Photochemical & Photobiological Sciences View Article Online
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
PAPER
Cite this: Photochem. Photobiol. Sci., 2014, 13, 799
View Journal | View Issue
Interaction of extrinsic chemical factors affecting photodegradation of dissolved organic matter in aquatic ecosystems† Petr Porcal,*a,c Peter J. Dillona and Lewis A. Molotb Photochemical degradation of dissolved organic matter (DOM) plays an important role in the carbon cycle. Irradiation experiments were conducted to evaluate the influence of chemical factors, specifically those expected to be altered in natural waters by atmospheric acid deposition, on photodegradation of DOM. These included pH, nitrate, iron and calcium. The experiments were carried out using stream and lake water samples with a wide range of dissolved organic carbon (DOC) concentration. Decreasing DOC concentration along with decreasing absorbance was observed during three-week exposures to natural solar radiation as well as during laboratory experiments with artificial solar radiation. The pH of the samples significantly affected degradation rates of DOM especially with elevated iron, while no influence
Received 10th January 2014, Accepted 17th March 2014 DOI: 10.1039/c4pp00011k www.rsc.org/pps
of nitrate or calcium concentration was observed. Addition of FeIII did not significantly affect photodegradation and photobleaching rate constants in samples at circumneutral pH. Acid pH increased photodegradation rates. The results suggest that photodegradation rates of DOM will decrease during recovery from acidification. Hence, lower photodegradation rates may be responsible for increases in DOM observed in some regions of North America and Europe.
Introduction Photochemical processes play an important role in carbon cycling in aquatic ecosystems. Dissolved organic matter (DOM) is composed of a complex mixture of organic material, originating both within the aquatic environment (autochthonous) as well as through the transport of partially-degraded organic material from the surrounding terrestrial environment (allochthonous). Photochemical processes involve direct mineralization or oxidation of DOM to carbon monoxide or carbon dioxide. In most studies involving terrestrial DOM, photobleaching has been shown to result in degradation of complex molecular structure of DOM towards lower molecular weight substances which are more susceptible to microbial degradation.1 Photochemical transformation of DOM is a complex process consisting of several interconnected pathways and
a Environment and Resource Studies, Trent University, 1600 West Bank Drive, Peterborough, Ontario, Canada, K9J 7B8. E-mail: [email protected] b Faculty of Environmental Studies, York University, 4700 Keele Street, Toronto, Ontario, Canada, M3J 1P3 c Biology Centre of the AS CR, v.v.i., Institute of Hydrobiology, České Budějovice, Czech Republic † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c4pp00011k
This journal is © The Royal Society of Chemistry and Owner Societies 2014
intermediates and due to its complexity could be easily affected by several chemical factors extrinsic to DOM. It has been shown that pH affects photodegradation of DOM.2,3 Molot et al.4 examined the effect of pH on photochemical loss of DOM by hydroxyl radicals (•OH), which are the most important intermediates in DOM photodegradation. Hydroxyl radicals are significantly produced in the presence of nitrite and nitrate5 and by the reaction of FeII with H2O2 (the photoFenton process).6,7 In the presence of iron, photolysis of FeIIIDOM complexes can also be a significant photodegradation pathway.8 The effects of other ions on photodegradation have not been studied in detail. Reche et al.9 found a positive correlation between photodegradation rates and the sum of calcium and magnesium cations. Significant increases in dissolved organic carbon (DOC) concentrations have been reported in some regions in North America and much of Europe. Several possible explanations have been suggested for the observed increases including increasing CO2, climate warming, decreased SO4 deposition, continued nitrogen deposition and hydrological changes including altered precipitation trends.10–15 Significant reductions in acid deposition observed in Europe and North America15 have resulted in recovery from acidification. Decrease in SO4 deposition results in decreasing SO4 concentration, increasing pH and acid neutralizing capacity in streams and lakes in Europe16 and some regions in
Photochem. Photobiol. Sci., 2014, 13, 799–812 | 799
View Article Online
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
Paper
North America.17 Will the increase in pH and acid neutralizing capacity affect photochemical processes of DOM in acidified waters? The trend in nitrate concentration is ambiguous. Evans et al.16 in a large spatial study of water quality data for 56 long-term monitoring sites in eight European countries observed increasing trends in Italy and UK, decreasing in Scandinavia and Central Europe, while there was no significant trend in 35 of 56 sites. Will this increasing trend in NO3− concentration have any effect on photochemical processes of DOM? Decreases in acid anion concentrations (SO4− and NO3−) should be balanced by smaller decreases (relative to acid anions) in base cations and increases in alkalinity.17 Because large quantities of exchangeable calcium and magnesium were lost during acidification, recovery will be a slow process.18 One possible way to speed up the recovery or suppress the effects of acidification is liming.19 Depending on intensity and duration the liming increases Ca concentration and pH in surface waters.20 Could these changes affect photochemical processes of DOM? This study examined how changes in pH, Fe, NO3− and Ca concentrations affect photochemical degradation of DOM in temperate regions impacted by acidification. The results of this study could have potential applicability also to the downstream ecosystems such as estuaries and coastal environments which exhibit the wide range of observed extrinsic properties.
