Annals qfGlaciology 27 1998 © International Glaciological Society

Inorganic carbon-isotope distribution and budget in the Lake Hoare and Lake Fryxell basins, Taylor Valley, Antarctica

KLAUS N EUMANN, I W. BERRY LYONS,I D AVID J. D ES M AR A IS2 1 University qf Alabama, De/Jart71lent qfGeology, nlscaloosa, A L 35487, US.A. 2 NASA Ames Research Centel; J1Ioffitt Field, CA 94035, US.A .

ABSTRACT. One of the unusua l features of Lakes Fryxell and Hoare in Taylor Vall ey, so uthern Victori a Land, Antarcti ca, is their perennial ice cover. This ice cover limits gas excha nge between the atmosphere and the la ke water, and causes a ve ry stabl e stratifica­ tion of the lakes. We ~ n a l yzed a series of water samples from profil es of these la kes and their tributaries for 813 C of the dissolved inorganic carbon (DIC) in order to qualify the carbon flu x fr om the streams into the lakes, and to i\1Vestigate the carbon cycling within the lakes. Isotopic values in the uppermost waters (813C = + 1.3%0 to 5.3%0 in Lake H oare, +0.4%0 to +3.0%0 in Lake Fryxell ) are close to the carbon-isotope values encountered in the streams feeding Lake Fryxell , but distinctively heavier than in streams feeding Lake H oare (813C = - 2.3%0 to 1.4%0). T hese ratios a re much heavier than ratios found in the moat that forms around the lakes inJanuary- February (813C = - 10.1 %0). In the oxic photic zones of the lakes, photosynthesis clearly influences the isotopic c

INTRODUCTION more moderate cl imates (Doran a nd others, 1994). Because they a re located in the d ri es t desert of the Ea rth, and are Closed-basin lakes are excellent indicators of changes in recha rged onl y by glacial meltwater, their water levels are cl imate. Due to their nature, water levels can vary dramati­ ,·ery sensitive to climate changes (Clow and others, 1988; cally, as precipitation and evaporati on change over time. Evi­ Whar to n and others, 1992). The amount of recharge can dence of these changes can be fo und in the form of perched or vary d rastically from year to year, a nd is controll ed by the fl ooded shorelines (Gilbert, 1890; J ones and others, 197 1; Cut­ temperature (Chinn, 1985). These lakes are also covered fi eld, 1974), and salt deposits in lake sediments (H ardie and with a perennial ice cover. This "lid" a ffects the lakes by re­ others, 1978; Eugs ter, 1980). Changes in lake level and water ducing gas exchange, excluding wind mixing and abso rbing volume ofte n correspond wi th changes in the water chemi s­ 97- 99% of the incoming li ght (Vincent, 1987; Lizotte and tr y. During water-level ri ses, solutes are diluted, while during Prisc u, 1992; Wharton and others, 1992). The balance water depletion solu tes can be concentrated. Minerals, such between ice-cover gain (by fr eezing of water to the bottom as calcite, can be precipitated if their saturati ons are reached. of the ice) and loss (via ablati on from the surface) itself is T hese chemical changes, in turn, affect the bi ology of the affected by climate variation, and as a res ult the thickness lakes, as hi gher nutrient concentrations cause higher produc­ changes over time (\Vharton and others, 1992). This in turn ti vit y, and hi gher solutes generall y limit the number of spe­ directly influences the transmissivity, and the productivity cies present in a lake (Melack, 1983). Bi ological acti vity in in these partia ll y light-limited lake systems. these types of la kes, as anywhere on Earth, is closely con­ nected to the carbon cycle. Photosy nthesis and respi ration de­ SITE DESCRIPTION plete and enrich inorganic carbon in lakes. T hese processes can change the stable carbon isotopic composition (813 C) of Lakes Hoare and Fryxell are located in Taylor Valley, so uth­ the lake water. Shifts in rates of productivity and mineraliza­ ern Victoria La nd, Anta rctica (Fig. 1). In 1993 a U. S. ti on can cause the organic contents oflake sediments to vary. National Science Foundation- (NSF-)supported long-term These changes in the quality (isotopic) and quantity (organi c ecological research (LTER ) site was established in Taylor vs inorgani c matter) of the carbon have been used widely to Va lley, a nd vari ous chemical, physical and biological pa­ deduce paleo-conditions from lake sedi ments (M cKenzie, rameters have been monito red by the LT ER team on a rou­ 1985). In order to inte rpret sedimentary records it is essential tine basis. The valley is located "-' 100 km northwest of to qualify and quantify present closed-basin lake systems. M cMurdo Stati on, at 76- 78° S, 160- 164° E. The vall ey is Closed-basin lakes in Antarctica respond like lakes in arid, with a precipitation of :::; 100 mm a 1 (Clow and others, Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 05:30:19, subject to the Cambridge Core terms of use. 685 Ne umann and others: Carbon-isotope distribution in Tay lor Valley lakes

