当前位置:首页 >> 农学 >>

Carbon fluxes,p CO2 and substrate weathering in a large northern river basin,


Chemical Geology 159 ?1999. 61–86

Carbon fluxes, p CO 2 and substrate weathering in a large northern river basin, Canada: carbon isotope perspectives
Kevin Telmer
a

a, ) ,1

, Jan Veizer

b,c

School of Earth and Ocean Science, Uni?ersity of Victoria, Victoria, BC, Canada, V8W 3P6 b Ottawa-Carleton Geoscience Centre, Uni?ersity of Ottawa, Ottawa, Canada, K1N 6N5 c Institut f ur ¨ Geologie, Ruhr Uni?ersitat, ¨ 44780 Bochum, Germany Received 20 May 1998; accepted 30 November 1998

Abstract Isotopic composition of dissolved inorganic carbon ? d13C DIC . in the Ottawa River basin is about y8 and y16‰ for lowland carbonate and upland silicate tributaries, respectively. This suggests that ?1. the source of DIC to the Ottawa River is soil respiration and carbonate weathering, ?2. exchange with the atmosphere is unidirectional or volumetrically unimportant, and ?3. in-river respiration and photosynthesis are not significant influences on the river carbon budget. Accepting these constraints, chemical and isotopic data are used to reconstitute soil p CO 2 for tributary catchments. Averages for upland silicate, mixed, and lowland carbonate basins are calculated to be roughly 2000, 5000, and 30,000 ppm, respectively. These values are used as input to model the pathway of carbon through the watershed—rain water to soil water to river water. The flux of carbon from the Ottawa River as DIC is calculated to be 4.3 = 10 10 mol Cra. Utilizing carbon isotopes, 75% and 25% of the Ca2qq Mg 2q flux is calculated to originate from carbonate and silicate weathering, respectively, and 61% of the DIC is calculated to originate from organic respiration. The latter represents some 6% of respired carbon in the basin, assuming an average respiration rate of 0.5 mmol C my2 hy1. Based on a diffusion model, CO 2 evasion to the atmosphere from the Ottawa River and its tributaries is estimated to be 1.3 = 10 10 mol Cra or 30% of the DIC flux. q 1999 Elsevier Science B.V. All rights reserved.
Keywords: Carbon; Isotope; River; Ottawa; Weathering; Watershed; p CO 2

1. Introduction Ri?ers and the carbon cycle. The impetus for this research comes mainly from some unresolved aspects of the global carbon budget which predicts that about 5.3 = 10 15 g of carbon is released annually
Corresponding author. Fax: q1-250-472-4184; e-mail: telmer@uvic.ca 1 Present address: School of Earth and Ocean Science, University of Victoria, Victoria, BC, Canada, V8W 3P6.
)

from burning of fossil fuels ?Berner and Berner, 1996.. About 2.8 = 10 15 g of this carbon accumulate as CO 2 in the atmosphere, resulting in the much-discussed greenhouse effect. Rivers, linking the terrestrial system and the ocean, transport yearly about 0.3–0.6 = 10 15 g of carbon in particulate and dissolved forms ?Degens et al., 1987.. Studies of dissolved CO 2 in some major rivers, such as the Amazon, Changjiang ?Yangtze. or Rhine ?Stallard, 1980; Kempe, 1982; Devol et al., 1987; Gao and Kempe, 1987; Richey et al., 1988; Buhl et al., 1991., show

0009-2541r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 9 - 2 5 4 1 ? 9 9 . 0 0 0 3 4 - 0

62

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

that CO 2 concentrations in these rivers are about 10–15 times greater than expected for an equilibrium with the atmosphere ?350 ppm.. Such high values are reached especially in the down-river sections of the rivers. These observations document that the rivers not only transport carbon from land to oceans, but also are actively degassing CO 2 into the atmosphere, with the residence time of CO 2 in the water approximated at roughly 4 days. The evasion rate is not yet quantified, but appears to be at least in the 10 13 gra range ?Kempe, 1982; Degens et al., 1991.. If the flux of CO 2 evasion from rivers approximates this magnitude, the consequences for the global carbon budget and for Global Change scenarios could be considerable. Such a large source would demand additional sinks; the identified sinks being already insufficient to accommodate some 1r2 of the anthropogenic CO 2 that is ‘missing’ in the atmosphere. In a review on the carbon cycle, Toggweiler ?1995. illuminates the problems, showing that the widely accepted oceanographic scenario of CO 2 uptake is not a solution to this dilemma and stresses the importance of the natural carbon cycle involving rivers. The major sources of carbon contributing to the total dissolved CO 2 in rivers are ?1. atmospheric CO 2 , ?2. CO 2 derived from the decay of organic matter and ?3. the dissolution of carbonate minerals. The d13 C of CO 2 in fresh water can be used to identify which of these source?s. predominates ?Tan, 1989.. For example, isotopic studies have confirmed that the CO 2 in the Amazon, Danube and the lower Rhine originates from decomposition of organic matter ?Richey et al., 1988; Buhl et al., 1991; Pawellek and Veizer, 1994.. Utilizing carbon isotopic techniques, this paper aims to quantify the sources and fluxes of carbon as it cycles through a large northern river basin, from rain water to soil water to river water, in an effort to understand which processes are dominant and to provide information useful for constraining the relative impact of riverine processes on the global carbon cycle. Basin characteristics. The Ottawa River is the largest tributary of the St. Lawrence River. The total drainage area of the Ottawa River basin above its outlet into the St. Lawrence is 149,000 km2 representing approximately 11.2% of the total drainage area of the St. Lawrence River basin ?1,315,000 km2 .. This drainage area is exceeded only by a few

Canadian rivers, namely: the Mackenzie ?1,805,000 km2 . and some of its tributaries ?Peace, Liard, and Athabasca., the Nelson ?1,070,000 km2 . and its tributary, the Saskatchewan ?414,000 km2 ., and by the Fraser ?232,000 km2 . ?Ontario Water Resources Commission and the Quebec Water Board, 1971.. It is the largest all-Canadian river in eastern Canada. The total length of the Ottawa River is 1160 km and its vertical descent is 365 m, producing an average slope of 0.315 mrkm ?Ottawa River Engineering Board, 1965.. Like many rivers that traverse recently glaciated terrain, the Ottawa River consists of a chain of lakes connected by sections of river, rapids and waterfalls. In many ways it can be considered a lacustrine river although not to the extent of the St. Lawrence system which includes the Great Lakes. It has a mean annual discharge of 1933 m3rs, mean annual precipitation of 872 mm and a mean annual temperature of 3.58C. The basin is underlain by 60% Precambrian plutonic rocks, 28% metamorphic rocks of Late Proterozoic age, 6% Paleozoic carbonate rocks, 4% Archean volcanics, and 2% Paleozoic clastic rocks ?Telmer, 1997.. The Paleozoic sediments reside mainly in the south of the basin and are extensively covered by Quaternary tills and glaciomarine sediments which have developed thick productive soils since the retreat of the Laurentian ice sheet. The physically resistant Canadian Shield, which predominates in the north of the basin, contains only thin and discontinuous tills and sediments and consequently soils are sparse and poorly developed. Fig. 1 shows the Ottawa River basin in the context of the Great Lakes–St. Lawrence system and Fig. 2 illustrates the geology of the basin. The Ottawa River has 28 significant tributaries, 24 of them below Lake Temiscaming, with an equal number in Quebec and Ontario. Of these 28 tributaries, 21 carry more than 95% of the cumulative tributary flow with the Riviere ` Gatineau being the largest. The 21 largest tributaries were selected for sampling in this study. 1.1. Sampling methodology Samples were taken along the entire length of the Ottawa River and from the mouths of its major tributaries during low water stand in September of 1991 and during high water stand in May of 1992.

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

63

Fig. 1. The Ottawa River basin in the context of the Great Lakes–St. Lawrence system. Statistics determined by GIS analysis ?Telmer, 1997..