Material and methods Study sites Stream and lake water samples used in this study were collected from Harp, Dickie, and Brandy Lakes and their inflows (Plastic Inflow #1 – PC1, Dickie Inflow #5 – DE5, Dickie Inflow #8 – DE8, Harp Inflow #4 – HP4, Brandy Inflow #2 – KS2) located in the districts of Muskoka and Haliburton, Ontario, Canada, near the southern edge of Precambrian Shield and the boundary of the Boreal ecozone. These locations are longterm study sites of the impacts of long range atmospheric transport, climate change and recreational development on water quality in forested headwater catchments and lakes e.g. ref. 21. The dominant soil types are acidic brunisols and podsol. Organics soils ( peat) are common throughout the region in areas with very poor drainage which are typically saturated with water for a large part of the year. All of the catchments are primarily forested, well-drained soils generally have deciduous or mixed forests, whereas the poorly drained soils have mixed or coniferous forest.21 The samples collected from Jackson and Mariposa Creek (Peterborough County, Ontario) represented rural, agriculture regions. The catchment areas of Jackson and Mariposa Creeks were 100 times larger than the catchments of PC1, DE5, HP4, and KS2. Pastures and cropland covered 61 and 81% of Jackson and Mariposa Creek catchments, respectively. Samples were collected from surface layer into 20 L HDPE carboys, previously thoroughly acid-washed and rinsed with de-mineralized water. Samples were immediately stored in the
800 | Photochem. Photobiol. Sci., 2014, 13, 799–812
Photochemical & Photobiological Sciences
dark and transported to a refrigerator. Within a week samples were filtered through a series of cartridge filters with decreasing nominal pore size from 25 µm to 0.5 µm (US Filter). Irradiation experiments Filtered samples were divided into at least three parts each. One part remained unaltered. The pH, NO3−, Ca, or Fe concentrations were amended in the remaining parts. The pH was increased or decreased with 0.1 M NaOH and 0.1 M HCl (Fischer Scientific). The NO3−, Ca and Fe concentrations were amended with NaNO3, CaCl2 and FeCl3 (Fischer Scientific), respectively. The pH of the samples with altered NO3−, Ca and Fe concentrations was adjusted to the pH value of the unaltered sample. Prepared samples were exposed to artificial solar radiation in laboratory experiments or to natural solar radiation in rooftop experiments. In a series of laboratory irradiation experiments with PC1 stream water (collected in June 2009) determining the simultaneous effect of pH and Fe on photodegradation, 0.5 mM NaF (reagent grade) was added to half of the samples to reduce the effect of free Fe ions.4 At pH 4.0, F− is almost as effective as EDTA at chelating Fe (however, it should be noted that because F− is a monodentate ligand it forms less stable complexes with Fe than multidentate ligands).22 Laboratory experiments Samples were exposed in 250 mL UV-transparent Teflon® (FEP) bottles2 to artificial solar radiation for different time periods up to 48 h in duplicate or triplicate. Subsamples were collected after 12, 24, and 36 hours. Samples were irradiated inside a Suntest XLS+ Solar Simulator (Atlas Gmbh., Germany) equipped with a Xenon lamp and quartz filters to simulate the noon time solar spectrum at sea level and continuously irradiated. Samples were partly submerged in a water bath to eliminate excess heat and maintain a relatively constant sample temperature of approximately 25 °C. The depth of the water bath was 5 cm and the water depth in the bottles was 11 cm. Dark samples, wrapped in a tin foil, were kept near the air exhaust at the same temperature as inside the simulator for the duration of the experiment. The leaching of DOM from Teflon bottles was previously tested with Milli-Q water and was negligible. The intensity of irradiation in the UV and visible range used in laboratory experiments was 700 W m−2. It is approximately two times higher than monthly averaged midday insolation on horizontal surface in June and approximately 30% higher than maximum outdoors irradiance under a cloudless, non-hazy sky on June 21 at our sampling location (NASA Surface Meteorology and Solar Energy, http://eosweb.larc.nasa. gov). The higher intensity was used to reduce irradiation time to accelerate the experimental work. Preliminary experiments with samples from Dickie and Harp Lake catchments (DOM ranged from 0.4 to 3.3 mmol L−1), when the samples were irradiated under two different intensities (400 and 700 W m−2) for the same amount of irradiation energy, showed that photodegradation rate constant does not depend on inten-