METHODS

o 1 2 3 4 5 6 7 Kilometers Water samples were collected at the location of the greatest water depth. Holes were drilled a nd melted at the beginning of each season. The holes could be used during the whole season ( October-J anuary). Five-liter Niskin bottles were lowered through the holes, and samples collected 3- 4 times per season. pH was measured electrometrically (Beck man portable pH meters with vari ous silver- silver-chloride glass electrodes; accuracy ± 0.05) in the fi eld within hours aft er sample coll ection. Temperature was measured with a Sea­ Bird GTD probe. Dissolved inorganic carbon (DIG) samples from the lakes were stabilized with chloroform,

~ Streams. _ Watershed!lilllJ Lakes 0 Glaciers stored in cooler boxes at ,..,A°e. The samples were inj ected Creeks - Boundary into 6 N H 2S04, sparged with nitrogen gas and measured with an MSA Lira infrared gas analyzer (personal commu­ Fig. 1. M ap if the eastern part if Tay lor Valley, Antarctica. ni cation fr omJ e. Priscu, 1997), generally within 2 weeks of Glaciers are light gray, lakes da rk gray; white areas consist if sampling. For streams, alkalinity was determined by titra­ morainal material and bedrock. All streams feeding Lake ti on and Gran plot (Drever, 1988). The relative standard de­ Hoareflow along Canada and Suess Glaciers. The lake levels viations for the DIG and alkalinity analyses were ± 3%. are afew meters above sea level. T he Ross Sea is about 5 km to The sampl es for major cations/ani ons were filtered through the east. Whatman 0.4 /1m filters in the fi eld, and analyzed in M cMurdo Station by ion chromatography (Welch and 1988). The average annual temperature is less than - 20o G, others, 1996), using a Dionex DX-300 chromatograph. but temperatures ri se above freezing during a few days of Those analyses were fini shed within "-'2 months after the summer. No vascular plants grow in the valley. There sampling, a nd the error, expressed as percentage errors in are algae and lichens growing in the soil (Campbell and the charge balance between cati ons and anions, was 3.4% Cla ridge, 1987), but most pla nts are located in algal mats in for the streams and 3.0- 1.4% for the lakes. The speciation the lakes and the streams (Wharton and others, 1983). As a models PHREEQE and PHRqJ'ITZ (Parkhurst and res ult, terrestrial input of organic matter into the la kes is others, 1980; Plummer and others, 1988) were used to cal­ ve ry limited (M cKnight and others, 1991, 1993). Lake H oare culate CO2 concentrations in the water. Samples [or bl3 C is a fresh-water lake (total dissolved solids (TDS) = 300- a nalysis were filtered into pre-evacuated 60 ml serum bot­ 750 mg I I) of 30 m depth, that is oxygenated all the way to tles, using ""hatman 0.4 1£m GF-F filters, a nd stabilized with the bottom, except for some small anoxic pockets in its dee­ 0.2 ml concentrated mercuric chloride solution. They were pest parts. Lake Fryxell is 18 m deep, is brackish analyzed at the NASA Ames Research Genter on a modi­ (TDS = 300- 8000 mg I- I) and has a steep chemo- and oxy­ fi ed Nuclide 6-60RMS mass spectrometer (Hayes and cline at 9 m depth. others, 1977).

Table 1. Stream length ( Alger and others, 1997), annual water discharge, alkalinity concentrations and alkalinityfluxes, and bl3C valuesJor streams in Lake Fryxell and Lake Hoare basins

/993-94 /994-95 /995- 96 Length Discharge Alk. cone Alk·flux Discharge Alk. cone. Alk.flux Discharge Alk. cone. Alk.flux D1C- bI3C

