Climate modulates the acidity of Arctic lakes on millennial time
scales
Alexander P. Wolfe Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB T6G 2E3, Canada
ABSTRACT
High-resolution diatom stratigraphies from two lakes on eastern Baffin Island (Nunavut,
Canadian Arctic) are used to reconstruct lake-water pH over the past 5 k.y. Despite contrasting geomorphic histories and markedly different diatom floras, the inferred pH of
both lakes declined in concert with Neoglacial cooling. This result confirms that climate
exerts a first-order influence on the acidity of these poorly buffered oligotrophic lakes,
through links between lake ice cover, primary productivity, and dissolved inorganic carbon dynamics. Diatom-based pH inferences from dilute lakes can therefore assist paleoclimatic interpretations from lake sediments.
Keywords: diatoms, lake acidity, Holocene, Neoglacial, Baffin Island.
INTRODUCTION
The suggestion that climate controls the pH
of oligotrophic lakes originated from studies
in the Alps (Psenner and Schmidt, 1992; Sommaruga-Wögrath et al., 1997). In these analyses, cold air temperatures induced a lowering
of lake-water pH, and vice versa, through
links between lake alkalinity and catchment
weathering, eolian dust fluxes, lake metabolism, and dissolved inorganic carbon speciation. However, these inferences are based on
paleolimnological records limited to the past
200 yr. In most lakes, the apparent relationship
between air temperature and pH collapsed
once anthropogenic overprinting, in the form
of atmospheric deposition of strong-acid anions, came into play during the twentieth century. Therefore, at present there is only tantalizing evidence for the direct coupling of
climate and lake-water acidity (Koinig et al.,
1998). It is of considerable interest to explore
whether climate regulates the acidity of dilute
lakes over longer time scales for two reasons.
First, lake-water pH can be reconstructed reliably from sediment diatom assemblages
(Battarbee et al., 1999). Second, growing interest in climate change is stimulating new approaches at gleaning paleoclimatic information from lacustrine archives, especially in
remote regions (Smol and Cumming, 2000).
Furthermore, the possibility exists that current
changes in the Arctic climate system may alter
the chemical regimes of high-latitude lakes,
potentially influencing their susceptibility to
anthropogenic atmospheric deposition. In this
study, high-resolution diatom-inferred lakewater pH histories for the past 5 k.y. are presented from two ultraoligotrophic Arctic
lakes. These reconstructions are assessed by
comparing them to independent late Holocene
paleoclimatic reconstructions. The results suggest that a direct link exists between climate
and lake acidity in this environment.
REGIONAL LATE HOLOCENE
PALEOCLIMATOLOGY
There is consensus that the late Holocene
was an interval of progressive cooling in the
Baffin Bay region, as inferred by isotopic, palynological, and glacial geologic evidence
(Fig. 1).
Although the area has undoubtedly been
subject to submillennial climate variability,
the low-frequency component of this cooling
appears to be regulated by high-latitude summer insolation. Late Holocene cooling (Neoglacial) culminated in the Little Ice Age ca.
A.D. 1450–1850. Thereafter, summer temperatures have increased markedly (Overpeck et
al., 1997), partially in response to greenhousegas forcing.
STUDY SITES AND CORE
CHRONOLOGIES
The two investigated lakes are 75 km apart
on the north coast of Cumberland Peninsula
(Fig. 2). Mean annual temperature is ;210 8C,
and total annual precipitation is ;350 mm.
Kekerturnak Lake (678539N, 648509W; 120 m
above sea level [masl]) occupies a bedrock basin in outer Qajon Fiord and is younger than
9 ka, having been overridden during Late
Foxe glaciation (Steig et al., 1998). In contrast, Fog Lake (678119N, 638159W; 460 masl)
is perched on an unglaciated upland and preserves a sediment record of at least 90 k.y.