During the first sampling campaign, 34 samples were collected, 16 from the Ottawa River at a spacing of 50–100 km, and 18 from major tributaries. Some minor adjustments to the sampling localities were made after viewing the preliminary results of the first sampling campaign such that during the second

campaign, 35 samples were collected, 16 from the Ottawa River, 19 from major tributaries. In addition to the ‘whole river’ sampling campaign, water was taken from one location near the mouth of the Ottawa River once a month for 13 months adding a temporal perspective to the study and allowing the

64

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

Fig. 2. Archean, Paleozoic and Pleistocene geology of the Ottawa River Basin. The extent of the incursion of the Champlain Sea—an arm of the Atlantic Ocean which occupied parts of the Ottawa River basin from 12,000 to 8500 B.P. due to Wisconsinan glacial isostatic depression—is also shown ?Champlain Sea data from Occhietti, 1989.. Leaching of the thick marine sediments that were deposited by this sea produces waters with elevated ionic strength ?Telmer, 1997..

nature and timing of the transition from high to low water stand to be determined. Fig. 3 shows the sampling localities for both sampling campaigns, the temporal monitoring station, and the sample procurement route. Vertical and longitudinal mixing concerns exerted a secondary influence on the choice of sample localities. Sanders et al. ?1983. state that rivers are usually sufficiently shallow and turbulent so that vertical

homogeneity is quickly attained below an influx of new material. Lateral homogeneity is attained more slowly. Thus wide, swift-flowing rivers may not be completely mixed for many kilometres below the influx point. Large inflows from tributaries can take up to 16 km to mix completely in the Columbia River ?Hem, 1985. and up to 150 km in the ‘longitudinally stratified’ St. Lawrence River ?Yang et al., 1996.. Sampling localities were chosen to minimize

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

Fig. 3. Basic hydrology, sampling localities and procurement route.

65

66

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

mixing concerns. Fortunately, areas where rivers are well mixed and where they are easily sampled are typically coincident. Bridges and ferry crossings, which make excellent sampling locations, are common where the river is narrow, swift and turbulent and hence well mixed. Below hydroelectrical generating stations the water is usually well mixed and there usually exists some infrastructure to allow sampling.

1.1.1. Collecting the sample River water samples were collected from bridges, hydroelectric dams, barrages, and ferries at a depth of 2 m and at a lateral position where the water was most likely to be well mixed. Generally, the centre of the rivers satisfied these conditions. Samples were collected using a narrow-mouthed glass bottle attached to a heavy metal retort stand ?sampling iron. and with a teflon groundwater bailer. The sampling iron was rapidly lowered to a depth of 2 m, preventing the collection of any surface water and floating materials ?litter, oils, etc.. and then slowly jigged between depths of 2 and 4 m producing a moderately ‘depth-integrated’ grab sample. Once full, the iron was raised, the glass bottle tightly capped and removed from the apparatus. The teflon bailer was also lowered to a depth of 2 m and jigged to allow through-flow for several minutes. When the bailer is raised, it automatically seals itself from the atmosphere preventing contamination from atmospheric oxygen and carbon dioxide. The bottles and bailer were then taken nearby where the field analyses were performed immediately. There are two main types of water samples typically collected by water quality surveys, ?1. grabsamples and ?2. composite-samples, each having different uses ?Hem et al., 1990.. Grab-samples characterize water quality at a particular time, and provide
Table 1 Average concentration of the Ottawa River at its mouth Naq ?m molrl. River water ?this study. River water ?Environment Canada, 1993. 139.0 127.0 Kq 19.8 26.0

information about minimum and maximum concentrations during the history of the sampling campaign. Composite samples are collected over a period of time and thus provide an estimate of average water quality over the period of sampling. Composite samples cannot detect changes in water quality that occur during the period of time represented by the composite. In this study, it would have been best to use a composite sample which represents a time interval slightly less than the interval being investigated ?on the order of days. rather than a grab sample which represents a time interval ?15 min. much less than the interval of interest. In this study, as for many water surveys, for practical reasons such as cost and field logistics, it was only possible to collect grab-samples. However, Taylor and Hamilton ?1994. assessed cross-channel and diurnal variation of solutes within sites, using data collected over 25 years at numerous sites throughout the Saskatchewan River basin—at 365,000 km2 , of a similar scale as the Ottawa River basin. Their results showed that the coefficient of variation ?CV. of multiple samples taken within 1 day at four sites was usually less than 5% ?analytical error is often of this magnitude. and always less than 10%, suggesting that daily mean concentration may be reliably estimated from a single grab sample. They also concluded that crosschannel variations were unlikely to be significant, except where tributary inflows have not yet mixed with the flow of the main river. To further validate the meaningfulness of data from this study, average concentrations of major dissolved species from this study were compared with those based on a much greater number of composite samples from the Envirodat water quality database ?Environment Canada, 1993. taken over a decade. Table 1 lists the averages. The two datasets strongly agree. Some differences are expected since the sampling campaigns were not synchronous.

Mg 2q 94.6 96.4

Ca2q 253.5 233.0

Cly 117.3 98.3

2y SO4

NOy 3 5.9 3.3

pH 7.0 7.1

91.9 100.8

Data from this study are averaged from 15 samples taken during 1993–1994; data from Environment Canada ?1993. are averaged from a maximum of 221 samples taken during 1978–1988 for pH and a minimum of 24 samples taken during 1979–1982 for Naq.

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

67

1.1.2. Sample preser?ation After being filtered through a 0.45 m Millipore membrane, duplicate samples were collected and stored at 48C for chemical and isotope analysis. Samples collected for anion analysis needed no further treatment. Samples for the determination of cations were acidified to 0.4% HNO 3 . For carbon isotopes, samples were collected in brown glass bot-

tles with air-tight polycone caps and poisoned with HgCl 2 to eliminate bacterial growth. Fig. 4 illustrates the analytical and preservation procedures used in the field and laboratory. 1.2. Laboratory analyses In the laboratory, major cations were analyzed by a Thermo Jarrell Ash Atomscan 25 Argon ICP ?AES..

Fig. 4. Preservation and analytical procedures used in the field and laboratory.

68

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86
q H 2 CO 3 s HCOy 3 qH 2y q HCOy 3 s CO 3 q H

The 13 Cr12 C ratios were determined by mass spectrometry ?VG SIRA 12. on CO 2 gas liberated by adding concentrated H 3 PO4 to the samples under vacuum. The d13 C are reported relative to Chicago PDB with a precision of "0.1‰. All these measurements were carried out at the Department of Earth Sciences, University of Ottawa. The results of chemical and isotopic measurements can be found in Appendix A. The partial pressure of CO 2 and the saturation index of calcite and dolomite in the river were calculated using the WATEQ computer program ?Truesdell and Jones, 1974.. The concentration of dissolved inorganic carbon ?DIC. was calculated from alkalinity and pH ?Skirrow, 1975..

? 2. ? 3.

These reactions are dependent on the activity of the hydrogen cation ?pH.. As a result, if the pH and partial pressure of CO 2 are known, the activity ?activity and concentration are essentially equivalent for dilute waters. of each carbonate species can be calculated from Henry’s Law, the law of mass action, and the expression for the equilibrium constant as follows: a H 2 CO 3 s K CO 2 p CO 2 2 a HC O y s K CO 2 p 3 CO 2 aHq

? 4. ? 5. ? 6.

2. Discussion Chemical system. The carbon cycle is the dominant control on acid-base reactions in the Ottawa River basin, as it is for most natural surface waters ?Drever, 1988.. It thus plays an important role in the behaviour of most dissolved chemical species, Cly and Naq being the main exceptions. DIC enters waters in the Ottawa River basin from primarily three sources: the atmosphere, soil and ground water respiration, and from dissolution of carbonate rocks. The first two sources involve equilibration of CO 2 with water which can be summarized by the following three equations ?Drever, 1988; Stumm and Morgan, 1996.: CO 2 q H 2 O s H 2 CO 3
2
2y s aCO 3

K 2 K 1 K CO 2 p CO 2
2 q aH

The sum of the activities of the three species is the total dissolved inorganic carbon ?henceforth DIC. ?Skirrow, 1975; Drever, 1988.. The values for the equilibrium constants are temperature dependent and are determined empirically by experiment. Regression of published data ?Drever, 1988. for the temperature range T s 0–308C yields the following algebraic expressions for K : K CO 2 s y2.22 = 10y6 T 3 y 1.91 = 10y5 T 2 q 1.63 = 10y2 T q 1.11 K 1 s 1.67 = 10y4 T 2 y 1.34 = 10y2 T q 6.58 K 2 s y2.22 = 10y6 T 3 q 2.29 = 10y4 T 2

? 7. ? 8.

? 1.