3 - \ 1 :i 1 3 - I km III a m M mol a m a mM mol a 1 m a mM mol a- I

Fryxell Basin 824200 0.596 440670 152750 0.737 77430 403860 0.799 240590 - 1.68 Canada Stream 1.5 130910 0.1 35 17690 41710 0.209 8720 75350 0.200 15 100 - 2.27 Huey Creek 2. 1 51390 1.076 55310 1.170 Lost Seal Stream 192 110 0.571 109630 20630 0.808 16670 56480 0.489 27600 - 0.89 Aiken Creek 57050 0.982 56050 9710 0.935 9080 23530 1.094 25740 - 1.65 Von G ucrard Str. 4.9 68 140 0.801 54590 7220 1.093 7890 33270 1.11 2 370/0 1.1 8 Crescent Stream 38960 1. 253 48810 211 0 1.453 3070 17200 1.346 23 160 1.1 2 Delta Stream 11.2 63310 0.467 29590 3280 0.915 3000 29400 1. 249 36710 - 4.85 Green Creek 1. 2 60950 0.407 24790 29 120 0.380 11 070 64810 0.367 23770 - 4.55 M cKnight Creek 40580 0.630 25580 6800 0.967 6580 16470 0.990 16300 Ha rni sh Creek 52840 2500 1. 264 3160 24370 1.111 7840 Bm'Vles Creek 0.9 21970 0.530 11 640 10480 0.570 5980 23330 0.601 14020 M ariah Creek 26310 0.177 4670 12520 0.094 11 80 27600 0.348 9610 Andrews Creek 19680 0. 11 8 2320 6670 0. 154 1030 12050 0.310 3730 Hoare Basin 136710 0.417 58640 71350 0.480 34950 85670 0.439 34820 Anderson Creek 1.4 53780 0.440 23650 16050 0.475 7620 39 180 0.292 11 430 - 0.32 House Stream 2 21740 0.430 9350 14 180 0.400 5670 11 920 0.438 5220 Wharton Creek 37970 0.479 18 180 25520 0.544 13880 21460 0.562 12060 McKay C reek 23220 0.321 7460 15600 0.499 7780 13 110 0.466 6110

Note: 1995- 96 data in italics are estimates based on the behavior orthe other streams over the three seasons. Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 05:30:19, subject to the Cambridge Core terms of use. 686 Neumann and others: Ca rbon-isotope distribution in Tay lor Va lll'J,lakes

RESULTS depth and are lowest at the bottom, with values of - 3.02%0 and -2.64%0 in 1994 and 1995, respecti vely. Values below the chemocline are identical for the two years, but the DIC in Data for the DIC nux in the streams entering La kes FryxeIl the upper, oxic part of the lake is heavi er in 1995, with a and H oare are shown in Tabl e 1. The values for the 1993- 94 m aximum (j13C of +3.03%0. In 1994, the heaviest value season had to be calculated from the cha rge balance since observed was +0.41% 0. The data for Lake H oare are fr om no direct alkalinity measurements were made. The alka­ vVh a rton and others (1993) and were measured in 1987. The linity in streams of Fryxell basin is generall y higher than in DIC has its heaviest value in the upper waters Hoare basin; the averages are 0.7 and 0.4 mM, res pectively. (b13C = +5.3%0). To a depth of 10 m the values are con tant, Both water and carbon discharge varied greatly: for then they drop to -4.0%0 at 30 m depth. example, during the 1993- 94 season more than 800000 m3 of water and 440000 mol a l kalinity we re discharged into Lake Fryxell, whereas in the 1994- 95 season the a mounts DISCUSSION were 152000 m 3 and 77 400 mol alkalinity (i. e. a reduction Stream.s of > 80% ). As a res ult, the carbon input into the lakes is pri­ marily controlled by the vari ation in annual discharge. The In absolute term , streams put a large amount ofDIC into the changes in flow of the streams entering Lake H oare were lakes (Tabl e I). All the streams in both basins are undersatu­ much small er than those of the streams entering Lake Fryx­ rated with res pect to calcite. Calcite has been obse rved in ell. In the same two seasons, fl ow was reduced by onl y 50 %. soils (Campbell and Claridge, 1987), and a longer expos ure In the 1995- 96 sea on the influx of alkalinity into Lake of the water to calcite will result in a higher DIC load in the Fryxell tripled over the previ ous season, while it remained water. For example, the substantia ll y longer streams in the about the same in Lake Hoare basin. Lake Fryxell bas in generall y exceed Lake Hoare basin 13C values of some streams during the seaso n (j 1994- 95 streams in their DIC concentration. Alkalinity, which under are s hown in Tabl e 1. Isotopic compositions ranged from the pH conditions encountered in these streams is roughly - 4.85%0 to + 1.1 8%0. The streams vary considerably in length, equivalent to [HC03 ], varies by one order of magnitude amount of algal mats present, gradients and exposure to within Fryxell basin streams. The shortes t streams (Table I), direct radiation. The heaviest isotopic compositions occur Andrews Mariah and Green Creeks, have the lowest [DIC]. in the streams with the highest alkalinities, Von Guerard Algal' mats a re cOl11m on in many of the stream beds, but and Crescent Streams (Fig. I). If weighted by discharge, the are not evenly distributed (Alger and others, 1997). Most of average inflow into Lake Fryxell has a (j13 C of - 1.7 %0. In the streams consist of slow-flowing stretches that are covered Lake Hoare basin, only Anderson Creek has been sampled with algae, and fast-moving parts that flow over soil and for carbon isotopes, and the values are - 0. 32%0 and - 3.21 %0 mora inalmaterial. Al gal mats could function both as a sink for samples collected at different dates. and as a source ofDIC to the streams. When they are photo- [DIC] and (j 13 C profiles for Lakes Fryxell and Hoare are shown in Figure 2. The [DIC] profiles are from November 1994. Both lakes show an increase in [DIC] with depth. Lake Fryxell Lake Hoare 4.5 4.5 - . Lake Hoare's values increase to 5.9 mM at 18 m depth, and i Ice i Ice I then remain consta nt, while values in Lake Fryxell increase Jan,x., ~ Dec. to the bottom, reaching 59 mM. At the end of the austral 5.5 5.5 summer both lakes show a trend of dec reasing DIC in the l< I I .. upp e nn ~s t parts of the water column (Fig. 3). The isotope 6.S ~6 . 5 a a> profiles of Lakes Fryxell and Hoare are similar in sh ape. ,3 0 Va lues are enriched in the upper part of the waler column. 7.5 7.5 In Lake Fryxell, they decrease below the chemoc1ine at 9 m · .•. Dec i< .. A B 8.5 8.5 0 0 2 4 6 8 10 1.6 1.8 2 2.2 2.4 Ice DIC{mM] DIC{mM] 5 5 Lake Fryxell Lake Hoare 1995, 10 4.5 . ---,,- - --, -·-1 4.5 i _. 7~r.p Ice Ice Fryxell ~x Jan. \. :[15 :[15 Oct. 5.5 5.5 P~)'"'" .s:: .s:: \ . . g.20 g.20 \> ~ ~ 4:·· i Cl Cl Cl I I d.5 £ 6.5 25 Cl Hoare 25 Co 8 15. a> Hoare 19S.7.. d : 0 0" Nov. ~ ~ 30 30 IY· A B 7.5 7.5 '.. j 35 35Lw~~~~~~LW~ Qct. Ded A ~A Dec. o 10 2030 405060 -6 -4 -2 0 2 4 6 e 0 mM HC0 • 8.5 3 8.5 0 0.2 0.4 0.6 0.8 0.005 0.01 0.015 0.02 0.025 Fig. 2. ( a) Concentrations qfbicarbonate in Lakes Hoare and CO,", {mM] CO,.. {mM] Fr)'xeLl. Lake FryxeLl has an oxycline and chemocline at 9 m