(Wolfe et al., 2000). The proximity of the
lakes implies they have undergone the same
Holocene climate forcing, despite large differences in geomorphic history and age. This facilitates the distinction of purely limnological
responses to climate, which should be equally
expressed at both sites, from nonclimatic
edaphic factors. Regional bedrock geology is
Archean granite and gneiss. Both lakes are
highly dilute (conductivities ,10 mS·cm21;
[Na11 K11 Mg211 Ca21] ,5 mg·L21) and
are therefore largely unbuffered. Modern summer epilimnetic pH is 6.32 in Fog Lake and
6.25 in Kekerturnak Lake, whereas HCO32
Figure 1. Summary of late Holocene climatic evolution of Baffin Bay. A: Summer insolation at 658N (thick line) (Berger and Loutre, 1991) and oxygen isotope stratigraphy (SMOW—standard mean ocean water) from Devon Island Ice
Cap (Paterson et al., 1977). B: Pollen-based summer temperature departures
integrated for Baffin Island, Labrador, and Keewatin (Andrews et al., 1981). C:
Time versus distance diagram of glacial activity in Pangnirtung Pass, southern
Cumberland Peninsula (Davis, 1985).
q 2002 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected].
Geology; March 2002; v. 30; no. 3; p. 215–218; 4 figures; 1 table.
215
Figure 2. Location of study sites on northern Cumberland Peninsula. Open circles show locations of lakes used to develop diatom-based inference models (Joynt and Wolfe, 2001).
concentrations are 0.09 and 0.11 mg·L21,
respectively.
Cores of 35 cm length were collected with
a gravity corer (Glew, 1989) and subsampled
into continuous 0.5 cm increments in the field.
Both cores are composed of massive olivebrown algal gyttja. Each core’s chronology is
based on five accelerator mass spectrometry
14C dates from sediment humic acid extracts
(Abbott and Stafford, 1996). Humic acid 14C
results were corrected by subtracting 300 yr
to account for the mean lag associated with
the sequestration of terrestrial dissolved organic carbon to the lakes (Miller et al., 1999)
and calibrated to calendar years before present
(Stuiver et al., 1998). The resulting chronologies produce age versus depth relationships
fitted by polynomial solutions (Table 1; Fig.
3, A and B).
Figure 3. A and B: Age models for Kekerturnak and Fog lake gravity cores, based on humic acid accelerator mass spectrometry 14C
dates. C: Relationship between surface-sediment diatom-inferred pH
and measured pH from 61 lake training set. Root-mean-squared error of prediction (RMSE) and bootstrapped RMSE (in pH units) are
included. D: Residuals for same relationship, demonstrating absence of bias at either extreme of pH.
DIATOM-INFERRED pH
RECONSTRUCTIONS
Diatom slide preparation, counting, and taxonomy followed standard protocols described
elsewhere (Wolfe et al., 2000). A known quantity of markers (Eucalyptus pollen) was added
to digested slurries in order to calculate diatom concentrations (Wolfe, 1997). At least
500 diatom valves were counted in contiguous
0.5 cm samples for the entire Fog Lake core
and the upper 20 cm of the Kekerturnak Lake
core, and every 1.0 cm below 20 cm.
Lake-water pH inferences were made by
several methods. (1) A regional transfer function was applied to down-core diatom assemblages, based on weighted-averaging (WA) regression and calibration of surface-sediment
diatom assemblages and measured limnological parameters from 61 Baffin Island lakes
TABLE 1. RADIOCARBON DATES FROM FOG AND KEKERTURNAK LAKE CORES
Depth
(cm)
Kekerturnak Lake
6.5–7.0
13.5–14.0
20.0–21.0
27.0–28.0
34.0–35.0
Fog Lake
6.5–7.0
13.5–14.0
20.5–21.0
27.5–28.0
34.5–35.0
216
Laboratory
number
14
C age
(yr B.P.)
6s
(yr)
Corrected age
(yr B.P.)
Calibrated age
(yr B.P.)
Range of solutions
(62s)
CAMS-44438
CAMS-44439
CAMS-44440
CAMS-44441
CAMS-44442
1190
2250
3110
3900
4780
50
50
60
50
50
890
1950
2810
3600
4480
790
1915
2920
3870
5205
675–925
1740–1990
2770–3070
3725–4075
4875–5300
CAMS-44443
CAMS-44444
CAMS-44445
CAMS-44446
CAMS-44447
1230
2080
3030
3720
4560
50
50
50
50
50
930
1780
2730
3420
4260
885
1695
2835
3675
4835
725–940
1550–1820
2755–2940
3485–3825
4645–4870
(Joynt and Wolfe, 2001). In this study, pH
emerged as a highly significant variable (Fig.