For convenience, it has been common practice in the literature to lump together the hydration of CO 2 ?g. into CO 2 ?aq. and the protolysis of CO 2 ?aq. into H 2 CO 3 ?aq., using the composite equilibrium constant K CO 2 , and to express the sum of the activities of CO 2 ?aq. and H 2 CO 3 ?aq. simply as the activity of H 2 CO 3 ?aq.. The true activity of H 2 CO 3 ?aq. is lower and can be determined from the relationship K CO 2 s K H 2 CO 3 r?1q K ., where K is the constant describing the hydration equilibrium ?Stumm and Morgan, 1996.. For the purposes of this study, the use of the composite equilibrium constant K CO 2 is sufficient and henceforth the activity of H 2 CO 3 ?aq. includes the activity of CO 2 ?aq..

2

y 1.62 = 10y2 T q 10.6

? 9.

2.1. Watershed carbon distribution 2.1.1. Rain water DIC Using Eqs. ?4. – ?9., a p CO 2 of 350 ppm for the modern atmosphere ?Schlesinger, 1991., a pH of 5

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

69

?Schindler et al., 1976; Barrie and Hales, 1984. and a temperature of 3.58C for average rain water in the Ottawa River basin ?Telmer, 1997. and ignoring the small influences of other ions, the activity of the carbonate species in rain water is calculated to be: y 6.21 2y H 2 CO 3 s 10y 4.75 ; HCO y ; CO 3 s 3 s 10 y1 1.7 10 ; making the average DIC of precipitation in the Ottawa River basin ; 10y4 .73 or 1.84 = 10y5 molrl which falls into the typical range of 10y5 to 5 = 10y5 for DIC in rain water ?Stumm and Morgan, 1996.. 2.1.2. Rain water d13C The d13 C for DIC in rain water is a function of the atmospheric d13 C composition and the fractionation of carbon isotopes between atmospheric CO 2 and aqueous carbonate species. The former ranges from y6 to y8‰ PDB ?Cerling et al., 1991. and has an average of y7‰ PDB ?Faure, 1986. and the latter ranges from 9.2 " 0.4‰ at 08C to 6.8 " 0.4‰ at 308C ?Deuser and Degens, 1967; Halas et al., 1997.. Using an average temperature of 3.58C and an average atmospheric d13 C value of y7‰, waters in equilibrium with the atmosphere in the Ottawa basin, including average rain water, should have d13 C of about q1.4‰ PDB. 2.1.3. Soil water DIC Due to root respiration, oxidation of organic material and bacterial respiration, the soil atmosphere is considerably richer in CO 2 than the atmosphere. Reardon et al. ?1979. have observed consistent values of 0.4% CO 2 at depths greater than 6 m in sandy calcareous soils in southern Ontario, soils similar to those of the lowlands in the Ottawa River basin. At depths less than 6 m, seasonal variations were observable with the maximum variation occurring at 2 m, 0.1% CO 2 in the winter and ) 1% in the summer. p CO 2 values as high as 10y0 .5 atm ?; 30%. have been observed in soils ?Stumm and Morgan, 1996.. Utilizing Eq. ?4., soil waters in equilibrium with 1% CO 2 would have DIC of 340 m molrl. 2.1.4. Soil water d13C The d13 C of soil CO 2 is derived primarily from 13 d C of organic material which typically has a value of y24 to y34‰, with an average of y28‰

?Faure, 1986.. No fractionation occurs when this material is respired to produce CO 2 gas. The difference in the diffusion coefficients of 12 CO 2 and 13 CO 2 cause soil gases to be enriched in 13 CO 2 by up to 4.4‰ ?Cerling et al., 1991., relative to the organic source material. As a consequence, the average soil gas has d13 C of y30 to y19‰, with an average of ; y23‰. The dissolution of this gas in soil water is subject to the same fractionation as the above-discussed dissolution of atmospheric CO 2 into rain water giving average soil waters a d13 C of y23 to y13‰, with an average of y17‰.

2.1.5. Ri?er water DIC Fig. 5 shows the downstream trend of DIC values in the Ottawa River basin. The distribution of DIC in the Ottawa River basin is strongly controlled by geology. On average, the DIC of the tributaries ranges from about 45 m molrl for those draining purely silicate terrain in the headwaters to about 3000 m molrl for those draining carbonate catchments. The values for the Ottawa River reflect the mixing of its tributaries, falling between 55 m molrl in its headwaters and 700 m molrl at its mouth. The Pearson correlation coefficient of DIC and the abundance of carbonate rocks in tributary basins is 0.89. The dilute waters in the uplands, with an average DIC of 55 m molrl, are only 4 times more concentrated than rain water and, after evapotranspiration is accounted for ?calculated to be 42% for upland basins; Telmer, 1997., ; 45% of the DIC can be explained by precipitation input. Note, however, that this calculation is less meaningful than for an anion like Cly because DIC is much less conservative and is included here only for perspective. These low values can be explained by two reasons: ?1. soils are thin and discontinuous in the Canadian Shield ?Kaszycki and Shilts, 1987. limiting the production of soil CO 2 ; ?2. lower growth rates in the harsh northern parts of the basin yield less CO 2 from root respiration; and ?3. low weathering rates. Weathering of minerals consumes CO 2 by transforming it into HCOy 3 . By reducing the p CO 2 of water, weathering increases the difference between the partial pressure of CO 2 in the gas and the liquid phase, thus increasing the rate of CO 2 diffusion into the liquid. For example, the weathering of albite ?Eq. ?10.. and

70

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

Fig. 5. DIC for the Ottawa River and its major tributaries versus distance downstream. Monthly variations at the mouth of the Ottawa River are also shown.

limestone ?Eq. ?11.. consumes CO 2 enriching product waters in DIC. 2NaAlSi 3 O 8 q 2CO 2 q 3H 2 O s Al 2 Si 2 O5 ? OH . 4 q 2Naqq 2HCOy 3 q 4SiO 2 ? 10 . CaCO 3 q H 2 O q CO 2 s Ca2qq 2HCOy 3

bonates, tills, clays and soils, has the warmest climate of all tributary basins, and is heavily used for agriculture which enhances weathering by increasing water–rock interaction by churning soils. 2.1.6. Ri?er water d13C The d13 C DIC along the length of the Ottawa River ranged from y17.4‰ in the headwaters to y8.4‰ in the lowlands in the fall and from y17.3 to y7.3‰ in the spring. In general, the spring values are lighter ?Fig. 6.. The d13 C DIC values for the tributaries fall within the same range with the exception of some lowland tributaries that have slightly heavier values of up to y6.6‰. The values for the Ottawa River and its tributaries are comparable to, or slightly lighter than, those for the Rhine ?y10.5 to y7.1‰. and its tributaries ?y14.4 to y6.7‰. ?Buhl et al., 1991; Flintrop et al., in press.. The values of the carbonate bedrock dominated lowland tributaries and of the lower reaches of the river are also compa-

? 11 .

Weathering rates in the Canadian Shield are relatively low ?Frape and Fritz, 1987. due to the resistant nature of the bedrock and the lack of soils and fine grained materials, which limits water–rock interaction, and thus waters draining these areas contain little DIC. Further down-stream, as more weatherable rocks, such as Proterozoic marbles and Paleozoic carbonates ?see Fig. 2., and more weatherable materials, such as soils and tills, become plentiful, DIC increases. The highest DIC was recorded in the South Nation River which drains a basin abundant in car-

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

71

Fig. 6. Carbon isotopic composition of dissolved inorganic carbon ? d13 C DIC . in the Ottawa River basin.

rable, but slightly lighter than, for the carbonate bedrock dominated Danube and Fraser Rivers ?; y7.5‰. ?Pawellek and Veizer, 1994; Cameron et al., 1995.. These 13 C-depleted values show that respiration plays an important role in the isotopic composition of waters in the Ottawa River basin. In fact, it will be shown that respiration provides half or more of the DIC in all these waters. 2.2. Sources of carbon Organic respiration and carbonate dissolution. The sources of carbon in river water are primarily: ?1. CO 2 enriched soil waters, ?2. the dissolution of carbonates, ?3. direct inputs from precipitation, ?4. exchange with atmospheric CO 2 , and ?5. photosynthesisrrespiration within the river. As discussed earlier, direct inputs from precipitation are certainly minuscule and accepting, for the moment, that sources ?4. and ?5. are inconsequential or secondary, we are left with only two endmembers, carbon from the respiration of organic material in soils and from carbonate weathering. These two endmembers are

easily determined. As discussed in the section on soil water, the respiration endmember has a d13 C DIC of ; y17‰. The weathering endmember is determined from the d13 C of basinal carbonates and the stoichiometry of the weathering reaction. Carbonates in the Ottawa Embayment of the St. Lawrence Lowlands have d13 C of y5.3 to q1.9‰, with an average of 0‰ ?Brand and Terasmae, 1984; Taylor and Sibley, 1986; Fortin and Tasse, 1987; Middleton et al., 1990.. The dissolution of carbonate by carbonic acid: CaC carbO 3 q H 2 C respO 3 s CA2qq 2HC cw Oy 3