depth; salinities increase sharply below. (b) DIC- (j 13C dis­ Fig. 3. Seasonal changes in DIC and C0 2aq concentrations in tributions in the same lakes. Dala for Lake Hoare are fro m Lakes Fryxell alld Hoare. Nole the dijJerence in the scalesfor Wharton and others (1993). the concen trations, especially in [ CO2l aq. Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 05:30:19, subject to the Cambridge Core terms of use. 687 Neumann and others: Carbon-isotope distribution in Taylor Valley lakes

syntheticall y active they may take up DIC, and increase the (7- 8; see Table I) than in Lake Hoare (8.5- 8.8). Stream water <5I3C of the DIC remaining in the water by preferential up­ shou ld be relatively buoyant in Lakes Hoare and Fryxell, take of light 12C (Des Marais and Canfield, 1994; Des Mar­ thus remaining direc tl y underneath the ice. Stream water 13 ais, 1995). When they res pire, they add isotopically light CO2 entering Lake Fryxell has negative (5 C values, and that to the water. Of the short streams, Mariah Creek had no sig­ can be refl ected in the li ghter composition of the uppermost nificant mats present over the period of sampling, while lake water. large parts of the bed of Green Creek were covered with To calculate the inorganic carbon balance for Lakes mats. However, the DIC is roughly the same in both streams. Fryxell and Hoare, we modified an equation from Quay On the other hand, Green Creek had a <5 J3 C of - 4.55%0. DIC a nd others (1986): in equilibrium with the atmosphere should have <5 13 C values V d(DIC)L = mIC - S of +3%0 to +2%0, for temperatures ofO- lO °e. The low values dt 1 observed in Green Creek can only be caused by the respira­ 3 where V is volume of the lake, I is inflow (m Si), S is con­ tion of light organic matter, i.e. algal mats. Accounting for the various discharge rates of the streams, the average DIC version of DIC to organic carbon and removal to deeper input into Lake rryxell has a <5 13 C of - 1.7%0. Even given the water (mol S- I), and DICL and DIC] are dissolved inorganic uncertainties caused by the incomplete measurements carbon concentrations in the lake and inflows (mol m -3). A factor that is not accounted for in this equation is the re­ (",20% of the inflow was not sampled ), this value is more moval ofDIC by calcite precipitation. In Lake Washington, negative than the value expected for DIC equilibrated with the dissolution and/or precipitation of CaC0 was not sig­ the atmosphere. The two samples in Lake Hoare basin were 3 taken at the mouth of Anderson Creek at two times of the nificant (Quay and others, 1986). Our calculations show that 13 season, a nd had (5 C values of - 0.32%0 and - 3.21%0. The al­ gal mats are less important for the DIC loading than the Table 2. Water discharge, pH, pC03 and [ CO zlaq fiux in stream length, but they seem to have a significant impact Lake Fryxell and Lake Hoare basins, 1994- 95 on the isotopic composition of the DIG