This journal is © The Royal Society of Chemistry and Owner Societies 2014
View Article Online
Photochemical & Photobiological Sciences
sity but on total amount of irradiation energy ( paired t-test, p > 0.05, n = 4).23
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
Roof experiments – solar radiation Natural solar radiation experiments were conducted in a shallow pool on the roof of a university building (N 44.36°, W 78.28°). The site was not shaded by any construction. The pool was filled and continuously cooled with tap water. The temperature in the pool did not exceed 25 °C. Samples were exposed in polyethylene bags, which were transparent to photosynthetically active radiation (at 400 nm, transmittance 94%), UV-A (315 nm, 90%), and UV-B (280 nm, 85%). The bags were filled with 250 mL of sample and floated in the cooling water during the exposure. The samples were exposed and measured in triplicate. The blank samples were wrapped in a tin foil and kept submerged in the pool during the experiment. The intensity of solar radiation was not measured directly; the average daily insolation was instead calculated from monthly averaged data by linear interpolation and is therefore only an approximation. Monthly averaged insolation data for the university location (22-year average, location N45°30′, W78°30′) was obtained from the NASA Atmospheric Science Data Center, Surface Meteorology and Solar Energy (http:// eosweb.larc.nasa.gov/sse/). Global horizontal radiation is measured over a wide spectrum range including UV, visible and near-infrared radiation. However, the radiation inside of the irradiation chamber was only measured within UV and visible spectrum range. The portion of near-infrared radiation in global horizontal radiation ranges from 46 to 52% (e.g. ref. 24 and 25 and references therein) depending on atmospheric conditions, e.g. amount of water vapor. In our calculations the global horizontal radiation data were reduced of 50% to eliminate the portion of near infrared radiation. The long term average irradiation may have differed from actual irradiation during the experimental periods. The goodness of fit between results obtained in laboratory and roof experiments was tested, coefficient of determination (R2) ranged from 0.67 to 0.97 (n = 3). The cumulative irradiation energy used in laboratory experiments during 48 hours was 121 MJ m−2 and corresponded approximately to 11 days of average solar insolation in June at sampling locations. The length of roof experiments ranged from 14 to 24 days. The samples were collected six times during the duration of experiment. The possible leaching of DOC from bags was tested with Milli-Q water exposed in bags for the duration of experiments. The polyethylene (PE) bags exposed to solar radiation leached 0.8 ± 0.03 mg L−1 of DOC (mean ± standard error, n = 6) during the three weeks experiment, while no significant leaching of DOC was measured in wrapped control samples. All results were corrected for DOC leaching. The amount of leaching DOC was linearly extrapolated for individual time period and distracted from measured DOC concentration. Analyses After irradiation the samples were filtered through 0.45 µm nylon syringe filters (Mandel, Canada) to remove particles
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Paper
formed during irradiation.26 Filters were previously rinsed with 20 mL of de-mineralized water and 10 mL of sample. The concentration of DOM in filtered samples expressed as dissolved organic carbon was determined with a TOC-VCP analyzer (Shimadzu, Japan) by catalytic combustion oxidation with nondispersive infrared detection. Inorganic and organic standards were analyzed at the beginning and at the end of each run to correct possible instrumental drift. The concentration of total organic carbon was determined in unfiltered samples. The concentration of particulate organic carbon was calculated as a difference between total and dissolved concentrations of organic carbon. The concentration of NO3− was determined by ion chromatography (Dionex DX-500) or by flow injection method (EPA 220.3) with quantitative reduction to nitrite followed by nitrite determination (by diazotizing with sulfanilamide followed by coupling with N-(1-naphthyl) ethylenediamine dihydrochloride). Nitrite alone was also determined and the nitrate concentration was obtained by subtraction. The concentration of Ca was determined by ion chromatography (Dionex DX-500). Conductivity, pH and alkalinity (Gran titration method) were measured by an automatic titrator (PCtitrate, Man-Tech, Canada). The concentration of Fe was determined by inductively coupled plasma mass spectroscopy (ICP-MS) on an ICP mass-spectrometer Thermo Finnigan ELEMENT2 (Thermo Electron Corporation). Samples for Fe determination were preserved by nitric acid (final pH < 2). The concentration of particulate Fe was calculated as a difference between Fe concentrations in unfiltered and filtered samples.