? 12 .

where C carb is carbon from carbonate with a d13 C of 0‰ and C resp is carbon from the respiration of organic material with a d13 C of y17‰, yields a d13 C of y8.5‰ for the carbonate weathering endmember ?C cw .. The carbon isotope composition of all waters in the Ottawa basin can be explained by these two endmembers ?Fig. 6.. Minor deviations can result from either the variability in the endmember composition ?e.g., carbonate rocks with d13 C of y5‰ or

72

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

organic material with d13 C of y14‰ from C4 plants. or from fluxes of carbon from other sources, such as an exchange with the atmosphere or photosynthesisrrespiration in the river. These fluxes are almost certainly of secondary importance since the observed isotopic discrepancies from the model are within "1.6‰. However, the temporal behaviour of dissolved oxygen and d13 C DIC support this suggestion more conclusively. Dissolved oxygen is undersaturated in the fall and oversaturated in the spring ?see Appendix A., the opposite of what should occur if the rate of photosynthesis exceeded that of respiration in the fall. Seasonal d13 C DIC trends are indistinct if compared to the spatial ones, with the average difference between spring and fall samples for the same locality being less than 1‰. Fig. 7 shows the fluctuations of d13 C DIC at the mouth of the Ottawa River for the year 1993r1994. The most 13 C depleted values occur in the summer, the opposite of what would occur if in situ photosynthesis were influencing the signal.

This phenomenon can be explained hydrologically. In the summer, a greater percentage of discharge is provided by the upland tributary basins, which have a more depleted signal. During spring and late fall when the hydrograph rises, greater percentages of water come from the lowland tributaries that have relatively 13 C enriched values. This behaviour also suggests the sources of DIC in the basin are relatively fixed and operate on time scales longer than 6 months. The fact that d13 C DIC values are slightly enriched 13 in C in the fall at most localities in the bi-seasonal ?springrfall. data, as would be expected for in-river photosynthesis, does suggest that photosynthesisrrespiration could influence the carbon budget in a secondary nature. However, Schiff et al. ?1990. and Aravena et al. ?1992., have documented that soil d13 C DIC values in the silicate Harp Lake watershed ?50 km outside the Ottawa River basin. have a seasonal cycle where they are lightest in the winter ?y25‰. and heaviest in the summerrfall ?y23‰..

Fig. 7. Monthly d13 C DIC at the mouth of the Ottawa River.

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

73

They attribute this to lower losses by diffusion to the atmosphere during winter because of the snow cover, which allows a slight accumulation of CO 2 in the soil zone. This seasonal 2‰ difference in the source of d13 C DIC can fully account for the seasonal differences observed in the waters of the Ottawa River basin ?"- 2‰., suggesting that fluctuations in soil and ground water conditions are a stronger secondary control on the d13 C DIC of waters in the Ottawa River basin than in-river photosynthesisrrespiration. Silicate ?ersus carbonate weathering. Fig. 8 plots DIC versus d13 C DIC and divides the plot into fields that can be explained as follows. Since silicate weathering does not yield any additional carbon from a solid phase ?Eq. ?10.. it fixes only soil CO 2 . It also proceeds at a much slower rate than carbonate weathering, and therefore produces waters with low DIC and ‘respiration’ d13 C DIC values of ; y17‰. These waters plot to the lower left of Fig. 8. In a

limited system ?fixed amount of CO 2 or carbonate material., such as might occur in sparse soils or the wintertime when soil CO 2 production is much lower, the weathering of carbonate would proceed until the supply of CO 2 is exhausted, producing waters with moderately increased DIC and d13 C DIC of ; y8 to y10‰ in accordance with dominantly carbonate weathering. Such waters are circled in the upper left of Fig. 8. A similar water could be produced by catchments with only limited amounts of carbonate. Runoff from the remaining areas of silicate bedrock would dilute the DIC concentration but because silicate waters are so low in DIC, the d13 C signal would carry a dominantly carbonate weathering signal. It is likely that both scenarios are responsible for the waters in question, since limited CO 2 supplies occur in the areas underlain by silicate geology in the Ottawa River basin. Waters in contact with a generous supply of both CO 2 and carbonate plot in the upper centre and

Fig. 8. DIC versus d13 C DIC for the Ottawa River and its tributaries. The points that lie in the ‘excess respiration carbon’ field represent waters draining lowland carbonate catchments.

74

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

Fig. 9. Calcite saturation index versus d13 C DIC for the waters of the Ottawa River basin.

upper right circular portion of Fig. 8. These waters dissolve carbonate until calcite saturation is reached at which time, if CO 2 is sufficiently abundant, they begin to acquire a stronger respiration d13 C DIC signal. Fig. 9, which plots calcite saturation versus d13 C DIC , illustrates the relationship between calcite saturation and respiration more clearly. 2.3. Carbon dioxide 2.3.1. Spatial distribution The partial pressure of carbon dioxide ? p CO 2 . in the Ottawa River and its tributaries is a function of soil respiration, which increases p CO 2 and lowers pH, and mineral weathering, which consumes Hq and converts carbonic acid into bicarbonate, increasing pH and lowering p CO 2 ?Eqs. ?10. and ?11... The relative effectiveness of these processes varies spatially in the basin with the abundance of carbonates and soils. Upland silicate basins are characterized by relatively low soil respiration levels due to

their sparse soils but, because they are resistant to chemical weathering, the neutralization of carbonic acid is limited and p CO 2 values remain at or higher than atmospheric levels. Lowland carbonate basins have thick productive soils but, because carbonate weathering rates are high, the neutralization of carbonic acid is rapid yielding p CO 2 values not much higher than those of the upland silicate basins. Fig. 10 illustrates the downstream trend of p CO 2 . 2.3.2. Temporal distribution Seasonally, p CO 2 is higher in the spring than the fall. Three reasons can account or contribute to this. ?1. Hysteresis, where waters rich in soil CO 2 are flushed from the system during seasonal flooding. ?2. Rain and snow have low pH ?; 5. and thus spring flooding releases a wintertime of stored acid into the drainage network driving the p CO 2 up. This can be augmented by anthropogenic contributions to the acidity of precipitation or what is commonly referred to as ‘acid rain’ which can be considerable

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

75

Fig. 10. p CO 2 versus distance downstream for the Ottawa River and its tributaries.

in the Ottawa River basin ?Environment Canada, Lands Directorate, Environmental Conservation Service, 1985; Environment Canada, Inland Waters and Lands Directorate, 1988.. ?3. Photosynthesis in the summer and fall consumes more CO 2 than respiration contributes, lowering the p CO 2 of fall waters. The fluctuations in pH and p CO 2 at the mouth of the Ottawa River ?Fig. 11. are of too large a magnitude and occur too early in the spring to be explained by respiration or photosynthesis alone, suggesting that hysteresis and spring acid release are the main controls on the seasonal fluctuations in p CO 2 . The d13 C DIC values at the mouth of the river ?Fig. 7., as discussed above, also belie photosynthetic CO 2 drawdown in the fall. 2.3.3. Origin of pCO2 Fig. 12 plots carbon isotopes against p CO 2 . Pathways and processes for the conversion of soil water into observed river waters are shown. The story is similar to that of DIC but, importantly, is also sensi-

tive to pH. Silicate weathering consumes CO 2 but contributes no additional carbon to waters ?Eq. ?10... As a result, p CO 2 values are reduced but the d13 C DIC of resulting waters retains a respiration signal. These waters plot in the lower left of Fig. 12. Carbonate weathering ?Eq. ?11.. proceeds at much higher rates than silicate weathering and contributes d13 C carbonate of ; 0‰ and under limited conditions ?fixed CO 2 ., CO 2 is rapidly consumed, reducing p CO 2 levels to below equilibrium with the atmosphere and increasing the d13 C DIC of waters to ; y8‰ ?upper left of Fig. 12.. Under open conditions, such as would exist in productive soils, equilibrium with soil p CO 2 would produce waters with p CO 2 higher than atmospheric equilibrium. This would be exacerbated if calcite saturation were reached. These waters are located in the right side of the ‘carbonate weathering’ box. The mixing of waters controlled by carbonate weathering and waters controlled by silicate weathering, such as is the case for the downstream stretches of the Ottawa River, produces waters with