Discharge COd'ux Lakes 3 m ol 111- m ala I The chemical compositIOns of Lakes Hoare a nd Fryxell have been discussed extensively (e.g. Angino and others, Fryxell 152750 0.0610 9900 Canada Stream 41710 7.4 - 3.08 0.0343 1430 1962; Wi lson, 1979; Green and others, 1988; Lawrence and Lost Seal Stream 20630 7.2 - 2.47 0.2021 4170 Hendy, 1989; Wharton a nd others, 1989). The isotopic data Aiken Creek 97 10 8.2 - 3.83 0.0175 170 for Lake Hoare from \Vharton and others (1993), data for Von Guerard Stream 7220 7.8 - 3. 19 0.0488 350 Lake Fryxell in 1994 from Neumann and others (1995) and Crescent Stream 2110 8.0 - 2.92 0.057 1 120 Delta Stream 3280 7.9 - 3.21 0.0407 130 additional data for both lakes from the 1995- 96 season are Green Creek 29 120 7.3 - 3.25 0.0582 1690 shown in Figure 2. Lake Hoare and Lake Fryxell profiles are McKnight Creek 6800 7.7 - 3.01 0.0551 380 strikingly similar: both have bottom waters with negative Harnish Creek 2500 7.8 - 3.25 0.0752 190 values that increase towa rd the surface. The steepest Bowles Creek 10 480 6.8 - 3.06 0.0544 570 Ma riah Creek 12520 7.3 - 3.65 0.0175 220 (5 13 increase of C values in Lake Fryxell occurs at the chemo­ Andrews Creek 6670 6.8 - 3.01 0.0715 480 dine at 9m depth. In 1995 the uppermost part of the water column of Lake Fryxell had the same structure as Lake H oare 71 350 0.0188 1310 13 Anderson Creek 16050 7.6 - 3.32 0.0370 590 Hoare, with the heaviest <5 C compositions with values House Stream 14180 8.3 - 4.19 0.0128 180 around +3%0 just below the ice. In 1994 the uppermost Wha rton Creek 25520 7.9 - 3.88 0.0137 350 sample had a value of + 0.41 %0. The fractionation of <5 l3 C in McKay Creek 15600 8.2 - 3.81 0.0119 190 the water column is a res ult of bi ological activity. The Fryx­ ell profi le refl ects primary production maxima at 9 m and just beneath the ice. The maxima are caused by availability Lake FryxeU Lake Hoare of nutrients at the chemodine (which also is a "nutridine"; 0 0 Priscu, 1995), and the highest irradiance in the lake just un­ A Ice B 2 derneath the ice. Ice 5 [DIC] also increases with depth in both lakes (Fig. 2). 4 [DIC] in the bottom water of Lake Fryxell is one order of Is I 10 magnitude larger than the concentration in Lake Hoare. .<: is. .<:a. [C02]aq also increases with depth. The difference between ~8 ~15 [DIC] and [C0 is that, in the uppermost meter of the 2]aq 10 water column, [DIC] decreases over the season, while 20 [C02] aq values increase (Fig. 3). Over the austral summer 12 1994- 95, pH values decreased from 8.8 to 8.5 in Lake Hoare, 14 25 and from 7.6 to 7.0 in Lake Fryxel1. This can be caused by two 10" 10" 10·' 10·' 10° 10' 10" 10" 10·' 10·' 10° processes: the melting of lake ice and associated percolation CO. conc.: mM CO. conc.: mM of low-DIC meltwater into the lake water, and the influx of CO. Uptake, Influx: mM yr' CO. Uptake, Influx: mM yr' stream water in the late season. Stream water is low in

