Geochimica et Cosmochimica Acta, Vol. 62, No. 12, pp. 2053–2075, 1998 Copyright © 1998 Elsevier Science Ltd Pergamon Printed in the USA. All rights reserved 0016-7037/98 $19.00 1 .00 PII S0016-7037(98)00127-6 The fluvial geochemistry of the rivers of Eastern Siberia: II. Tributaries of the Lena, Omoloy, Yana, Indigirka, Kolyma, and Anadyr draining the collisional/accretionary zone of the Verkhoyansk and Cherskiy ranges 1,2 1,2 3 3 1 YOUNGSOOK HUH, GERA PANTELEYEV, DMITRY BABICH, ALEXANDR ZAITSEV, and JOHN M. EDMOND 1Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA 2MIT/WHOI Joint Program in Oceanography/Applied Ocean Science and Engineering, Cambridge, Massachusetts 02139, USA 3Department of Geography, Laboratory of Erosion and Fluvial Processes, Moscow State University, Moscow, Russia (Received August 11, 1997; accepted in revised form March 5, 1998) Abstract—Fundamental to the global carbon cycle over geologic time scales is the control of atmospheric CO2 by aluminosilicate weathering. Much of the information on the rates of this process comes from rivers in the tropics and subtropics. To understand the possible climatic influences systematic studies are needed for the arctic/subarctic regions. This is the second in a series of papers addressing this problem by systematic studies of the pristine rivers of the Russian Far East. The region to the east of the Siberian Platform (Huh et al., 1998) is a geologically complex terrain formed by the Mesozoic collision and accretion of the Siberian and Kolyma plates. Because of the arid continental climate, it has not been glaciated in the recent past. Thus, it is possible to study weathering processes in an arctic environment dominated solely by cryogenic interactions without contamination by heterogeneous components derived from scouring glaciers. All the major rivers and their tributaries in this area have been sampled on expeditions to individual basins (;100 samples) on a reconnaissance basis at falling stage, usually in July and August. The total dissolved cation levels (TZ1) are moderate (up to ;3,100 mEq), and the major ion chemistry is indicative of Ca-aluminosilicate and carbonate weathering with significant contributions from black shales in some tributaries. The Si/TZ1*, Si/(Na* 1 K) and 87Sr/86Sr ratios indicate that the weathering is superficial, i.e., only to cation-rich secondary minerals. The areal total dissolved solid fluxes range from 0.04 to 0.39 3 106 mol/km2/yr, up to an order of magnitude lower than for the Amazon-Orinoco draining the Andes 3 6 2 in the tropics (0.6–4.1 10 mol/km /yr). The CO2 consumption by aluminosilicate weathering (18–230 3 103 mol/km2/yr) is also at the lower end of the range observed in the Amazon-Orinoco headwaters (143–1,000 3 103 mol/km2/yr). However, as the North American counterparts in similar latitudes and with comparable relief, the Mackenzie, Yukon, and Fraser draining the Rockies, also have high dissolved solids 3 6 2 3 3 2 (0.2–2.9 10 mol/km /yr) and CO2 (19–1,750 10 mol/km /yr) fluxes, these low values seem to be more a function of lithology than simply climate. Ice action in cold environments appears to overcome the inhibiting effects of the decreased temperatures and lack of precipitation in producing a high chemical yield but results in superficial weathering in the case of aluminosilicates. Copyright © 1998 Elsevier Science Ltd 1. INTRODUCTION that the global warming effects of carbon dioxide and other greenhouse gases will be amplified two to four times in the The possible responses of the continental weathering regime to arctic relative to lower latitudes. How fast and how far the climatic and tectonic variation on a variety of time scales is weathering regime will respond are important questions. The drawing much interest (Froelich et al., 1992; Raymo and Rud- fluvial output to the shelf seas also affects the density stratifi- diman, 1992; Blum and Erel, 1995). While global deep-sea cation of the Arctic Ocean, which can have effects extending to hydrothermal and subduction-related fluid inputs to the ocean the whole ocean through deep-water formation mechanisms are probably smoothly varying over time (Gaffin, 1987; Kast- (Aagaard and Carmack, 1989). Possible changes in the fluxes ner et al., 1991) and not directly influenced by climatic changes, the fluvial fluxes could be quite variable. In order to of nutrients like silica, nitrogen, and phosphorus could affect understand the link between these and climate, it is useful to the productivity and, therefore, the ecosystem in the shelf seas. study the current weathering patterns in high latitudes relative At present major signals are being imposed on the geochemi- to those in the tropics as a surrogate for climatic deterioration. cal cycle by the active orogenies in the continental arc of the The very low temperatures and the cryogenic processes result- western Americas and the Tethyan collision zone (Sarin et al., ing from them make the arctic/subarctic a unique environment. 1989; Edmond, 1992; Krishnaswami et al., 1992; Edmond et The weathering fluxes and patterns might be expected to be al., 1996). The effect of climate per se is more problematic. In quite distinct from those in the relatively well studied hot and global carbon cycle models, Arrhenius-type temperature-de- humid tropics. Unusual magnitudes in the fluxes of certain pendent weathering kinetics are central to the negative feed- elements or elemental and isotopic ratios associated with arctic back mechanism of CO2 over geologic time scales, presumed to fluvial weathering potentially could be used in interpreting the be mediated by increased water vapor transport and precipita- sedimentary record in terms of environmental change. tion (Berner et al., 1983; Berner, 1994). The temperature feed- The arctic is a climatically sensitive region (Hansen et al., back function of Berner (1994) is doubled if the temperature is 1983; Cuffey et al., 1995). General Circulation Models predict 6°C higher than present and halved if the temperature is 6°C 2053 2054 Y. Huh et al. Fig. 1. Schematic map of the rivers of Eastern Siberia which have been sampled as part of this project. Lake Baikal occupies the southwestern corner, Kamchatka and Alaska the east. Total drainage area investigated is ;4 3 106 km2, half the size of the U.S. (;9 3 106 km2). Dashed lines indicate watershed boundaries. lower. This very large predicted variation can be tested with et al., 1972; Wadleigh et al., 1985; Newton et al., 1987). The contemporary data on fluvial chemistry and fluxes as a function data reported for the Lena delta gives a monitor of the flux of of climate. the major elements from the entire basin to the Arctic Ocean In order to constrain the magnitude of the changes in weath- using information based on monthly samples. ering rates, and hence CO2 consumption over time, and to study The effect of climate on riverine dissolved loads has been the effects of climatic and tectonic changes on the weathering addressed previously in some reconnaissance and budget stud- yields from various geological settings, studies of pristine sys- ies. Peters (1984) surveyed the available data for U.S. rivers but tems from a range of geologic, topographic, and climatologic did not find any conclusive evidence for an effect of tempera- environments are needed. Various studies using the mass bal- ture on weathering. He attributed this to enhanced physical and ance of major ions and the strontium isotope systematics have chemical weathering due to frost cracking and wedging in cold shown the close link between the geology of the drainage environments outweighing the effects of warm, humid pro- basins and the dissolved load and chemical composition of the cesses in the tropics. Bluth and Kump (1994) looked at small rivers draining them—the Amazon (Gibbs, 1972; Stallard and rivers draining the basaltic terrains of Iceland, the Colombia Edmond, 1983, 1987), the Orinoco (Lewis and Weibezahn, Plateau, and Hawaii and found comparable weathering rates, as 1981; Edmond et al., 1995, 1996), the Yangtze and the represented by silica and bicarbonate fluxes. On the other hand, Huanghe (Hu et al., 1982; Zhang et al., 1990), the Indus (Pande Meybeck (1986) examined French streams draining monolitho- et al., 1994), the Ganga-Brahmaputra (Sarin et al., 1989; logic watersheds and found a positive relationship with tem- Palmer and Edmond, 1992), the Mekong (Carbonnel and Mey- perature. Drever and Zobrist (1992) showed that the concen- beck, 1975), and the Congo (Ne´grel et al., 1993). However, trations of the major cations, silica, and alkalinity decreased these studies have been concentrated in the tropical/temperate with elevation (decreasing temperature) within a single alpine zones. The only comparable work reported for high latitudes is watershed but that the weathering stoichiometry remained con- on the Mackenzie (Reeder et al., 1972), the Fraser (Cameron et stant. In a comparison of two drainages at two different eleva- al., 1995), the Lena delta outlets (Gordeev and Sidorov, 1993), tions and temperatures (11.7 vs. 10.6°C), Velbel (1993) ob- the tributaries of the Siberian Platform (Huh et al., 1998), and served higher dissolution rates of feldspars at greater the rivers of Iceland and Sweden (Gı´slason et al., 1996; temperatures and calculated the activation energy to be higher Andersson et al., 1994). For basins that have been glaciated in than previous estimates suggesting a larger weathering-medi- the recent past, like the Mackenzie, it is difficult to distinguish ated negative feedback between global temperature and atmo- between the weathering of in situ basement rocks from that of spheric CO2. However the database remains very small. the transported and disaggregated glacial overburden (Reeder New data from large basins are needed in other lithologic Siberian river water chemistry 2055 Fig.