76

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

Fig. 11. Monthly p CO 2 and pH at the mouth of the Ottawa River, 1993r1994.

d13 C DIC of ; y11‰ and high p CO 2 ?centre-right of Fig. 12.. The d13 C DIC values are closer to that of ‘carbonate’ waters because of the greater flux of DIC from those waters. The high p CO 2 results from lowering the pH by mixing low pH waters of the uplands with high pH, high DIC waters of the lowlands. The direction of shifts caused by in-river photosynthesisrrespiration are also shown on the plot, although, as already discussed, there is no evidence to suggest that these processes affect the p CO 2 or the d13 C DIC signals to a large degree.
2.3.4. pCO2 of soil gas It is possible to reconstitute the soil atmosphere that must have been in equilibrium with soil water to produce the observed carbon chemistry of river water. This can be calculated from the river’s d13 C DIC and DIC concentration if we use the evidence presented above and assume that: ?1. in-river photosynthesisrrespiration are insignificant; ?2. carbonate precipitation is not a major carbon sink. Only a few

tributaries are weakly calcite oversaturated and this state is usually seasonal, suggesting that this assumption is reasonable. If calcite precipitation were a major sink for riverine carbon, the calculated p CO 2 would be a minimum; ?3. exchange of CO 2 with the atmosphere is minimal. The lack of an atmospheric d13 C signal ?f 0‰. in the carbon isotope data suggests that this is the case; ?4. as calculated earlier the d13 C DIC of average soil water is y17‰; and ?5. as discussed, the d13 C DIC contributed from carbonates is 0‰. The expressions for the reconstituted p CO 2 of the soil atmosphere are: H 2 CO 3? biological . s DIC p CO 2 ? soil . s

?

d13 C DIC d13 C carb q d13 C resp

/

? 13 . ? 14 .

H 2 CO 3? biological . K CO 2

where d13 C DIC and DIC are the values measured in the river; d13 C carb is the value of d13 C DIC con-

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

77

Fig. 12. d13 C DIC versus p CO 2 for the Ottawa River and its tributaries and pathways for the conversion of soil water into river water.

tributed from carbonate weathering; d13 C resp is the value of d13 C DIC contributed from respiration of soil organic material; and H 2 CO 3?biological. is the carbonic acid calculated to have originated from biological carbon. Eq. ?14. converts the carbonic acid originating from biological material to the p CO 2 of the soil atmosphere which must have been present to produce the observed river waters. Table 2 lists the reconstituted soil p CO 2 for major tributary basins as well as the whole Ottawa River basin. The average for lowland carbonate catchments, upland silicate catchments, and mixed catchments is 10y1 .5, 10y2 .85, and 10y2 .3 ppm, respectively. As expected, the upland silicate basins that have only sparse soils have low soil p CO 2 while the lowland basins and their well-developed soils generate relatively high soil p CO 2 , the mixed basins fall in-between. In reality, it is likely that the existing soils in the upland basins have higher p CO 2 than that listed in Table 2 but inputs from such soils are diluted by soil poor or soil absent areas. The

same is likely true for the lowland and mixed basins but to a lesser degree. 2.3.5. CO2 e?asion to the atmosphere Riverine DIC losses can occur by CO 2 evasion to the atmosphere, CO 2 uptake by aquatic plants, and precipitation of carbonate minerals. As discussed previously, any signals caused by CO 2 consumption by aquatic plants are undetectable and the Ottawa River remains calcite undersaturated year round, leaving evasion to the atmosphere as the only possible sink term. According to Mook ?1970., up to 6 months are required for some lakes to reach isotopic equilibrium with the atmosphere. Yang et al. ?1996. found that Lake Ontario and the main channel of the St. Lawrence River were roughly in isotopic equilibrium with the atmosphere owing to the long residence time ?; 100 years. of water in the Great Lakes where the majority of St. Lawrence water originates. There are no estimates of mean residence time for

78

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

Table 2 Hypothetical p CO 2 of soil atmosphere for the Ottawa River basin and its tributary subbasins Basin Ottawa River Camachigama Kenojevis Blanche Montreal Kipawa Mattawa Petawawa Dumoine Noire Coulonge Bonnechere Madawaska Mississippi Rideau Gatineau Riviere ` du Lievre Petite Nation South Nation Rouge Nord Rigaud p CO 2 ?ppm. 7769 1108 6186 12,371 4561 722 2298 3036 1559 2846 1842 16,855 8930 18,892 23,864 1758 4704 6235 42,728 2557 16,167 52,128

values attainable by carbonate weathering are not evident. Clearly, riverine carbon exchange with the atmosphere is dominantly unidirectional evasion to the atmosphere. Some attempts at calculating fluxes of CO 2 from rivers to the atmosphere exist in the literature. For the Rhine River, between 1963 and 1978, the mean p CO 2 and net flux of CO 2 to the atmosphere was estimated to be 3300 ppm and 9.16 = 10 9 molra, respectively ?Kempe, 1982.. In 1976–1977, p CO 2 values of 4000 ppm and a net flux of 1.5 = 10 11 molra were estimated for the Amazon River ?Stallard, 1980.. Yang et al. ?1996. found that p CO 2 and CO 2 fluxes in the St. Lawrence River fluctuated from 576 to 207 ppm and 10 8 and y4 = 10 7 molr day in the spring and fall, respectively, and that this essentially mimics biologically controlled p CO 2 conditions and fluctuations in the epilimnion water of the Great Lakes. Note that the fall term is negative indicating an influx from the atmosphere to the river. These calculations are based on a theoretical diffusion model of the CO 2 flux between river and atmosphere and can be calculated from the following formula ?Broeker, 1974.: F s D ? Cair y C water . rZ

water in the Ottawa River but dye tracer studies ?Merritt, 1964. and a tritium spill at the Chalk River nuclear research facility ?Natural Resources Canada S.T.F., 1994. document peak concentrations of dye and tritium at Cheneaux station, 110 km downriver of the injection, after roughly 10 days, a rate of 11 kmrday. The Ottawa River widens to a lake-like 2.5 km width through much of this section and water movement rates are thus likely much slower through this section than through the remainder of the river. If we assume an average water movement rate of 15 kmrday then water originating in the headwaters ?1000 km from the mouth. would leave the basin in roughly 2–3 months. Distances from the headwaters of most tributaries to the mouth of the Ottawa River are also roughly 1000 km and so the same residence time is applicable to them. Two to three months is a maximum since only a small fraction of the water in the Ottawa River and its tributaries originates in their headwaters. Two to three months is not enough time to attain atmospheric isotopic equilibration according to the Mook estimate and the carbon isotopes bear this out, as shifts toward heavy d13 C DIC beyond

? 15 .

where ? Cair – C water . is the concentration difference of gas between the overlying air and the bulk of the water, D is the gas- and temperature-specific diffusion coefficient, and Z is the thickness of the boundary layer: a thin film existing at the air–water interface. Z depends largely on wind speed ?Broeker et al., 1978. and water turbulence ?Holley, 1977. and ranges from 3 = 10y5 m for the world ocean to 2 = 10y4 m for small fresh water lakes. DrZ, therefore, is the gas exchange rate, which gives the height of a water column that will equilibrate with the atmosphere per unit time ?e.g., mrday.. Z for the Ottawa River is estimated to be 10y4 m, in-between the quiescent environment of a small fresh water lake and that of the world ocean, and D is 10y4 m2rday at 108C ?Kempe, 1982. yielding a gas exchange rate DrZ of 1 mrday. Depths of the Ottawa River range from 10 m to 50 m with the average roughly 13 m ?Natural Resources Canada S.T.F., 1994.. Accordingly, the Ottawa River would need 13 days to equilibrate with the atmosphere. Perhaps