[DIC], but high in [C02] aq (Table 2), while the uppermost Fig. 4. CO 2 uptake ( Priscu, 1995), recharge via streams, and lake water in Lake Hoare is depleted in [C02] aq (Fig. 4). C02aq concentrations in Lakes Fryxell and Hoare. Note the T he pH values are also slightly lower in the stream waters different scalesfor depth and concentrationslfluxes. Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 05:30:19, subject to the Cambridge Core terms of use. 688 Neumann and others: Carbon-isotope distribution in Taylor Valley lakes

the upper meters in Lake Fryxell change from super- to Lake Fryxell water level would also drop, but the high undersaturation with respect to CaC03 through the season. [DIC] will reduce effects on the carbon cycle. These shifts In Lake Hoare, water above 15 m is supersaturated with re­ in isotopic composition are possible only because the lake spect to calcite, and CaC03 minerals have been found in waters are separated from the atmosphere by the perennial sediments (Wharton and others, 1989). At present, we have ice cover. In lakes in higher latitudes deficits in C02aq are no data that would all ow any qualitative estimate of calcite compensated by dissolution of atmospheric CO2 into the precipitation in these lakes, and we cannot include this pro­ water. Thus Lakes Fryxell and H oare offer sedimentary re­ cess in our calculation. positories that could yield important paleoclimatic and pa­ Lake Fryxell should be affected more strongly by the DIC leobiological information. input from streams than Lake Hoare because it receives much higher loads ofDIC (Tables I and 2). Using the modified for­ CONCLUSIONS mula, normalizing the DIC input with the lake area and assuming a mixing of the incoming stream water with the The carbon cycle in the perennially ice-covered lakes in uppermost meter of the lake's water column, Lake Fryxell an­ Taylor Valley is mainly governed by three factors: (1) the ice nually received 3.8%,0.7% and 2.1 % (average = 2.2% ) of cover that limits gas exchange with the atmosphere, (2) the the DIC pool in the uppermost meter. The values for Lake amount of DIC recharge via streams, and (3) the primary Hoare are 9.3%, 5.5% and 5.5% (average = 6.8% ). The ab­ productivity in the lakes. Removal of inorganic carbon via solute influx into Lake Hoare is much smaller, but because calcite precipitation may also be important, but currently the lake has such a low concentration of DIC it plays a more this process cannot be quantified. The amount of DIC in important role than in Lake Fryxell. the streams is dependent on the stream length, with longer Primary productivity rates (PPRs) from Priscu (1995) streams having higher DIC values due to more extensive were used to compare the average annual uptake of carbon dissolution of calcites. Algal mats in the streams seem to at various lake depths with the available DIe. At 5- 8 m have little influence on the DIC concentration, but might depth, the annual uptake accounts for 0.43 % of the DIC in have an effect on the isotopic composition. In the lakes, Lake Hoare, and for 0.76% of the DIC in Lake Fryxell. [DIC] and [C02]aq increase with depth, while PPR and iso­ These ratios are very similar, given that the PPR in Lake topic composition decrease. A seasonal decrease ofDIC and Fryxell is roughly four times as high as in Lake H oare, and pH and an increase ofpC02 occur in the uppermost water do not explain the different isotopic signatures found in the column. The annual carbon uptake by phytoplankton at 5- lakes. Most algae do not use all DIC as source for carbon, 9 m depth represents 9% of the C02aq in Lake Fryxell, but but rather a small fraction of it, [C0 Jaq' [C0 Jaq is much 2 2 64% of the C02aq in Lake Hoare. This high utilization rate smaller in Lake Hoare (see Fig. 3), due to lower DIC and causes the continuously heavier (j1 3C composition in Lake higher pH. The resulting ratio of [C02]aq uptake vs supply H oare. Lakes Hoare and Fryxell will react differently to shows a large difference between the two lakes (Fig. 4). climate changes, with Lake Hoare, with its small pool of While in Lake Fryxell the annual carbon uptake accounts DIC, being afIected more directly than Lake .Fryxell, with for only 9% of the [C02Jaq at 5- 9 m depth, in Lake Hoare its higher DIC concentrations. (5- 8 m ) it accounts for 64%. This higher percentage uptake