Absorbance Absorbance spectrum was measured by a Cary 50 UV-Vis spectrophotometer (Varian Inc.) in a 1 cm quartz cuvette from 200 to 800 nm. The fraction of DOM that absorbs ultraviolet (UV) and visible light is usually referred to as chromophoric or colored dissolved organic matter (CDOM). An index of the CDOM content was calculated as the integrated absorption 4P 50 coefficient from 250 to 450 nm (1 nm resolution) ( ABS).27 250
Spectral slopes for the intervals of 350–400 nm (S350–400) and 275–295 nm (S275–295) were determined by fitting the absorption spectra within corresponding ranges to a single exponential decay function according to Helms et al.27
Calculations The difference between initial concentration and final concentration in dark samples represents microbial degradation and chemical changes and ranged up to 4.5% of the initial concentration. The dark changes were linearly extrapolated for individual time period. The net photochemical degradation was thus calculated as the difference between the concentration at the beginning of experiment and the measured concentration plus dark change at any time.
Photochem. Photobiol. Sci., 2014, 13, 799–812 | 801
View Article Online
Paper
Photochemical & Photobiological Sciences
Degradation rate constants The equation for a pseudo first-order kinetic reaction was used to calculate the photochemical degradation rate constant:
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
DOC ¼ DOC0 eK DOC E ;
ð1Þ
where KDOC is the pseudo first-order photodegradation rate constant, DOC0 is the initial concentration of DOC [mg L−1], DOC is the DOC concentration [mg L−1] after exposure, and E is the cumulative energy of irradiation incident on horizontal surface [GJ m−2] calculated as a product of intensity of irradiation and time. Photobleaching was defined as a decrease in CDOM content. The equation for a pseudo first-order kinetic reaction was used to calculate photobleaching rate constant: 450 X
ABS ¼
450 X
250
ABS0 eK CDOM E ;
ð2Þ
250
where KCDOM is the pseudo first-order photobleaching rate 450 450 P P constant, ABS0 is the initial CDOM content [m−1], ABS is 250
250
the CDOM content [m−1] after exposure, and E is the cumulative energy of irradiation incident on horizontal surface [GJ m−2] calculated as a product of intensity of irradiation and time. A non-linear regression analysis was used to fit eqn (1) and (2) to the measured data.28
Statistics A repeated measures one-way ANOVA was used to test the effects of each manipulation within each experiment. Post-hoc
Tukey’s multiple comparison tests were performed when the p was
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
PAPER
Cite this: Photochem. Photobiol. Sci., 2014, 13, 799
View Journal | View Issue
Interaction of extrinsic chemical factors affecting photodegradation of dissolved organic matter in aquatic ecosystems† Petr Porcal,*a,c Peter J. Dillona and Lewis A. Molotb Photochemical degradation of dissolved organic matter (DOM) plays an important role in the carbon cycle. Irradiation experiments were conducted to evaluate the influence of chemical factors, specifically those expected to be altered in natural waters by atmospheric acid deposition, on photodegradation of DOM. These included pH, nitrate, iron and calcium. The experiments were carried out using stream and lake water samples with a wide range of dissolved organic carbon (DOC) concentration. Decreasing DOC concentration along with decreasing absorbance was observed during three-week exposures to natural solar radiation as well as during laboratory experiments with artificial solar radiation. The pH of the samples significantly affected degradation rates of DOM especially with elevated iron, while no influence
Received 10th January 2014, Accepted 17th March 2014 DOI: 10.1039/c4pp00011k www.rsc.org/pps
of nitrate or calcium concentration was observed. Addition of FeIII did not significantly affect photodegradation and photobleaching rate constants in samples at circumneutral pH. Acid pH increased photodegradation rates. The results suggest that photodegradation rates of DOM will decrease during recovery from acidification. Hence, lower photodegradation rates may be responsible for increases in DOM observed in some regions of North America and Europe.