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

79

constant inputs of baseflow prevent this from happening. Using a DrZ of 1 mrday, Cair of 350 ppm ?1.60 = 10y8 mol CO 2rcm3 . and a mean p CO 2 of the Ottawa River over its length ? C water . of 1200 ppm ?5.49 = 10y8 mol CO 2rcm3 ., the calculated flux of CO 2 from the Ottawa River to the atmosphere is 3.89 = 10y6 mol cmy2 dayy1 or 14.2 mol my2 ay1 . Estimating the river to be 400 m wide and 1100 km long, the annual CO 2 evasion would be 6.2 = 10 9 mol C. Evasion from the tributaries would likely be as large making the CO 2 evasion from both sources ; 1.3 = 10 10 mol Cra. This represents 30% of the river’s annual DIC budget. Considering the many assumptions needed to arrive at this figure ?for example, the p CO 2 at the mouth of the Ottawa River varies by an order of magnitude from 1200 to 12,000 ppm and so arriving at a reliable average p CO 2 for the entire length of

the river is prone to error., it can represent only a hunch for the order of magnitude of evasion. It is, however, worth mentioning that if this were to represent typical degassing behaviour for all rivers, CO 2 evasion to the atmosphere would be in the 10 14 gra range in contrast to earlier estimates of 10 13 gra ?Kempe, 1982; Degens et al., 1991.. 2.4. Watershed carbon cycling 2.4.1. The hydrological pathway Using the data from Table 2 as a guide for soil conditions and Eqs. ?4. – ?6., the DIC species and concentrations through the hydrologic cycle, from rain water to soil water to river water, can be modelled using the following boundary conditions: ?1. in the soil zone, DIC is proportional to p CO 2 ; ?2. for carbonate weathering, DIC at calcite saturation s

Fig. 13. Evolution of dissolved inorganic carbon for hypothetical average water from ?A. rain water to ?B. soil water to ?C. calcite saturation and toward ?D. river water for weathering in lowland carbonate catchments; from ?A. rain water to ?M. soil water and towards ?N. river water for weathering in mixed silicatercarbonate catchments; and from ?x. rain water to ?y. soil water and towards ?z. river water for upland silicate catchments. Dissolved inorganic carbon species lines are for DIC s 10y3 .5 molrl.

80

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

2DIC in initial soil water ?DIC calcite saturation s 2DIC soil ?3. for silicate weathering, total DIC s DIC in initial soil water ?DIC total s DIC soil atmosphere .. The relationship among p CO 2 , pH, and calcite saturation was calculated from the following equation:
atmosphere .;

literature ?Reardon et al., 1979; Drever, 1988; Schiff et al., 1990; Aravena et al., 1992; Stumm and Morgan, 1996.. 2.4.2. The balance of carbon The flux of DIC along the length of the Ottawa River and for its tributaries for the spring and fall is shown in Fig. 14. The flux from the lowland basins is dominant in the spring, with the South Nation River contributing the most. In the fall the Riviere ` Gatineau is the major contributor. This is purely a consequence of hydrology where the flow of the Riviere ` Gatineau is much less variable than flows from the lowland rivers, such as the South Nation and Rideau Rivers, and thus provides the lion’s share of the discharge in the fall. This annual hydrological pattern is the primary process responsible for elevated concentrations of DIC and other major ions at the mouth of the Ottawa River in the spring and fall. The average flux of DIC from the mouth of the Ottawa River is 1373 molrs or 4.33 = 10 10 molra.

a3 Hqs

2 2 2 K2 K1 K CO 2 p CO 2

2 K cal

? 16 .

where K cal is the equilibrium constant for calcite dissolution and all other terms are as in Eqs. ?4. – ?6.. Activity coefficients for all species were assumed to be 1 so that Debye–Huckel considerations were avoided. This is a reasonable assumption for dilute waters. Fig. 13 shows the results of the model plotted on modified Bjerrum plot along with the Ottawa River basin data. Almost all of the data can be explained by the three endmember curves suggesting that the estimates for the p CO 2 of the reconstituted soil atmosphere based on d13 C DIC as well as the above boundary conditions are reliable. The estimates fall well within the range documented in the

Fig. 14. Flux of DIC from the Ottawa River and its tributaries.

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

81

This represents 0.16% of the annual global DIC flux from rivers of 2.67 = 10 13 mol ?Ludwig et al., 1996. and 11% of the annual DIC flux of the St. Lawrence River of 3.9 = 10 11 mol C ?Yang et al., 1996. into which the Ottawa River flows. Eq. ?13. can also be used to calculate the concentration of DIC attributable to respiration and to carbonate dissolution. For the Ottawa River, this amounts to 61% or 2.65 = 10 10 mol Cra and 39% or 1.68 = 10 10 mol Cra, respectively. Utilizing this calculation in conjunction with the stoichiometry of carbonate dissolution, it follows that 75% of the total wCa2qq Mg 2q x originates from carbonate weathering and 25% from silicate weathering plus atmospheric and other inputs. This calculation may be significant in terms of the global carbon budget. Since only silicate weathering consumes CO 2 over the long term versus the more ephemeral CO 2 consumption produced by carbonate weathering ?Walker and Hays, 1981., this method can theoretically be applied globally to help constrain the global silicate versus carbonate weathering rate, ultimately providing an estimate of the long-term ‘silicate’ versus short-term ‘carbonate’ CO 2 draw-down. The separation of DIC into its respiration and carbonate dissolution components also allows us to estimate the rivers relative importance as a transporter of respired carbon. If we crudely assume that the average respiration rate for the whole basin is 0.5 mmol C my2 hy1 , based on the data published by Cerling et al. ?1991., then total annual respiration for the Ottawa River basin with an area of 149,000 km2 will be 6.5 = 10 11 mol Cra. This means that the river transports 6.6% of respired carbon from the basin, the remainder diffusing directly to the atmosphere from the soil. If the CO 2 that evades to the atmosphere from the surface of the river and its tributaries is included, the estimate becomes 8.6%.

3. Conclusion Spatially, the carbon chemistry of waters in the Ottawa River basin is controlled by the rates and products of rock weathering, which in turn are controlled by the susceptibility of different minerals to weathering as well as by soil respiration rates, the latter controlling the amount of carbonic acid avail-

able for weathering. Chemical signals caused by respiration and photosynthesis within the river itself are not evident or are secondary. The generation of carbonic acid by organic respiration in the soil zone is much higher in the well-developed soils of the lowlands than in the sparse soils of the uplands. A hypothetical soil p CO 2 for upland silicate, mixed, and lowland carbonate basins is calculated to be roughly 2000, 5000, and 30,000 ppm, respectively, although locally higher values can be expected. Waters draining these soils react with minerals in the soils and bedrock until either calcite saturation is reached, as is the case for the lowland carbonate tributaries, or until they discharge from the waterrrock interface as is the case for the silicate basins. In mixed basins, both processes occur and the resulting waters represent a mix of the two endmembers. This is the case for the Ottawa River. For an entirely carbonate basin, the relative contribution to the DIC budget from organic respiration and carbonate dissolution would be 50% from each source. The Ottawa River basin is underlain by only ; 8% calcareous rocks, yet 13 C mass balance calculations show that 61% of the DIC in the Ottawa River originates from organic respiration and 39% from carbonate dissolution. This illustrates the resistant nature of silicate rocks and the dominant role of carbonates in controlling the chemistry of surface waters. From this calculation and from the stoichiometry of carbonate dissolution, it follows that 75% of the total wCa2qq Mg 2q x originates from carbonate weathering and 25% from silicate weathering plus atmospheric and other inputs. Since only silicate weathering consumes CO 2 over the long term versus the more ephemeral CO 2 consumption provided by carbonate weathering, this isotopic method can be used to help constrain the global silicate versus carbonate weathering rate allowing the determination of long-term ‘silicate’ versus short-term ‘carbonate’ atmospheric CO 2 draw-down. Calculations based on a diffusion model produce a first order estimate of CO 2 evasion to the atmosphere. For the Ottawa River and its tributaries it is ; 1.3 = 10 10 mol Cra or roughly 30% of the DIC flux. Carbon isotopes suggest that this CO 2 exchange with the atmosphere is unidirectional ?evasion to the atmosphere. supporting the supposition that rivers are a source and not a sink of CO 2 to the

