Glacial chemical weathering

The principal reactions that comminuted bedrock undergo in glaciated terrain are summarized below. We assume that the bedrock is primarily composed of silicates and aluminosilicates.

Glacial comminution crushes bedrock and exposes the trace reactive components within crystal aggregates more rapidly than would be the case in temperate and tropical soils, where new minerals are ultimately accessed via solubilization of aluminosilicate lattices. Hence, glaciers are effective at promoting the solubiliza-tion of trace reactive components in the bedrock, which include carbonates, sulphides and fluid inclusions. Laboratory experiments and direct sampling of waters from the glacier bed (Tranter et al., 1997, 2002b) show that the initial reactions to occur when dilute snow and ice melt first access glacial flour are carbonate and silicate hydrolysis (Equations 1 & 2). These reactions raise the pH to high values (>9), lower the PCO2 (to ca. 10-6 atms) and maximize the water's potential to adsorb CO2. Carbonate hydrolysis produces a solution with a Ca2+ concentration of ca. 200|leqL-1, with HCO3- the dominant anion.

Ca1_%(MgJCO3(s) + H2O(l) ^ (1 - x) Ca2+(aq) + xMg2+(aq) calcite + HCO3- (aq) + OH-(aq) (1)

KAlSi3O8(s) + H2OQ) ^ HAlSi3O8 + K+(aq) + OH-(aq) (2) K-feldspar weathered feldspar surface

The relatively dilute meltwater in contact with fine-grained glacial flour promotes the exchange of divalent ions from solution for monovalent ions on surface exchange sites. Hence, some of the Ca2+ and Mg2+ released from carbonate and silicate hydrolysis is exchanged for Na+ and K+.

glacial flour - Na(2-z),Kz (s) + (1 - x) Ca2+ (aq) + xMg2+ (aq) ^ glacial flour - Ca(1-x), Mgx(s) + (2 - z) Na+(aq)

The high pH derived from hydrolysis enhances the dissolution of aluminosilicate lattices, as Al and Si become more soluble at pH > 9. Hydrolysis of carbonates results in a solution that is near

Table 14.1 The concentration of major ions in glacial runoff from different regions of the world (after Brown, 2002). Concentrations are reported in |ieqL-1

Region

S+

Ca2+

Mg2+

Na+

K+

HCO3-

SO42-

Cl+

Source*

Canadian High

280-

3500

260-2600

21-640

1-190

0.1-39

210-690

59-3900

1

Arctic

Antarctica

550-

3100

72-1300

120-336

360-1400

0.8-110

91-1600

34-1200

0.6-

1000

2

Svalbard

330-

1900

120-1000

99-540

110-270

5.1—41

110-940

96-760

5-

310

3-5

Canadian Rockies

1300-

1500

960-1100

290-310

3.7-36

5.8-9.2

890-920

380-520

1.7-

25

6

Iceland

170-

960

110-350

30-120

30-480

2.8-12

190-570

26-130

30-

87

7,8

Himalayas

130-

940

75-590

6.6-230

25-65

22-51

200-730

160^10

1-

22

9

Norway

20-

930

8.8-623

1.6-66

8.3-210

1.0-29

1.4-680

7-140

0.9-

190

10

European Alps

37-

910

20-640

6-140

4.9-92

5.9-33

11^00

10-240

0.9-

92

11-13

Alaska

670

550

36

25

61

430

260

2

14

Greenland

280-

387

130-170

68-98

78-110

5-9

220-340

90-200

16-

30

15

Cascades

56-

150

35-80

8.3-20

2.5-17

9.7-37

83-100

7.9-29

16

Global mean runoff

1200

670

280

220

33

850

170

160

17

*1, Skidmore & Sharp (1999); 2, De Mora et al. (1994); 3, Hodgkins et al. (1997); 4, Hodson et al. (2000); 5,Wadham et al. (1997); 6, Sharp et al. (2002); 7, Raiswell & Thomas (1984); 8, SigurQur Steinkorsson & Oskarsson (1983); 9, Hasnain et al. (1989); 10, Brown (2002); 11, Brown et al. (1993); 12, Collins (1979); 13, Thomas & Raiswell (1984); 14, Anderson et al. (2000); 15, Rasch et al, 2000; 16, Axtman & Stallard, 1995; 17, Livingstone (1963).