Introduction Photochemical processes play an important role in carbon cycling in aquatic ecosystems. Dissolved organic matter (DOM) is composed of a complex mixture of organic material, originating both within the aquatic environment (autochthonous) as well as through the transport of partially-degraded organic material from the surrounding terrestrial environment (allochthonous). Photochemical processes involve direct mineralization or oxidation of DOM to carbon monoxide or carbon dioxide. In most studies involving terrestrial DOM, photobleaching has been shown to result in degradation of complex molecular structure of DOM towards lower molecular weight substances which are more susceptible to microbial degradation.1 Photochemical transformation of DOM is a complex process consisting of several interconnected pathways and
a Environment and Resource Studies, Trent University, 1600 West Bank Drive, Peterborough, Ontario, Canada, K9J 7B8. E-mail: [email protected] b Faculty of Environmental Studies, York University, 4700 Keele Street, Toronto, Ontario, Canada, M3J 1P3 c Biology Centre of the AS CR, v.v.i., Institute of Hydrobiology, České Budějovice, Czech Republic † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c4pp00011k
This journal is © The Royal Society of Chemistry and Owner Societies 2014
intermediates and due to its complexity could be easily affected by several chemical factors extrinsic to DOM. It has been shown that pH affects photodegradation of DOM.2,3 Molot et al.4 examined the effect of pH on photochemical loss of DOM by hydroxyl radicals (•OH), which are the most important intermediates in DOM photodegradation. Hydroxyl radicals are significantly produced in the presence of nitrite and nitrate5 and by the reaction of FeII with H2O2 (the photoFenton process).6,7 In the presence of iron, photolysis of FeIIIDOM complexes can also be a significant photodegradation pathway.8 The effects of other ions on photodegradation have not been studied in detail. Reche et al.9 found a positive correlation between photodegradation rates and the sum of calcium and magnesium cations. Significant increases in dissolved organic carbon (DOC) concentrations have been reported in some regions in North America and much of Europe. Several possible explanations have been suggested for the observed increases including increasing CO2, climate warming, decreased SO4 deposition, continued nitrogen deposition and hydrological changes including altered precipitation trends.10–15 Significant reductions in acid deposition observed in Europe and North America15 have resulted in recovery from acidification. Decrease in SO4 deposition results in decreasing SO4 concentration, increasing pH and acid neutralizing capacity in streams and lakes in Europe16 and some regions in
Photochem. Photobiol. Sci., 2014, 13, 799–812 | 799
View Article Online
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
Paper
North America.17 Will the increase in pH and acid neutralizing capacity affect photochemical processes of DOM in acidified waters? The trend in nitrate concentration is ambiguous. Evans et al.16 in a large spatial study of water quality data for 56 long-term monitoring sites in eight European countries observed increasing trends in Italy and UK, decreasing in Scandinavia and Central Europe, while there was no significant trend in 35 of 56 sites. Will this increasing trend in NO3− concentration have any effect on photochemical processes of DOM? Decreases in acid anion concentrations (SO4− and NO3−) should be balanced by smaller decreases (relative to acid anions) in base cations and increases in alkalinity.17 Because large quantities of exchangeable calcium and magnesium were lost during acidification, recovery will be a slow process.18 One possible way to speed up the recovery or suppress the effects of acidification is liming.19 Depending on intensity and duration the liming increases Ca concentration and pH in surface waters.20 Could these changes affect photochemical processes of DOM? This study examined how changes in pH, Fe, NO3− and Ca concentrations affect photochemical degradation of DOM in temperate regions impacted by acidification. The results of this study could have potential applicability also to the downstream ecosystems such as estuaries and coastal environments which exhibit the wide range of observed extrinsic properties.