82

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

83

84

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86

85

atmosphere, further exacerbating the existing ‘missing’ CO 2 sink problem. Interestingly, if all rivers evaded 30% of their DIC flux to the atmosphere as CO 2 as the Ottawa River does, then the flux of CO 2 to the atmosphere from rivers would be 1.3 = 10 14 gra an order of magnitude higher than the earlier estimates. The Ottawa River discharges 4.3 = 10 10 mol Cra as DIC, representing 0.13% of the global DIC riverine flux and 11% of the flux from the St. Lawrence River. Carbon isotopes indicate that 61% of this is derived from respiration of soil organic material and, assuming a respiration rate of 0.5 mmol C my2 hy1 for the basin, the Ottawa River transports about 6.6% of total respired carbon from its basin, the remainder evading directly to the atmosphere from soils. Acknowledgements We gratefully acknowledged the following for their support and contribution to this work: The Natural Sciences and Engineering Research Council of Canada for financial support; Chao Yang and Hairou Qing for assistance with field work; Gilles St. Jean, Margaret Mclaren, Natalie Morrisette, Wendy Abdi, and John Loop for laboratory support; the Geological Survey of Canada, Environment Canada and Agriculture Canada for providing baseline data. References
Aravena, R., Schiff, S.L., Trumbore, S.E., Dillon, P.J., Elgood, R., 1992. Evaluating dissolved inorganic carbon cycling in a forested lake watershed using carbon isotopes. Radiocarbon 34, 636–645. Barrie, L.A., Hales, J.M., 1984. The spatial distributions of precipitation acidity and major ion wet deposition in North America during 1980. Tellus 36, 333–355. Berner, E.K., Berner, R.A., 1996. Global Environment, Water, Air, and Geochemical Cycles. Prentice-Hall, Upper Saddle River, NJ, pp. 27–61. Brand, U., Terasmae, J., 1984. Source rock geochemistry of Pleistocene tills of Southern Ontario. Ont. Geol. Survey Misc. Pap. 121, 65–71. Broeker, W.S., 1974. Chemical Oceanography. Jovanovich, Harcourt-Brace. Broeker, H.C., Peterman, J., Siems, W., 1978. The influence of wind on CO 2 exchange in a windwave tunnel, including the effects of monolayers. J. Mar. Res. 36, 595–610. Buhl, D., Neuser, R.D., Richter, D.K., Riedel, D., Roberts, B.,

Strauss, H., Veizer, J., 1991. Nature and nurture: environmental isotope story of the river Rhine. Naturwissenschaften 78, 337–346. Cameron, E.M., Hall, G.E.M., Veizer, J., Krouse, H.R., 1995. Isotopic and elemental hydrogeochemistry of a major river system: Fraser River, British Columbia, Canada. Chem. Geol. 122, 149–169. Cerling, T.E., Solomon, D.K., Quade, J., Bowman, J.R., 1991. On the isotopic composition of carbon in soil carbon dioxide. Geochim. Cosmochim. Acta 55, 3403–3405. Degens, E.T., Kempe, S., Gan-Wei-Bin, 1987. Transport of Carbon and Minerals in Major World Rivers, Part 4. Vol. 64. Mitt. Geol.-Palaont. Inst. Univ. Hamburg, SCOPErUNEP ¨ Sonderbd. Degens, E.T., Kempe, S., Richey, J.E., 1991. Biogeochemistry of Major World Rivers. SCOPE 42, Wiley, Toronto. Deuser, W.G., Degens, E.T., 1967. Carbon isotope fractionation in ?aqueous.. Nathe system CO 2 ?gas. –CO 2 ?aqueous. –HCOy 3 ture 215, 1033–1035. Devol, A.H., Quay, P.D., Richey, J.E., Martinelli, L.A., 1987. The role of gas exchange in the inorganic carbon, oxygen and 222 Rn budgets of the Amazon River. Limnol. Oceanogr. 32, 235–248. Drever, J.I., 1988. The Geochemistry of Natural Waters. Prentice-Hall, Englewood Cliffs, NJ, pp. 48–75. Environment Canada ?Digital., 1993. In Envirodat Water Quality Database. Inland Waters Directorate, Water Quality Branch, Environment Canada, Ottawa, Canada. Environment Canada, Inland Waters and Lands Directorate, 1988. Acid Rain: A National Sensitivity Assessment. Environmental Fact Sheet 88-1, Ottawa, Canada. Environment Canada, Lands Directorate, Environmental Conservation Service, 1985. Sensitivity to Acidification of Lakes in the Canadian Shield: Analysis and Interpretation of Geological and Ecological Data for 158 Watersheds in Southern Quebec. Working Paper no. 41, Ottawa, Canada. Faure, G., 1986. Principles of Isotope Geology. Wiley, Toronto, pp. 492–493. Flintrop, C., Hohlmann, B., Jasper, T., Korte, C., Podlaha, O.G., Scheele, S., Veizer, J., in press. Anatomy of pollution: rivers of North Rhine–Westphalia, Germany. Am. J. Sci. Fortin, D., Tasse, N., 1987. Burial dedolimitization within the Beekmantown Group, St. Lawrence Lowlands, Montreal area. GAC-MAC Annual Meeting Progress Abstracts 12, 42. Frape, S.K., Fritz, P., 1987. Geochemical trends for groundwaters from the Canadian shield. In: Fritz, P., Frape, S.K. ?Eds.., Saline Water and Gases in Crystalline Rocks. Geological Association of Canada, St. John’s, Newfoundland, pp. 19–38. Gao, W., Kempe, S., 1987. The Chanjiang: its long-term changes in p CO2 and carbonate mineral saturation. In: Degens, E.T. et al. ?Eds.., Transport of Carbon and Minerals in Major World Rivers, Part 4. Vol. 64. Mitt. Geol.-Palaont. Inst., Univ. ¨ Hamburg, pp. 207–216. Halas, S., Szaran, J., Niezgoda, H., 1997. Experimental determination of carbon isotope equilibrium fractionation between dissolved carbonate and carbon dioxide. Geochim. Cosmochim. Acta 61, 2691–2695.

86

K. Telmer, J. Veizer r Chemical Geology 159 (1999) 61–86 Richey, J.E., Devol, A.H., Wofsy, S.C., Victoria, R., Riberio, M.N.G., 1988. Biogenic gases and the oxidation and reduction of carbon in Amazon River and floodplain waters. Limnol. Oceanogr. 33, 551–561. Sanders, T.G., Ward, R.C., Loftis, J.C., Steele, T.D., Adrian, D.D., Yevjevich, V., 1983. Design of Networks for Monitoring Water Quality. Water Resources Publications, Littleton, CO. Schiff, S.L., Aravena, R., Trumbore, S.E., Dillon, P.J., 1990. Dissolved organic carbon cycling in forested watersheds: a carbon isotope approach. Water Resour. Res. 26, 2949–2957. Schindler, D.W., Newbury, R.W., Beatty, K.G., Campell, P., 1976. Natural water and chemical budgets for a small Precambrian lake basin in central Canada. J. Fish. Res. 33, 2526–2543. Schlesinger, W.H., 1991. Biogeochemistry, An Analysis of Global Change. Academic Press, Toronto, p. 309. Skirrow, G., 1975. The dissolved gases-carbon dioxide. In: Riley, J.P., Skirrow, G. ?Eds.., Chemical Oceanography, Vol. 2. Academic Press, pp. 1–192. Stallard, R.F., 1980. Major Element Geochemistry of the Amazon River System. Ph.D. thesis. MITrWHOI, Boston. Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry, Chemical Equilibria and Rates in Natural Waters. Wiley, Toronto, pp. 148–251. Tan, F.C., 1989. Stable carbon isotopes in dissolved inorganic carbon in marine and estuarine environment. In: Fritz, P., Fontes, J.C. ?Eds.., Handbook of Environmental Isotope Geochemistry, Vol. 3. Elsevier, pp. 171–190. Taylor, B.R., Hamilton, H.R., 1994. Comparison of methods for determination of total solutes in flowing waters. J. Hydrol. 154, 291–300. Taylor, T.R., Sibley, D.F., 1986. Ferroan dolomite in the Trenton Formation, Ordovician Michigan Basin. Sedimentology 33, 61–86. Telmer, K.H., 1997. Biogeochemistry and Water Balance of the Ottawa River Basin. Ph.D. thesis. University of Ottawa, Ottawa, Canada, pp. 55–74. Toggweiler, J.R., 1995. Anthropogenic CO 2 : the natural carbon cycle reclaims centre stage. Rev. Geophys., 1249–1252 ?supplement.. Truesdell, A.H., Jones, B.F., 1974. WATEQ: a computer program for calculating chemical equilibria of natural waters. J. Res., U.S. Geol. Surv. 2, 233–248. Walker, J.C.W., Hays, P.B., 1981. A negative feedback mechanism for the long-term stabilization of earth’s surface temperature. J. Geophys. Res. 86 ?C10., 9776–9782. Yang, C., Telmer, K., Veizer, J., 1996. Chemical dynamics of the ‘St. Lawrence’ Riverine system: d DH 2 O, d18 O H 2 O , d13 C DIC , d34 S Sulfate , and dissolved 87 Srr86 Sr. Geochim. Cosmochim. Acta 60, 851–866.