of CO2 in Lake Hoare explains why the carbon-i sotope composition in the uppermost water column is consistently ACKNOWLEDGEMENTS heavier than in Lake Fryxell. Wharton and others (1993) We would like to thank K. Welch of the University of Ala­ showed that a large part of the organic matter produced in bama, and R. Edwards of Montana State University, Boze­ the upper part of the water column is respired before it man, for their extensive help in the field and in the becomes incorporated into the sediment. This respiration laboratori es of Antarctica. A. Tharpe of NASA Ames results in increased [DIC] values and light isotopic compo­ Research Center performed the (jl3C measurements. We sition in the lower part of the water column. thank]. Priddle and H. Kennedy for reviewing this paper. How will the carbon cycle and the isotopic signatures of This work was supported by NSF grant OPP-92/1773. Lakes Fryxell and Hoare react to climate changes? An increase in temperature would increase water and DIC dis­ charge into the lakes. The main effect would be to raise lake REFERENCES level. The carbon budget of Lake Hoare will be greatly im­ pacted, because the original [DIC] is much lower than that Alger, A. S. and 8 others. 1997. Ecological processes in a cold desert ecosystem: the abundance and distribution cif algal mats in glacial meltwater streams in Taylor of Lake Fryxell. However, inflow of low-TDS stream water f'allry, Antllldira. Boulder, CO, University of Colorado. Institute of Arc­ would dilute the nutrients dissolved in the lake and probably tic and Alpine Research. (INSTAAR Occasional Paper 51.) decrease the productivity. This wi ll lead to less carbon uptake, Angino, E. E. , K. B. Armitage and]. Glllsh. 1962. Chemical stratification and lighter carbon-isotopic compositions in the lake too. In in Lake Fryxell, Victoria Land, Antarctica. Science, 138(3536), 34-36. Campbell, 1. B. and C. C. G Cia ridge. 1987. Antarctica: soils, weathering pro ­ situ produced calcite will have light (j13C values, or the cesses and environment. Amsterdam, etc., El sevier. (Developments in Soil [DIC] will become so low that calcite will be undersaturated. Science 16.) A decrease in temperature and discharge would lower Chinn, T.J. H. 1985. Structure and equilibrium of the Dry Valleys glaciers. the water and DIC input into the lakes. In Lake Hoare this NZ Alltare. Ree., 6, Special Supplement, 73-88. Clow, C. D. , GP. M cKay, C. M. Simmons,Jr and :!{. A. Wharton,Jr. 1988. should lead to a heavier carbon-isotope composition. The Climatological observations and predicted sublimation rates at Lake already small carbon supply would be utilized furthe l~ and Hoare, Amarclica. J Climate, 1 (7),715- 728. the DIC would become heavier. Increased calcite precipita­ Cutfield, S. K. 1974. Hydrological aspects of' Lake Vanda, Wright Valley, Victoria Land, Amarctica. NZ Geol. Geophys., 17 (3),645- 657. tion would effectively remove carbon from the system, po­ J Des Marais, 0..J. 1995. The biogeochemist ry ofhypersaline microbial mats. tentially making the system carbon-deficient. Calcite Adv. Microbial E[OI., 14, 251 - 274. precipitated in these conditions will be isotopically heavy. Des Marais, D.]. and 0. E. Canficld. 1994. The carbon isotope biogeochem-

Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 05:30:19, subject to the Cambridge Core terms of use. 689 Neumann and others: Carbon-isotope distribution in Taylor VallI!Y lakes