Material and methods Study sites Stream and lake water samples used in this study were collected from Harp, Dickie, and Brandy Lakes and their inflows (Plastic Inflow #1 – PC1, Dickie Inflow #5 – DE5, Dickie Inflow #8 – DE8, Harp Inflow #4 – HP4, Brandy Inflow #2 – KS2) located in the districts of Muskoka and Haliburton, Ontario, Canada, near the southern edge of Precambrian Shield and the boundary of the Boreal ecozone. These locations are longterm study sites of the impacts of long range atmospheric transport, climate change and recreational development on water quality in forested headwater catchments and lakes e.g. ref. 21. The dominant soil types are acidic brunisols and podsol. Organics soils ( peat) are common throughout the region in areas with very poor drainage which are typically saturated with water for a large part of the year. All of the catchments are primarily forested, well-drained soils generally have deciduous or mixed forests, whereas the poorly drained soils have mixed or coniferous forest.21 The samples collected from Jackson and Mariposa Creek (Peterborough County, Ontario) represented rural, agriculture regions. The catchment areas of Jackson and Mariposa Creeks were 100 times larger than the catchments of PC1, DE5, HP4, and KS2. Pastures and cropland covered 61 and 81% of Jackson and Mariposa Creek catchments, respectively. Samples were collected from surface layer into 20 L HDPE carboys, previously thoroughly acid-washed and rinsed with de-mineralized water. Samples were immediately stored in the
800 | Photochem. Photobiol. Sci., 2014, 13, 799–812
Photochemical & Photobiological Sciences
dark and transported to a refrigerator. Within a week samples were filtered through a series of cartridge filters with decreasing nominal pore size from 25 µm to 0.5 µm (US Filter). Irradiation experiments Filtered samples were divided into at least three parts each. One part remained unaltered. The pH, NO3−, Ca, or Fe concentrations were amended in the remaining parts. The pH was increased or decreased with 0.1 M NaOH and 0.1 M HCl (Fischer Scientific). The NO3−, Ca and Fe concentrations were amended with NaNO3, CaCl2 and FeCl3 (Fischer Scientific), respectively. The pH of the samples with altered NO3−, Ca and Fe concentrations was adjusted to the pH value of the unaltered sample. Prepared samples were exposed to artificial solar radiation in laboratory experiments or to natural solar radiation in rooftop experiments. In a series of laboratory irradiation experiments with PC1 stream water (collected in June 2009) determining the simultaneous effect of pH and Fe on photodegradation, 0.5 mM NaF (reagent grade) was added to half of the samples to reduce the effect of free Fe ions.4 At pH 4.0, F− is almost as effective as EDTA at chelating Fe (however, it should be noted that because F− is a monodentate ligand it forms less stable complexes with Fe than multidentate ligands).22 Laboratory experiments Samples were exposed in 250 mL UV-transparent Teflon® (FEP) bottles2 to artificial solar radiation for different time periods up to 48 h in duplicate or triplicate. Subsamples were collected after 12, 24, and 36 hours. Samples were irradiated inside a Suntest XLS+ Solar Simulator (Atlas Gmbh., Germany) equipped with a Xenon lamp and quartz filters to simulate the noon time solar spectrum at sea level and continuously irradiated. Samples were partly submerged in a water bath to eliminate excess heat and maintain a relatively constant sample temperature of approximately 25 °C. The depth of the water bath was 5 cm and the water depth in the bottles was 11 cm. Dark samples, wrapped in a tin foil, were kept near the air exhaust at the same temperature as inside the simulator for the duration of the experiment. The leaching of DOM from Teflon bottles was previously tested with Milli-Q water and was negligible. The intensity of irradiation in the UV and visible range used in laboratory experiments was 700 W m−2. It is approximately two times higher than monthly averaged midday insolation on horizontal surface in June and approximately 30% higher than maximum outdoors irradiance under a cloudless, non-hazy sky on June 21 at our sampling location (NASA Surface Meteorology and Solar Energy, http://eosweb.larc.nasa. gov). The higher intensity was used to reduce irradiation time to accelerate the experimental work. Preliminary experiments with samples from Dickie and Harp Lake catchments (DOM ranged from 0.4 to 3.3 mmol L−1), when the samples were irradiated under two different intensities (400 and 700 W m−2) for the same amount of irradiation energy, showed that photodegradation rate constant does not depend on inten-