Hem, J.D., 1985. Study and Interpretation of the Chemical Characteristics of Natural Water. 2254. US Geological Survey, Alexandria, VA. Hem, J.D., Demayo, A., Smith, R.A., 1990. Hydrogeochemistry of rivers and lakes. In: Wolman, M.G., Riggs, H.C. ?Eds.., Surface Water Hydrology, The Geology of North America, Vol. O-1. Geological Society of America, Boulder, CO, pp. 189–231. Holley, E.H., 1977. Oxygen transfer at the air–water interface. In: Proc. Symp. on Transp. Processes in the Ocean at the 82nd Nat. Meet. of the AICE. Plenum, pp. 117–150. Kaszycki, C.A., Shilts, W.W., 1987. Glacial erosion of the Canadian Shield. In: Graf, W.L. ?Ed.., Geomorphic Systems of North America, Centennial Special. Vol. 2. Geological Society of America, Boulder, CO, pp. 155–159. Kempe, S., 1982. Long-term records of CO 2 pressure fluctuations in fresh water. In: Degens, E.T. ?Ed.., Transport of Carbon and Minerals in Major World Rivers, Part 1, Vol. 52. Mitt. Geol-Palaont. Inst., Univ. Hamburg, Hamburg, pp. 91–332. ¨ Ludwig, W., Amiotte, S.P., Probst, J.L., 1996. River discharges of carbon to the world’s oceans; determining local inputs of alkalinity and of dissolved and particulate organic carbon. Comptes Rendus de l’Academie des Sciences, Serie II. Sciences de la Terre et des Planetes 323, 1007–1014. Merritt, W.F., 1964. Studies of Dilution in the Ottawa River Using Rhodamine B. Atomic Energy of Canada ?AECL. Research, Chalk River. Middleton, K., Coniglio, M., Frape, S.K., 1990. Burial dedolimitization, Middle Ordovician carbonates, southwestern Ontario. AAPG Bull. 74, 1308. Mook, W.G., 1970. Stable carbon and oxygen isotopes of natural water in the Netherlands. Isotopic Hydrology, IAEA, pp. 163–190. Natural Resources Canada S.T.F., 1994. Dilution Downstream of the Deep River Candidate Area, STF Tech. Bib. No. 385. Ottawa, Canada. Occhietti, 1989. Ontario Water Resources Commission and the Quebec Water Board, 1971. Ottawa River Basin, Water Quality and Its Control in the Ottawa River. Paper no. 1. Environment Canada, Water Quality Branch, Ottawa. Ottawa River Engineering Board, 1965. In: Report on Hydrology and Regulation of the Ottawa River. Report on Hydrology and Regulation of the Ottawa River, Ottawa. Pawellek, F., Veizer, J., 1994. Carbon cycle in the upper Danube and its tributaries: d13 C DIC constraints. Israel J. Earth Sci. 43, 187–194. Reardon, E.J., Allison, G.B., Fritz, P., 1979. Seasonal chemical and isotopic variations of soil CO 2 at Trout Creek, Ontario. J. Hydrol. 43, 355–371.


相关文章:
CO2 Capture and Storage - What it has to offer Gunter_图文_....pdf
Weathering Fossilisation Lim Re sp e g n i ...Carbon Intensity Adapted from Kaya, 1995 Carbon ...Basin CO2 Sinks CO2 Sinks/CO2 Sources Aquifers (...
yi森林生态学-5-CNS循环_图文.ppt
N、P Flux 通量 Reservoirs 存 ? Living ...Carbon with mainly gaseous Cycles Atmosphere CO2 ...Weathering ? Ca(PO4)3OH + H2CO3 > Ca + ...
生物学_图文.ppt
CO2 and global warming Impacts of global warming ...N+P co-limiting in Summer; In Northern Gulf ...increasing number of river basins in the state. ...
...CO2 consumption in the Congo and Amazon river basins.unkown
//www.researchgate.net/publication/42582755 Carbon river fluxes and weathering CO2 consumption in the Congo and Amazon river basins ARTICLE in APPLIED ...
Feasibility of CO2 Fixation via Artificial Rock Weathering_....pdf
weathering can potentially fix a large amount of ...carbon mass on earth.3 Therefore, storing CO2 ...P.; Joyce, E. L., Jr.; Sharp, D. H. ...
同位素地球化学38.pdf
e P Carboniferous apparent in IP Figure 38.3. ... the flux of carbon due to weathering of ...CO2 and burial as carbonate (Fws) is expressed ...
...CO2NET lectures 24-4-2006-3 Storage, risk assessment and ....ppt
CO2NET Lectures on Carbon Capture and Storage 1....basins Formation of fossil fuels Radioactive waste...Originates from mechanical and chemical weathering ...
A new strategy to conserve Acacia tortilis subsp. raddianain ....unkown
of Applied Chemistry and Microbiology, P.O. ...carbon (C) assimilation activity in a temporal ...CO2 fluxes, biomass, colonization and hyphal ...
...atmospheric/soil CO2 consumed by major world river basins.unkown
weathering and atmospheric/soil CO2 consumed by major world river basins J....minerals to produce dissolved inorganic carbon (Garrels and Mackenzie, 1971; ...
...RATE CONTROL PARAMETERS IN A HUMID TROPICAL RIVER BASIN, ....unkown
CHAPTER 3 SILICATE WEATHERING, CO2 CONSUMPTION AND ITS RATE CONTROL PARAMETERS IN A HUMID TROPICAL RIVER BASIN, SOUTHWESTERN INDIA Chapter 3 Abstract: The ...
...weathering and CO2 consumption in the Xijiang River basin,....unkown
c o m / l o c a t e / g e o m o r p h Chemical weathering and CO2 consumption in the Xijiang River basin, South China Quanzhou Gao a,b...
Silicate rock weathering and atmospheric/soil CO2 uptake in ....unkown
Silicate rock weathering and atmospheric/soil CO2 uptake in the Amazon basin estimated from river water geochemistry: seasonal and spatial variations Jefferson ...
...and past high CO2 atmospheres enhance mineral weathering ....unkown
CO2 atmospheres enhance mineral weathering through ...belowground carbon-energy fluxes Joe Quirk*, ...Northern Ireland (described in Quirk et al. 2012...
..., andimplications for long-term changes in atmospheric pCO2.unkown
weathering rates would lower atmospheric pCO2 by 7...CO2, with a temperature dependency of $0.1 per...River drainage after 2.3 Ma and discharged light...
...weathering rates and CO2 consumption in Yangtze River basin.unkown
of chemical weathering rates and CO2 consumption in Yangtze River basin S.L...Rivers play an important role in long-term carbon cycle by transporting ...
Program with abstracts.unkown
river systems, both large and small, involving weathering, erosion, ... and glacio-eustatic sea-level changes associated with Northern Hemisphere ...
...EELGRASS BEDS ON THE WATER CHEMISTRY OF PORT GAMBLE, PUGET....unkown
eelgrass (Zostera marina) beds in order to utilize them as a carbon sink...Given that eelgrass takes up CO2 through photosynthesis, we hypothesized that...
...to Contrasting Environmentsin Baffin Bay and the Northern ....unkown
CO2-rich bottom watersrelated to the Baffin Bay...northern Labrador Sea, contains the largest fora... oxygenated deep waters; and high carbonflux. ...
1. PALEOCEANOGRAPHY OF THE EASTERN EQUATORIAL PACIFIC DURING ....unkown
(NADW), the rapid uplift of the Himalayas, the major intensification of Northern Hemisphere glaciation), and the response of the ocean before and after ...
Effect of Crude Oil and Chemical Additives on Metabolic ....unkown
The importance of soil organic N and P mineralization in pro- viding labile nutrients to plants is well established [67], and the only nutrients ...
更多相关标签: