected between 0 and 2.5 salinity, suggesting the existence of internal N₂O sources in this zone of the estuary, presumably from manufacturing industries located nearby. N₂O saturation values ranged from 101% to 166% along the Minho estuary and from 113% to 227% along the Lima estuary (see Table 4), indicating that both estuaries are potential N₂O sources to the atmosphere.

Surface waters of both estuaries were well oxygenated during our study period. Concentrations of DO were higher than 280 µmol L^–1 in the upper Lima estuary at salinity 0.2, decreasing to 240 µmol L^–1 at salinity 2.8 (Fig. 2C). Seawards of this salinity, an increasing tendency was observed and a concentration of 255 µmol L^–1 was also reached in the vicinity of the estuary mouth (salinity 30.2). In the Minho estuary, a decrease in DO was also detected in the upper estuary and the concentration dropped from 243 µmol L^–1 at 0.1 salinity to 199 µmol L^–1 at 2.3 salinity. Afterwards, a sharp increase in DO was measured along the estuary and a maximum value of 264 µmol L^–1 was reached at the estuary mouth (salinity 29.7). DO-enriched seawater probably accounted for the similar increasing trend seaward in both estuaries. Saturation values were higher than 70% in the Minho estuary and 90% in the Lima estuary.

In the Minho estuary pH increased from 7.3 in the more river-influenced zone to 8.1 at the estuary mouth (Fig. 2D). In the Lima estuary pH showed a larger range of values, increasing from 6.7 in the most river-influenced site to 8.0 at the estuary mouth.

NO₃^– was the dominant species of inorganic nitrogen in both estuaries, reaching a similar maximum concentration in the river input (47.6 µmol L^–1 in Minho and 44.4 µmol L^–1 in Lima) (Fig. 2E). Values decreased seawards and in general followed the theoretical conservative mixing line. Along the Minho estuary both NO₂^– and NH₄⁺ exhibited an irregular behaviour, though the system seemed to function as an NO₂^– source and an NH₄⁺ sink (Fig. 2F, G). Between salinity 0 and ~5-7 the decline in NH₄⁺ (~4.0 to 0.5 µM L^–1) was simultaneous with an increase in NO₂^– (~0.6 to 1.0 µM L^–1) and N₂O (~10 to 13-14 nmol L^–1), suggesting the occurrence of nitrification.

Figure 3 displays relationships between ΔN₂O and AOU, NH₄⁺, NO₂^–, NO₃^– and pH in the Minho estuary. In the mentioned salinity zone (0 and ~5-7) a significant positive correlation was found between ΔN₂O and AOU (R²=0.75). This indicates the occurrence of nitrification as a source of N₂O and the respective slope provides an estimate of the biological N₂O yield per mole O₂ consumed (Yoshinari 1976). Further, the simultaneous (negative) correlations between ΔN₂O and the primary substrate for nitrification, NH₄⁺ (R²=0.40), and pH (R²=0.80), and the positive correlation between ΔN₂O and the byproduct of nitrification, NO₂^– (R²=0.41), are consistent with the predominance of nitrification as a mechanism of N₂O production in the upper part of the Minho estuary. No correlation was found with NO₃^–_, whose high concentrations were mostly riverine derived.

In the Lima estuary no relationships were found, suggesting the occurrence of nitrification (Fig. 4).

The application of PCA to the studied environmental variables in the Minho and Lima estuaries allowed us to identify two main composite variables, PC1 and PC2 (eigenvalues >1.0), which explain 82% of the variance (Table 3) and represent a good description of the environmental structure across the estuarine sampled sites.

PC1 explained 53% of variance and had the highest positive loading for NO₃^–, T and N₂O and a negative loading for S and pH (Fig. 5). This component represents the separation of major river-influenced stations from major marine-influenced ones in both estuaries. PC2 explained 27% of the variance and correlated positively with DO and negatively with NO₂^– and NH₄⁺. This component appears to represent relevant parameters to N dynamics, particularly in Minho estuary. In fact, projection of stations along PC2 reveals a clear separation of sites from the upper Lima estuary (L1-L5) (Fig. 5), mostly associated with higher values of N₂O, and sites from the upper Minho (M1-M5) more associated with higher NO₂^– and NH₄⁺, apparently from river origin, as the N load from the Minho River is considerable at this point (Table 1). It was also observed that the stations from both middle/lower estuaries did not differ in terms of studied environmental variables.