This journal is © The Royal Society of Chemistry and Owner Societies 2014
View Article Online
Photochemical & Photobiological Sciences
sity but on total amount of irradiation energy ( paired t-test, p > 0.05, n = 4).23
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
Roof experiments – solar radiation Natural solar radiation experiments were conducted in a shallow pool on the roof of a university building (N 44.36°, W 78.28°). The site was not shaded by any construction. The pool was filled and continuously cooled with tap water. The temperature in the pool did not exceed 25 °C. Samples were exposed in polyethylene bags, which were transparent to photosynthetically active radiation (at 400 nm, transmittance 94%), UV-A (315 nm, 90%), and UV-B (280 nm, 85%). The bags were filled with 250 mL of sample and floated in the cooling water during the exposure. The samples were exposed and measured in triplicate. The blank samples were wrapped in a tin foil and kept submerged in the pool during the experiment. The intensity of solar radiation was not measured directly; the average daily insolation was instead calculated from monthly averaged data by linear interpolation and is therefore only an approximation. Monthly averaged insolation data for the university location (22-year average, location N45°30′, W78°30′) was obtained from the NASA Atmospheric Science Data Center, Surface Meteorology and Solar Energy (http:// eosweb.larc.nasa.gov/sse/). Global horizontal radiation is measured over a wide spectrum range including UV, visible and near-infrared radiation. However, the radiation inside of the irradiation chamber was only measured within UV and visible spectrum range. The portion of near-infrared radiation in global horizontal radiation ranges from 46 to 52% (e.g. ref. 24 and 25 and references therein) depending on atmospheric conditions, e.g. amount of water vapor. In our calculations the global horizontal radiation data were reduced of 50% to eliminate the portion of near infrared radiation. The long term average irradiation may have differed from actual irradiation during the experimental periods. The goodness of fit between results obtained in laboratory and roof experiments was tested, coefficient of determination (R2) ranged from 0.67 to 0.97 (n = 3). The cumulative irradiation energy used in laboratory experiments during 48 hours was 121 MJ m−2 and corresponded approximately to 11 days of average solar insolation in June at sampling locations. The length of roof experiments ranged from 14 to 24 days. The samples were collected six times during the duration of experiment. The possible leaching of DOC from bags was tested with Milli-Q water exposed in bags for the duration of experiments. The polyethylene (PE) bags exposed to solar radiation leached 0.8 ± 0.03 mg L−1 of DOC (mean ± standard error, n = 6) during the three weeks experiment, while no significant leaching of DOC was measured in wrapped control samples. All results were corrected for DOC leaching. The amount of leaching DOC was linearly extrapolated for individual time period and distracted from measured DOC concentration. Analyses After irradiation the samples were filtered through 0.45 µm nylon syringe filters (Mandel, Canada) to remove particles
This journal is © The Royal Society of Chemistry and Owner Societies 2014
Paper