*1, Skidmore & Sharp (1999); 2, De Mora et al. (1994); 3, Hodgkins et al. (1997); 4, Hodson et al. (2000); 5,Wadham et al. (1997); 6, Sharp et al. (2002); 7, Raiswell & Thomas (1984); 8, SigurQur Steinkorsson & Oskarsson (1983); 9, Hasnain et al. (1989); 10, Brown (2002); 11, Brown et al. (1993); 12, Collins (1979); 13, Thomas & Raiswell (1984); 14, Anderson et al. (2000); 15, Rasch et al, 2000; 16, Axtman & Stallard, 1995; 17, Livingstone (1963).

saturation with calcite and aragonite. It is only in these types of waters that aluminosilicate dissolution is greater than carbonate dissolution. The influx of gases (including CO2 and O2), either from the atmosphere or from basal ice, and CO2 produced by microbial respiration (see below) both lowers the pH and the saturation with respect to carbonates. In addition, sulphide oxidation produces acidity (see below). Hence, almost all subglacial meltwaters are undersaturated with respect to calcium carbonate. The rapid dissolution kinetics of carbonates with respect to silicates means that carbonate dissolution continues to have a large impact on meltwater chemistry, despite carbonates being present often in only trace concentrations in the bedrock. For example, Haut Glacier d'Arolla has a bedrock which is composed of meta-morphic silicate rocks. Carbonates and sulphides are present in trace quantities in bedrock samples (0.00-0.58% and <0.005-0.71% respectively). There are also occasional carbonate veins present in the schistose granite. Despite the bedrock being dominated by silicates, sulphide oxidation in subglacial environments dissolves carbonate to silicate in a ratio of ca. 5:1 (Tranter et al., 2002b), compared with the global average of ca. 1.3:1 (Holland, 1978).

The acid hydrolysis of silicates and carbonates (Equations 4 & 5) that arises from the dissociation of CO2 in solution is known as carbonation. Carbonation occurs in a restricted number of subglacial environments because ingress of atmospheric gases to these water-filled environments is restricted. It largely occurs in the major arterial channels at low flow, particularly near the terminus, and at the bottom of crevasses and moulins that reach the bed. Fine-grained sediment is flushed rapidly from these environments, and there is little time for the formation of secondary weathering products, such as clays. Hence, silicates dissolve incongruently, as crudely represented by Equation 4.

CaAl2Si2O8(s) + 2CO2(aq) + 2H2O (l) ^ Ca2+(aq) anorthite + 2HCO3-(aq) + H2Al2Si2O8(s) (4)

weathered feldspar surfaces

Ca1_%(MgJCO3(s) + CO2(aq) + H2O (l) ^ (1 - x)Ca2+(aq) calcite + xMg2+(aq) + 2HCO3(aq) (5)

There is a limited body of evidence which suggests that microbial oxidation of bedrock kerogen occurs (Wadham et al., 2004), and if this is the case, carbonation as a consequence of microbial respiration may occur in debris-rich environments, such as in the distributed drainage system and the channel marginal zone.

The dominant reaction in subglacial environments is sulphide oxidation, because, following hydrolysis, this is the major reaction which provides protons to solution, so lowering the pH, decreasing the saturation index of carbonates, so allowing more carbonate dissolution (Equation 7). Sulphide oxidation occurs predominantly in debris-rich environments where comminuted bedrock is first in contact with water. It is microbially mediated, occurring several orders of magnitude faster than in sterile systems (Sharp et al., 1999). It consumes oxygen, driving down the pO2 of the water.

4FeS2(s) + 16Ca1-x(MgJCO3(s) + 15O2(aq) + 14H2O(l) ^ (7) pyrite

16(1 - x) Ca2+(aq) + 16xMg2+(aq) + 16HCO3-(aq) + 8SO42-(aq) + 4Fe(OH)3(s)

ferric oxyhydroxides

Earlier studies suggested that the limit on sulphide oxidation was the oxygen content of supraglacial melt, because subglacial supplies of oxygen are limited to that released from bubbles in the ice during regelation, the process of basal ice melting and refreezing as it flows around bedrock obstacles. Studies of water samples from boreholes drilled to the glacier bed, however, show that the SO42- concentrations may be two or three times that allowed by the oxygen content of supraglacial meltwaters (Tranter et al.,

2002b). This suggests that oxidizing agents other than oxygen are present at the glacier bed. It seems very likely that microbially mediated sulphide oxidation drives certain sectors of the bed towards anoxia, and that in these anoxic conditions, Fe(III), rather than O2, is used as an oxidizing agent (Equation 8). Sources of Fe(III) include the products of the oxidation of pyrite and other Fe(II) silicates in a previous oxic environment, as well as that found in magnetite and haematite.

FeS2(s) + 14Fe3+(aq) + 8H2O(l) ^ 15Fe2+(aq) + 2SO42- (aq)

Support for anoxia within subglacial environments comes from the S18O-SO4, which is enriched in 16O when sulphide is oxidized in the absence of oxygen (Bottrell & Tranter, 2002).