N₂O water-air fluxes are shown in Figure 6. Positive values prevailed at all stations but decreased, in general, between the upper and lower zone of the estuaries. This tendency was more pronounced in the Lima estuary, where higher N₂O fluxes in the river-influenced area were about twice (~12.0 μmol m^–2 d^–1; C96) (Fig. 6B) those observed in the upper part of the Minho estuary (~6.0 μmol m^–2 d^–1; C96) (Fig. 6A). The higher fluxes in the Lima estuary were mainly associated with the higher levels of N₂O observed in the upper estuary area (St.1 to St.3; Fig. 2), indicating that the low salinity zone (0-5) is an important source of N₂O to the atmosphere.

Using the two different parameterizations (C96, RC01) to calculate the gas transfer coefficients, the averaged N₂O water-air fluxes from Minho estuary to the atmosphere ranged between 4.0±3.3 µmol m^–2 d^–1 (RC01) and 4.1±2.8 µmol m^–2 d^–1 (C96), corresponding to a mean N₂O concentration of 11.3±1.3 nmol L^–1 (132±22% saturation) and a mean wind speed of 3.9±0.1 m s^–1. Slightly higher N₂O water-air fluxes were found in the Lima estuary, with averaged values ranging between 4.7±1.9 µmol m^–2 d^–1 (RC01) and 5.0±2.0 µmol m^–2 d^–1 (C96), corresponding to a higher mean N₂O concentration of 13.7±1.6 nmol L^–1 (153±26% saturation) and a lower mean wind speed level (2.4±0.1 m s^–1).

N₂O fluxes from the Minho and Lima estuaries were regressed versus the first two PCs’ ordination of station scores to test their ability to predict the fluxes. We found out that only PC1 showed to be correlated with N₂O flux, with a strong positive correlation (R²=0.61) (Fig. 7). N₂O flux also increased along a gradient of increasing NO₃^–, T and N₂O. These results suggest that future global changes in these parameters will result in an increase of N₂O flux in these estuarine systems.

DiscussionTop

The present study reveals that the Minho and Lima estuaries, particularly the upper reaches, behave differently regarding N₂O levels, sources and fluxes. N₂O distribution exhibits a pronounced spatial variability in both estuaries but in the Lima estuary concentrations are higher than in Minho. The Lima River was the main N₂O contributor to the Lima estuary, whereas the occurrence of nitrification seems to represent an additional N₂O source within the Minho estuary. As NH₄⁺ is a primary substrate for nitrification, low NH₄⁺ concentration may limit nitrification. It has been suggested that the AOA and AOB niches are defined by ammonium concentrations (Martens-Habbena et al. 2009Martens-Habbena W., Berube P.M., Urakawa H., et al. 2009. Ammonia oxidation kinetics determine niche separation of nitrifying Archaea and Bacteria. Nature 461: 976-981.), with AOA dominating in ammonia-limited acid, whereas AOB have a tolerance of high ammonia concentrations.

Though no information on benthic AOA and AOB communities along the Minho and Lima estuaries is available, NH₄⁺ concentration in the upper Minho estuary (maximum 4.4 μmol L^–1) seems more suitable for the occurrence of nitrification than in the Lima estuary (maximum 1.8 μmol L^–1). Nitrification reactions typically happen within a DO range of 15.6-78.0 µmol L^–1, and the Minho and Lima surface waters were well above these concentrations, leading to nitrification conditions. However, only in the upper Minho do correlations found between ΔN₂O and AOU, NH₄⁺ and NO₂^– suggest that nitrification may have been acting as an NH₄⁺ sink and a source of N₂O. The calculated biological N₂O yield (0.060 nmol per μmol O₂ consumed) falls within the range observed in marine systems and in particular in the Atlantic off the Iberian coast (Nevison et al. 2003Nevison C.D., Butler J.H., Elkins J.W. 2003. Global distribution of N₂O and the ΔN₂O-AOU yield in the subsurface ocean. Global Biogeochem Cycles 17: 1119. ). The effect of salinity on nitrification is well documented and in many estuarine systems nitrification rates are highest at lower and intermediate salinities (Bianchi et al. 1999Bianchi M., Feliatra F., Lefevre D. 1999. Regulation of nitrification in the land-ocean contact area of the Rhône River plume (NW Mediterranean). Aquat. Microb. Ecol. 18: 301-312., Teixeira et al. 2013Teixeira C., Magalhães C., Joye S.B., et al. 2013. The role of salinity in shaping dissolved inorganic nitrogen and N₂O dynamics in estuarine sediment-water interface. Mar. Pollut. Bull. 66: 225-229.). Our results from the upper Minho are in accordance with these findings, as the potential nitrification occurred at low salinity (between ~2 and ~10). The community composition of nitrifying microbes is very dependent on salinity, but a combination of other environmental factors may shape AOB diversity along an estuary (Mosier et al. 2008Mosier A.C., Francis C.A. 2008. Relative abundance and diversity of ammonia-oxidizing archaea and bacteria in the San Francisco Bay estuary. Environ. Microbiol. 10: 3002-3016.).

pH may regulate nitrification, and Wild et al. (1971)Wild H.E., Sawyer C.N., McMahon T.C. 1971. Factors affecting nitrification kinetics. J. Water Pollut. Control Fed. 43: 1845-1854. found an ideal pH range for nitrification between 7.5 and 8.5. As nitrifiers are known to decrease pH, the sharp negative correlation found between pH and ΔN₂O in the pH range 7.2-7.4 in the upper Minho estuary (Fig. 3E) may be a direct result of nitrification. N₂O saturation values ranging from 101% to 166% in the Minho estuary and 113% to 227% in the Lima estuary indicate that both estuaries behave as a potential N₂O source to the atmosphere. Positive N₂O water-air fluxes prevailed in all sampling stations, decreasing in general from upper to lower estuaries. However, this tendency was more pronounced in the Lima estuary, where higher N₂O fluxes in the river-influenced area were about twice (~12.0 μmol m^–2 d^–1; C96) those observed in the upper part of the Minho estuary (~6.0 μmol m^–2 d^–1; C96). It is likely that a greater turbulence of Minho upper estuary waters leads to a more rapid degassing of N₂O to the atmosphere in this part of the system.

Estimated N₂O fluxes were similar to those reported from the Portuguese Tagus and Sado estuaries (Gonçalves et al. 2010Gonçalves C., Brogueira M.J., Camões M.F. 2010. Seasonal and tidal influence on the variability of nitrous oxide in the Tagus estuary, Portugal. Sci. Mar. 74S1: 57-66. , 2015Gonçalves C., Brogueira M.J, Nogueira M. 2015. Tidal and spatial variability of nitrous oxide (N₂O) in Sado estuary (Portugal). Est. Coast. Shelf Sci. 167: 466-474. ) (Table 4) but much lower than that of the Douro estuary (Barnes and Upstill-Goddard 2011Barnes J., Upstill-Goddard R.C. 2011. N₂O seasonal distributions and air-sea exchange in UK estuaries: Implications for the tropospheric N₂O source from European coastal waters. J. Geophys. Res. B. 116: G01006.). These fluxes were also much lower than those reported from the Scheldt estuary in Belgium, the Ems estuary in Germany, the Humber estuary in the UK, the Guadalete estuary in Spain and the Pearl River estuary in China (Table 4).

Though N₂O water-air fluxes were obtained during a single sampling, aware of the seasonal variability that characterizes these estuarine systems, we estimated the annual contribution of Minho and Lima estuaries to the global N₂O emissions. Taking into account estuarine areas (23 km² for Minho and 5.4 km² for Lima) and the estimated mean N₂O fluxes, we extrapolated an emission of 0.94-0.96 Mg N₂O-N yr^–1 (estimated using RC01 and C96, respectively) for the Minho estuary and 0.26-0.28 Mg N₂O-N yr^–1 (RC01 and C96, respectively) for the Lima estuary.

On a global perspective, estimated N₂O emissions from the Minho and Lima estuaries (<1.0 Mg N₂O-N yr^–1) represent a reduced fraction (<0.02%) of emissions from European estuaries (6.8 Gg N₂O yr^–1, Barnes and Upstill-Goddard 2011Barnes J., Upstill-Goddard R.C. 2011. N₂O seasonal distributions and air-sea exchange in UK estuaries: Implications for the tropospheric N₂O source from European coastal waters. J. Geophys. Res. B. 116: G01006.). Nevertheless, being aware of our unique seasonal sampling, and particularly the higher values of N₂O emissions measured in winter spring and in other Portuguese estuaries (Gonçalves et al. 2010Gonçalves C., Brogueira M.J., Camões M.F. 2010. Seasonal and tidal influence on the variability of nitrous oxide in the Tagus estuary, Portugal. Sci. Mar. 74S1: 57-66. ), more studies assessing the seasonal variability of N₂O emissions in our systems are needed.

ACKNOWLEDGEMENTSTop

Acknowledgements are due to colleagues from the IPMA Oceanography Laboratory for their assistance in sampling, technical and analytical procedures. The authors also want to thank the Instituto Hidrográfico for their assistance during sampling. The research was supported by the PoPesca MARE project (22-05-01-FDR-001) and by the FCT-Portuguese Foundation of Science and Technology (POCI 2010 and FSE) through grant SFRH/BD/28569/2006.