formed during irradiation.26 Filters were previously rinsed with 20 mL of de-mineralized water and 10 mL of sample. The concentration of DOM in filtered samples expressed as dissolved organic carbon was determined with a TOC-VCP analyzer (Shimadzu, Japan) by catalytic combustion oxidation with nondispersive infrared detection. Inorganic and organic standards were analyzed at the beginning and at the end of each run to correct possible instrumental drift. The concentration of total organic carbon was determined in unfiltered samples. The concentration of particulate organic carbon was calculated as a difference between total and dissolved concentrations of organic carbon. The concentration of NO3− was determined by ion chromatography (Dionex DX-500) or by flow injection method (EPA 220.3) with quantitative reduction to nitrite followed by nitrite determination (by diazotizing with sulfanilamide followed by coupling with N-(1-naphthyl) ethylenediamine dihydrochloride). Nitrite alone was also determined and the nitrate concentration was obtained by subtraction. The concentration of Ca was determined by ion chromatography (Dionex DX-500). Conductivity, pH and alkalinity (Gran titration method) were measured by an automatic titrator (PCtitrate, Man-Tech, Canada). The concentration of Fe was determined by inductively coupled plasma mass spectroscopy (ICP-MS) on an ICP mass-spectrometer Thermo Finnigan ELEMENT2 (Thermo Electron Corporation). Samples for Fe determination were preserved by nitric acid (final pH < 2). The concentration of particulate Fe was calculated as a difference between Fe concentrations in unfiltered and filtered samples.
Absorbance Absorbance spectrum was measured by a Cary 50 UV-Vis spectrophotometer (Varian Inc.) in a 1 cm quartz cuvette from 200 to 800 nm. The fraction of DOM that absorbs ultraviolet (UV) and visible light is usually referred to as chromophoric or colored dissolved organic matter (CDOM). An index of the CDOM content was calculated as the integrated absorption 4P 50 coefficient from 250 to 450 nm (1 nm resolution) ( ABS).27 250
Spectral slopes for the intervals of 350–400 nm (S350–400) and 275–295 nm (S275–295) were determined by fitting the absorption spectra within corresponding ranges to a single exponential decay function according to Helms et al.27
Calculations The difference between initial concentration and final concentration in dark samples represents microbial degradation and chemical changes and ranged up to 4.5% of the initial concentration. The dark changes were linearly extrapolated for individual time period. The net photochemical degradation was thus calculated as the difference between the concentration at the beginning of experiment and the measured concentration plus dark change at any time.
Photochem. Photobiol. Sci., 2014, 13, 799–812 | 801
View Article Online
Paper
Photochemical & Photobiological Sciences
Degradation rate constants The equation for a pseudo first-order kinetic reaction was used to calculate the photochemical degradation rate constant:
Published on 20 March 2014. Downloaded by Universitat Politècnica de València on 27/10/2014 02:42:45.
DOC ¼ DOC0 eK DOC E ;
ð1Þ
where KDOC is the pseudo first-order photodegradation rate constant, DOC0 is the initial concentration of DOC [mg L−1], DOC is the DOC concentration [mg L−1] after exposure, and E is the cumulative energy of irradiation incident on horizontal surface [GJ m−2] calculated as a product of intensity of irradiation and time. Photobleaching was defined as a decrease in CDOM content. The equation for a pseudo first-order kinetic reaction was used to calculate photobleaching rate constant: 450 X
ABS ¼
450 X
250
ABS0 eK CDOM E ;
ð2Þ
250
where KCDOM is the pseudo first-order photobleaching rate 450 450 P P constant, ABS0 is the initial CDOM content [m−1], ABS is 250
250
the CDOM content [m−1] after exposure, and E is the cumulative energy of irradiation incident on horizontal surface [GJ m−2] calculated as a product of intensity of irradiation and time. A non-linear regression analysis was used to fit eqn (1) and (2) to the measured data.28
Statistics A repeated measures one-way ANOVA was used to test the effects of each manipulation within each experiment. Post-hoc
Tukey’s multiple comparison tests were performed when the p was