The realization that there is microbial mediation of certain chemical weathering reactions in subglacial environments (Sharp et al., 1999; Skidmore et al., 2000, Bottrell & Tranter, 2002) has resulted in a paradigm shift, as the types of reactions that may occur in anoxic sectors of the bed include the common redox reactions that occur, for instance, in lake or marine sediments (Drever, 1988). A key difference in glacial systems is that the supply of new or recent organic matter is limited to that inwashed from the glacier surface, such as algae, insects and animal faeces, or overridden soils during glacier advance. By contrast, the supply of old organic matter from comminuted rocks is plentiful. Given the thermodynamic instability of organic matter in the presence of O2 or SO42-, it seems likely that microbes will have evolved to colonize subglacial environments and utilize kerogen as an energy source. The first data to support this assertion is stable isotope analysis from Finsterwalderbreen, a small polythermal-based glacier on Svalbard that has shale as a significant component of its bedrock (Wadham et al., 2004). The S18O-SO4 of waters upwelling from subglacial sediments are very enriched in S18O; the S34S is enriched in 34S, which suggests that cyclical sulphate reduction and oxidation has been occurring. The S13C of DIC (dissolved inorganic carbon) is negative, consistent with the assertion that organic matter has been oxidized. Mass balance calculations suggest that a possible source of organic matter is kerogen, but the necromass of dead bacteria cannot be discounted. Whatever is the source of organic matter, sectors of the bed at Finster-walderbreen are so anoxic that sulphate reduction is occurring (Equation 9).

organic carbon

It is possible that methanogenesis occurs under certain ice masses, because methanogens have been isolated from subglacial debris (Skidmore et al., 2000). The low S13C-CH4 and high concentration of methane found in gas bubbles within the basal ice of the Greenland Ice Sheet are consistent with there being methanogen-esis within the basal organic-rich palaeosols.

The colonization of subglacial environments by microbes suggests that both energy and nutrient sources are readily available. Energy sources, such as sulphides and kerogen, have been discussed above. Comminuted bedrock may also provide a source of nutrient. Average crustal rock contains 1050 ppm of P. Typically, this is contained in sparingly soluble minerals such as apatite, and calcium, aluminium and ferrous phosphates (O'Neill, 1985). Comminuted bedrock and basal debris provides a renewable source of P on mineral surfaces, and it is likely that uptake of P by microbes maximizes the extraction of P from these activated surfaces. Hodson et al.(2004) suggest that 1-23 |lgPg-1 is present as readily extractable P on the surface of glacial flour. Sources of N also may be derived from comminuted rock (Holloway & Dahlgren, 2002). The N content of rocks is typically 20 ppm (Krauskopf, 1967), but may exceed 1000ppm in some sedimentary and metasedimentary rock (Holloway & Dahlgren, 2002). For example, bedrock has been shown to be a source of NH4+ from schists in the Sierra Nevadas, California (Holloway et al., 1998), and there may be appreciable concentrations of NH4+, which substitutes for K+, in biotite, muscovite, K-feldspar and plagioclase (Mingram & Brauer, 2001). It follows that glacial comminution of bedrock and basal debris maximizes the likelihood that N-pro-ducing surfaces are exposed to meltwaters and microbes, and, given that bedrock in the Sierra Nevadas can act as an N source, it is likely that comminuted glacial debris is also a potential source of N.

The predominance of carbonate hydrolysis, carbonation and sulphide oxidation in subglacial weathering reactions on alumi-nosilicate/silicate bedrock is also found on carbonate bedrock. The balance between carbonate dissolution and sulphide oxidation, however, depends on the spatial distribution of sulphides in the bedrock and basal debris (Fairchild et al., 1999). Non-congruent dissolution of Sr and Mg from carbonate is also observed in high rock:water weathering environments, such as the distributed drainage system, in which water flow is also low (Fairchild et al., 1999).

To date, there are few studies of glacial chemical weathering on bedrock with a significant evaporitic content. Work at John Evans Glacier in the Canadian High Arctic has shown that gypsum is dissolved in some areas of the bed, and that mixing of relatively concentrated Ca2+-SO42- waters with more dilute Ca2+-HCO3--SO42- waters results in CaCO3 precipitation owing to the common ion effect (M.L. Skidmore, personal communication, 2002). Kennicott Glacier, Alaska, is underlain by a sabkha facies limestone, which contains trace quantities of halite. Waters accessing sites of active erosion readily acquire Na+ and Cl- (Anderson et al., 2003).

A key feature of the above chemical weathering scenarios is that relatively little atmospheric or biogenic CO2 is involved. Hence, whereas ca. 23% and ca. 77% of solutes, excluding recycled sea salt, found in global mean river water is derived from the atmosphere and rock respectively (Holland, 1978), atmospheric sources account for a maximum of 3-11% of solute in glacial runoff (after Hodson et al., 2000).

Was this article helpful?

0 0

Post a comment