ReferencesTop

Aminot A., Chaussepied M. 1983. Manuel des analyses chimiques en milieu marin. Centre National pour l’Exploitation des Océans. CNEXO, Brest, France, 395 pp.

Bange H.W., Freing A., Kock A., et al. 2010. Marine pathways to nitrous oxide. In: Smith K.A. (ed.), Nitrous oxide and climate change. Earthscan, London. U.K. pp. 36-62.

Barnes J., Upstill-Goddard R.C. 2011. N₂O seasonal distributions and air-sea exchange in UK estuaries: Implications for the tropospheric N₂O source from European coastal waters. J. Geophys. Res. B. 116: G01006.
https://doi.org/10.1029/2009JG001156

Bettencourt A., Ramos L., Gomes V., et al. 2003. Estuários Portugueses. Editions INAG - Ministério das Cidades, Ordenamento do Território e Ambiente, Lisboa, 311 pp.

Bianchi M., Feliatra F., Lefevre D. 1999. Regulation of nitrification in the land-ocean contact area of the Rhône River plume (NW Mediterranean). Aquat. Microb. Ecol. 18: 301-312.
https://doi.org/10.3354/ame018301

Bollmann A., Laanbroek H.J. 2002. Influence of oxygen partial pressure and salinity on the community composition of ammonia-oxidizing bacteria in the Schelde estuary. Aquat. Microb. Ecol. 28: 239-247.
https://doi.org/10.3354/ame028239

Burgos M., Sierra A., Ortega T., et al. 2015. Anthropogenic effects on greenhouse gas (CH₄ and N₂O) emissions in the Guadalete River Estuary (SW Spain). Sci. Total Environ. 503-504: 179-189.
https://doi.org/10.1016/j.scitotenv.2014.06.038

Carini S., Weston N., Hopkinson C., et al. 1996. Gas exchange rates in the Parker River estuary, Massachusetts. Biol. Bull. 191: 333-334.
https://doi.org/10.1086/BBLv191n2p333

Clark J.F., Schlosser P., Simpson K.J, et al. 1995. Relationship between gas transfer velocities and wind speeds in the tidal Hudson River determined by the dual tracer technique. In: Jhane B., Monahan E. (eds), Air-Water Gas Transfer. AEON Verlag and Studio, Germany, pp. 785-800.

Dai M., Wang L., Guo X., et al. 2008. Nitrification and inorganic nitrogen distribution in a large perturbed river/estuarine system: the Pearl River Estuary, China. Biogeosci. 5: 1227-1244.
https://doi.org/10.5194/bg-5-1227-2008

de Bie M.J.M., Middelburg J.J. Starink M., et al. 2002. Factors controlling nitrous oxide at the microbial community and estuarine scale. Mar. Ecol. Prog. Ser. 240: 1-9.
https://doi.org/10.3354/meps240001

de Wilde H.P.J., de Bie M.J.M. 2000. Nitrous oxide in the Schelde estuary: Production by nitrification and emission to the atmosphere. Mar. Chem. 69: 203-216.
https://doi.org/10.1016/S0304-4203(99)00106-1

Dong L.F., Nedwell D.B. 2006. Sources of nitrogen used for denitrification and nitrous oxide formation in sediments of the hypernutrified Colne, the nutrified Humber and oligotrophic Conwy Estuaries, United Kingdom. Limnol. Oceanogr. 51(1, part 2): 545-557.
https://doi.org/10.4319/lo.2006.51.1_part_2.0545

Ferreira J.G., Nobre A.M., Simas T.C., et al. 2005. Monitoring plan for water quality and ecology for Portuguese traditional and coastal waters: Development of guidelines for the application of the European Union Water Framework Directive. Editions INAG-Instituto da Água IMAR-Institute of Marine Research. Lisbon, Portugal, 142 pp.

Gonçalves C., Brogueira M.J., Camões M.F. 2010. Seasonal and tidal influence on the variability of nitrous oxide in the Tagus estuary, Portugal. Sci. Mar. 74S1: 57-66.
https://doi.org/10.3989/scimar.2010.74s1057

Gonçalves C., Brogueira M.J, Nogueira M. 2015. Tidal and spatial variability of nitrous oxide (N₂O) in Sado estuary (Portugal). Est. Coast. Shelf Sci. 167: 466-474.
https://doi.org/10.1016/j.ecss.2015.10.028

Hartman B., Hammond D. 1985. Gas exchange in San Francisco Bay. Hydrobiology 129: 59-68.
https://doi.org/10.1007/BF00048687

Lin H., Dai M., Kao S.J. et al. 2016. Spatio temporal variability of nitrous oxide in a large eutrophic estuarine system: The Pearl River Estuary, China. Mar. Chem. 182: 14-24.
https://doi.org/10.1016/j.marchem.2016.03.005

Martens-Habbena W., Berube P.M., Urakawa H., et al. 2009. Ammonia oxidation kinetics determine niche separation of nitrifying Archaea and Bacteria. Nature 461: 976-981.
https://doi.org/10.1038/nature08465

Mosier A.C., Francis C.A. 2008. Relative abundance and diversity of ammonia-oxidizing archaea and bacteria in the San Francisco Bay estuary. Environ. Microbiol. 10: 3002-3016.
https://doi.org/10.1111/j.1462-2920.2008.01764.x

Murray R.H., Erler D.V., Eyre B.D. 2015. Nitrous oxide fluxes in estuarine environments: response to global change. Global Change Biol. 21: 3219-3245.
https://doi.org/10.1111/gcb.12923

Myhre G., Shindell D., Bréon F.M., et al. 2013. Anthropogenic and Natural Radiative Forcing. In: Stocker T.F., Qin D., Plattner G.K., et al. (eds), Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, pp. 659-740.
https://doi.org/10.1017/CBO9781107415324.018

Nevison C.D., Butler J.H., Elkins J.W. 2003. Global distribution of N₂O and the ΔN₂O-AOU yield in the subsurface ocean. Global Biogeochem Cycles 17: 1119.
https://doi.org/10.1029/2003GB002068

Prather M.J., Holmes C.D., Hsu J. 2012. Reactive greenhouse gas scenarios: Systematic exploration of uncertainties and the role of atmospheric chemistry. Geophys. Res. Lett. 39: L09803.
https://doi.org/10.1029/2012GL051440

Ramos S., Cowen R.K., Ré P., et al. 2006. Temporal and spatial distributions of larval fish assemblages in the Lima estuary (Portugal). Est. Coast. Shelf Sci. 66: 303-314.
https://doi.org/10.1016/j.ecss.2005.09.012

Raymond P.A., Cole J.J. 2001. Gas exchange in rivers and estuaries: choosing a gas transfer velocity. Estuaries 24: 312-317.
https://doi.org/10.2307/1352954

SNIRH (Sistema Nacional de Informação de Recursos Hídricos). 2013. Instituto da Água IP.
http://snirh.pt

Sousa C., Vaz M.N., Alvarez I., et al. 2013. Effect of Minho estuarine plume on Rias Baixas: numerical modeling approach. J. Coast. Res. Spec. 65: 2059-2065.
https://doi.org/10.2112/SI65-348.1

Strauss E.A., Mitchell N.L., Lamberti G.A. 2002. Factors regulating nitrification in aquatic sediments: effects of organic carbon, nitrogen availability, and pH. Can. J. Fish. Aquat. Sci. 59: 554-563.
https://doi.org/10.1139/f02-032

Teixeira C., Magalhães C., Joye S.B., et al. 2013. The role of salinity in shaping dissolved inorganic nitrogen and N₂O dynamics in estuarine sediment-water interface. Mar. Pollut. Bull. 66: 225-229.
https://doi.org/10.1016/j.marpolbul.2012.11.004

Wanninkhof R. 1992. Relationship between wind speed and gas exchange over the ocean. J. Geophys. Res. C 97(C5): 7373-7382.
https://doi.org/10.1029/92JC00188

Weiss R.F. 1970. The solubility of nitrogen, oxygen, and argon in water and seawater. Deep-Sea Res. Oceanogr. Abstr. 17: 721-735.
https://doi.org/10.1016/0011-7471(70)90037-9

Weiss R.F., Price B.A. 1980. Nitrous oxide in water and seawater. Mar. Chem. 8: 347-359.
https://doi.org/10.1016/0304-4203(80)90024-9

Wild H.E., Sawyer C.N., McMahon T.C. 1971. Factors affecting nitrification kinetics. J. Water Pollut. Control Fed. 43: 1845-1854.

WMO (World Meteorological Organization). 2006. World Meteorological Organization greenhouse gas bulletin: The State of Greenhouse Gases in the Atmosphere Using Global Observations through 2006. Bulletin No. 3. Global Atmosphere Watch, Geneva, Switzerland, 4 pp.

Yoshinari T. 1976. Nitrous oxide in the sea. Mar. Chem. 2: 189-202.