istry of microbial mats. In Stahl, L.]. and P. Caumette, eds. Microbial Melack, ]. A 1983. Large, deep salt lakes: a comparative Iimnological ana­ mats. Berlin, etc. , Springer-Verlag, 289- 298. (NATO AS} Series G35.) lysis. Hydrobiologia, 105, 223-230. Doran, P. T , R. A. Whanon, j r and W. B. Lyons. 1994. Paleoli m nology of Ncumann, K. , W. B. Lyo ns, nJ. Des Marais, ]. c. Prisc u and R. L. Ed­ the McMurdo Dry Va ll eys, Antarctica. ] PaleoLi",nol., 10, 85- 114. wards. 1995. 13C and pCOz distribution in Antarctic la kes: ice cover Drever, J. r. 1988. The geochemistry of natural waters. Second edition. Englewood and bi ota leave their imprint. [Abstract] EOS, 76 (46), Fall Meeting Sup­ Cliffs, l'{J, Prentice H all. pl ement, 1'215. Eugster, H. P. 1980. Geochemi stry of evaporitic lacustrine deposits. Annu. Parkhurst, 0. L. , 0. C. Thorstenson and 1. N. Plummer. 1980. PHREEQE Rev. Earth Planet. Sci., 8, 35- 63. - a computer program for geochemical eva lu ations. Us. Geoi. Su m Gilbert, G. K. 1890. Lake Bonnevill e. Us. Geol. Sum Monogr: I. Water-Resour. Invest. Rep. 80-96. Gree n, Vi. L. , M. P. Angle and K. E. Chave. 1988. The geochemistry ofAnt ­ Plummer, 1. N., 0. L. Parkhurst, G. W. Fleming and S. A. Dunkle. 1988. A arctic streams a nd their role in the evolu tion of four lakes in the computer program incorporating Pitzer's equations for calculation of McMurdo DryValleys. Geochim. Cosmochim. Acta, 52, 1265-1274. geochemical reactions in brines. u.s. Geo!. Sum Water-Resour. Invest. Rep. H ardie, L.A .,j. P. Smoot and H. P. Eugster. 1978. Saline lakes a nd their de­ 88-4135. posits: a sedimentological approach. In Matter, A. a nd M. Tucker. , eds. Prisc Ll,j. C. 1995. Phytoplankton nutrient deficiency in lakes of the MclvIur­ Modern and ancient lake sediments. O xford, International Association of do Dry Vall eys, Antarcti ca. Freshwater Bioi., 34(2), 215-227. Sedimentology, 7- 41. (Special Publication 2.) Quay, P. 0., S. R. Emerson, B. M. Quay and A H. Devol. 1986. The carbon H ayes,]. M. , nJ. Des Marais, 0. W. Peterson, 0. A. Schoell er and S. P. Tay- cycle for Lake Wa shington - a stable isotope study. Limn ol. OcearlOgr. , 31, 101'. 1977. High precision stablc isotope ratios from mi crogram samples. 596- 611. Adv. ill Ma ss Spectrosc., 7, 475- 480. Vincent, "". F. 1987. Anta rctic limnology. InViner, A. 8. , ed. inland waters of jones, L. M., R . E. Carver, E. R. McSaveney and T. Tickhill. 1971. Sediment New Zealand. \Vellington, ="ew Zealand, Department of Scientific and a nalys is of beaches of Lake Vanda, Wright Valley. Antarct. ] US., 6(5), Industrial Research, 379- 412. (DS1R Bulletin 24 1.) 199- 200. Welch, K. A , W. B. Lyo ns, E. Graham, K. Neuma nn, j. M. Thomas and D. Lawrence, M.]. F. and C. H. Hendy. 1989. Carbonate depos ition and Ross Mikesel l. 1996. The determination of major element chemistry in terres­ Sea ice advance, Fryxell Basin, Taylor Valley, Antarctica. N.Z ] Geol. trial waters from Anta rctica using ion chromatography. ] Chromatogr. , Geoph)!s., 32 (2), 267- 277. Ser. A , 739, 257- 263. Lizotlc, M. P and]. C. Prisc u. 1992. Photosynthcs is-irradiance relationships Wharton, R. A. , B. C. Pa rker and G. M. Simmons, Jr. 1983. Distribution, in phytoplanklon from thc physicall y stable water column of a peren­ species composition a nd morphology of algal mats in Antarctic Dry nially ice-covered lake (Lake Bonney, Antarctica). J Phycol., 28, 179-185. Valley lakes. Phycologia, 22 (4),355- 365. McKenzie, ]. 1985. Carbon isotopes and productivity in the lac Ll strine and Whanon, R. A. , C. P. M cK ay and G. M. Simmons, Jr. 1989. Perenni all y marine environment. In Stumm, W. , ed. Chemical processes in lakes. New ice-covered Lake Hoa re, Antarctica: phys ical environment, biology York, Wiley lnterscience, 99- 11 8. and sedimentation. Hydrobiologia, 172, 306- 320. Mc Knight, 0. M. , G. R. Aiken and R. L. Sm ith. 1991. Aquatic fulvi c acids Wharton, R. A., C. P. McKay, G. 0. Clow, 0. T Andersen, G. M. Simmons in microbially bascd ecosystcms: results from two dcsertlakes in Antarc­ and F. G. Love. 1992. Changes in ice cover thickness and lake level of tica. Limnol. Oceanogr., 36(5),998- 1006. Lake H oare, Antarctica: implications for local climate change. ] Geo ­ McKnight, 0. M. , G. R. Aiken, E. D. Andrews, E. C. Bowles and R. A. ph),s. Res., 97 (C3), 3503- 3513. Harnish. 1993. Dissolved organic matter in Dry Valley lakes: a compar­ Wharton, R. A. , W. B. Lyons and 0.]. Des Marais. 1993. Stable isotope bio­ iso n of Lake Fryxell , Lake Hoare and Lake Vanda. In Green, W]. and geochemistry of carbon and nitrogen in a perennially ice-covered Ant­ E. 1. Fried man, eds. Physical and hiogeochemical processes ill Antarctic lakes. arctic lake. Cltem. Genl. , 107 (1- 2), 159- 172. Was hington, DC, American Geophysica l Union, 11 9- 134. (Anta rctic Wil son, A. T. 1979. Geochemical problems of the Antarctic dry areas. Nature, Resea rch Series 59.) 280(5719), 205- 208.

Downloaded from https://www.cambridge.org/core. 01 Oct 2021 at 05:30:19, subject to the Cambridge Core terms of use. 690