3, C and D). (2) Because the diatom flora of
Baffin Island lakes is taxonomically similar to
that in dilute lakes elsewhere in eastern North
America, pH inferences were also made by
WA calibration using the pH optima from the
Paleoecological Investigation of Recent Lake
Acidification (PIRLA; Camburn and Charles,
2000), and the Environmental Monitoring and
Assessment Program (EMAP; Dixit et al.,
1999). (3) The overall importance of pH as a
predictor of diatom assemblage composition
was evaluated by comparing transfer function
results to sample scores on the first principal
component extracted from the diatom relative
frequencies. Principal components analysis
(PCA) is an indirect ordination technique that
models species abundances as linear responses
along synthetic orthogonal axes (ter Braak,
1987). In contrast to WA inferences, PCA
does not employ any a priori ecological characterization of diatom taxa with respect to pH
or any other environmental variable.
Although the two lakes are chemically similar, their diatom floras differ markedly. Before ca. 2500 yr B.P., the sediments of Kekerturnak Lake were dominated by the
circumneutral taxa Cyclotella rossi, Brachysira vitrea, Aulacoseira lirata, and Stauroforma exiguiformis. Thereafter, these taxa were
largely replaced by the more acidophilic A.
GEOLOGY, March 2002
more complicated (Fig. 4); two intervals of
increasing pH (5000–4000 yr B.P. and 3000–
2000 yr B.P.) are separated by periods of more
acidic conditions, the latter of which (1800–
200 yr B.P.) was characterized by a pH of
;6.0. A rapid pH increase is registered in the
uppermost samples.
The robustness of these trends is verified by
the PCA results (Fig. 4), which closely mimic
WA pH inferences. Alkaliphilic diatoms (fragilarioid taxa) load positively on axis 1, whereas low species scores are typical of acidophils
(Aulacoseira distans, A. perglabra, Eunotia
spp.). The remarkable correspondence between
the PCA and WA results indicates (1) that pH
is the dominant environmental control over diatom assemblage composition, and (2) that WA
pH inferences faithfully track past variability in
this parameter.
Figure 4. Diatom-inferred pH reconstructions from (A) Kekerturnak Lake and (B) Fog Lake
over 5 k.y. Three curves at left are based on weighted average calibration, using Baffin Island
training set (circles), and pH optima from Paleoecological Investigation of Recent Lake Acidification (PIRLA) and Environmental Monitoring and Assessment Program (EMAP) (gray
lines). Arrows indicate measured pH; pH inferences are shown next to principal components
analysis (PCA) first axis sample scores. At right are profiles of diatom concentrations.
distans and C. stelligera. In Fog Lake, sediments deposited between 5000 and 2000 yr
B.P. are characterized by S. exiguiformis and
Staurosira construens var. venter, circumneutral and alkaliphilic taxa, respectively. The
acidophils A. distans, Eunotia rhomboidea, E.
vanheurkii, and Peronia fibula became conspicuous after 2000 yr B.P., until the upper 2
cm of the core, when S. exiguiformis rebounded, accompanied by increased frequencies of
Nitzschia perminuta.
Despite these ecological differences, diatom-inferred pH reconstructions from both
lakes record a coherent pattern of decline
throughout the late Holocene (Fig. 4). Although WA reconstructions using the PIRLA
GEOLOGY, March 2002
and EMAP optima differ from those based on
the Baffin Island training set, the directions
and amplitudes of inferred pH change are parallel. Discrepancies between the inference
models arise from sampling that alternately biased more acidic (PIRLA) and more alkaline
(EMAP) lakes than those on Baffin Island.
Close correspondence between measured pH
and core-top reconstructions (Fig. 4) indicates,
as expected, that the Baffin Island model provides the most accurate pH inferences. Diatominferred pH consistently exceeded 6.6 prior to
2500 yr B.P. in Kekerturnak Lake; peak values
were ;6.8 between 4000 and 3000 yr B.P. After 2500 yr B.P., pH decreased to modern values (6.2–6.3). The record from Fog Lake is
DISCUSSION AND CONCLUSIONS
Several processes linking lake-water pH
and climate are known (Sommaruga-Wögrath
et al., 1997; Koinig et al., 1998). In lakes on
eastern Baffin Island, edaphic contributions to
alkalinity from base cations derived by weathering and eolian dust are unlikely to be as important as in other regions. In addition to the
highly resistant character of local bedrock,
lakes are icebound for more than 10 months
of the year, effectively isolating them from external influences for all but a few weeks annually. For example, increased dust flux over
Cumberland Peninsula during the late Holocene (Zdanowicz et al., 2000) did not attenuate lake acidification trends through enhanced
atmospheric loading of base cations.
Instead, it is argued that the coupling of climate and pH is regulated by within-lake processes, the most important of which involve
dissolved inorganic carbon (DIC) speciation.
Given low solute concentrations, lake-water
pH, whether measured directly or remotely
sensed by diatom assemblages, is regulated by
the DIC system, which is in turn modulated
by the duration of ice cover. Although CO2
supersaturation occurs in a variety of lake
types worldwide (Cole et al., 1994), it is prevalent when ice precludes the evasion of respired CO2 to the atmosphere. For example,
CO2 partial pressures in Alaskan lakes are 3–14
times greater under ice than during the openwater season (Kling et al., 1992). Furthermore, ice cover reduces light penetration and
hence primary productivity, so that photosynthetic drawdown of dissolved CO2 is minimal.
Because CO2 dominates the DIC pool under
such conditions, pH declines accordingly,
which is the situation envisaged for cold paleoclimatic intervals. Conversely, during
warmer periods, longer ice-free seasons promote equilibration between dissolved and atmospheric CO2, eliminating supersaturation in
217
the water column. Enhanced algal production
further reduces limnetic CO2, shifting DIC toward greater proportions of HCO3–, and driving pH increases. When applied to the Baffin
Island lakes, this model predicts a reduction
of algal productivity in concert with Neoglacial cooling, which is supported by progressive decreases in diatom concentrations (and
fluxes) in both lakes (Fig. 4).
The late Holocene pH trends from Kekerturnak and Fog Lakes are not monotonic. Both
lakes record increased diatom-inferred pH between ca. 3500 and 2500 yr B.P., followed by
marked acidification. The latter may reflect
limnological responses to cooling associated
with enhanced Arctic Ocean outflow into Baffin Bay (Dyke et al., 1996). Both lakes
reached their pH minima after 2000 yr B.P.,
in agreement with independent records of cold
prevailing conditions (Fig. 1). The rapid increase in pH observed in the post–Little Ice
Age sediments of Fog Lake is not captured in
Kekerturnak, although both lakes register increases in diatom concentrations during this
interval. It remains to be shown how directly
these observations relate to recent warming of
the Arctic (Overpeck et al., 1997).
The sensitivity of lake-water pH to temperature can only be estimated at present. If the
amplitude of Neoglacial summer cooling at
the latitude of the study sites is estimated as
3.0 8C (Andrews et al., 1981; Dahl-Jensen et
al., 1998), then pH appears to be coupled to
air temperature on the order of 0.2 units per
degree Celsius. From the diatom-inferred pH
results, late Holocene cooling therefore appears to have been punctuated by several
events of 1–2 8C amplitude. Although additional records are needed to fully test this possibility, it is consistent with suggestions of
pronounced Holocene climate variability in
the North Atlantic sector (Bond et al., 1997).
Although climate warming may raise the pH
of Baffin Island lakes by prolonging the icefree season, this relationship is fully dependent upon the unbuffered character of these
lakes, which allows the direct control of pH
by DIC dynamics. This same condition enhances the vulnerability of these lakes to atmospheric inputs of strong-acid anions of anthropogenic origin, implying that the
relationship between climate and pH may easily be disrupted.
ACKNOWLEDGMENTS
This research was funded by the National Science Foundation (ATM-9402657). The support of
the Nunavut Research Institute and the hamlet of
Qikiqtarjuaq is gratefully acknowledged. I thank
218
Gifford Miller and Ernest Joynt for data and ideas
and the reviewers for improving clarity and content.
REFERENCES CITED
Abbott, M.B., and Stafford, T.W., Jr., 1996, Radiocarbon geochemistry of modern and ancient
Arctic lake systems, Baffin Island, Canada:
Quaternary Research, v. 45, p. 300–311.
Andrews, J.T., Davis, P.T., Mode, W.N., Nichols, H.,
and Short, S.K., 1981, Relative departures in
July temperatures in northern Canada for the
past 6000 yr: Nature, v. 289, p. 164–167.
Battarbee, R.W., Charles, D.F., Dixit, S.S., and Renberg, I., 1999, Diatoms as indicators of surface
water acidity, in Stoermer, E.F., and Smol, J.P.,
eds., The diatoms: Applications for the environmental and earth sciences: Cambridge, UK,
Cambridge University Press, Ltd., p. 85–127.
Berger, A., and Loutre, M.F., 1991, Insolation values for the climate of the last 10 million
years: Quaternary Science Reviews, v. 10,
p. 291–310.
Bond, G., Showers, W., Cheseby, M., Lotti, R.,
deMenocal, P., Priori, P., Cullen, H., Hadjas,
I., and Bonani, G., 1997, A pervasive, millennial-scale cycle in the North Atlantic Holocene and late glacial climates: Science, v. 278,
p. 1257–1266.
Camburn, K.E., and Charles, D.F., 2000, Diatoms of
low-alkalinity lakes in the northeastern United
States: Philadelphia Academy of Natural Sciences Special Publication 18, 152 p.
Cole, J.J., Caraco, N.F., Kling, G.W., and Kratz,
T.K., 1994, Carbon dioxide supersaturation in
the surface waters of lakes: Science, v. 265,
p. 1568–1570.
Dahl-Jensen, D., Mosegaard, K., Gundestrup, N.,
Johnsen, S.J., Hansen, A.W., Clow, G.D., and
Balling, N., 1998, Past temperatures directly
from the Greenland Ice Sheet: Science, v. 282,
p. 268–271.
Davis, P.T., 1985, Neoglacial moraines on Baffin Island, in Andrews, J.T., ed., Quaternary environments: Eastern Canadian Arctic, Baffin
Bay and western Greenland: Boston, Allen
and Unwin, Ltd., p. 682–718.
Dixit, S.S., Smol, J.P., Charles, D.F., Hughes, R.M.,
Paulsen, S.G., and Collins, G.B., 1999, Assessing water quality changes in the lakes of
the northeastern United States using sediment
diatoms: Canadian Journal of Fisheries and
Aquatic Sciences, v. 56, p. 131–152.
Dyke, A.S., Dale, J.E., and McNeely, R.N., 1996,
Marine molluscs as indicators of environmental change in glaciated North America
and Greenland during the last 18 000 years:
Géographie Physique et Quaternaire, v. 50,
p. 125–184.
Glew, J.R., 1989, A new trigger mechanism for sediment samplers: Journal of Paleolimnology,
v. 2, p. 241–243.
Joynt, E.H., III, and Wolfe, A.P., 2001, Paleoenvironmental inference models from sediment
diatom assemblages in Baffin Island lakes
(Nunavut, Canada) and reconstruction of
summer water temperature: Canadian Journal
of Fisheries and Aquatic Sciences, v. 58,
p. 1222–1243.
Kling, G.W., Kipphut, G.W., and Miller, M.C.,
1992, The flux of CO2 and CH4 from lakes
and rivers in Arctic Alaska: Hydrobiologia,
v. 240, p. 23–36.
Koinig, K.A., Schmidt, R., Sommaruga-Wögrath,
S., Tessandri, R., and Psenner, R., 1998, Climate change as the primary cause for pH
shifts in a high alpine lake: Water, Air, and
Soil Pollution, v. 104, p. 167–180.
Miller, G.H., Mode, W.N., Wolfe, A.P., Sauer, P.E.,
Bennike, O., Forman, S.L., Short, S.K., and
Stafford, T.W., Jr., 1999, Stratified interglacial
lacustrine sediments from Baffin Island, Arctic
Canada: Chronology and paleoenvironmental
implications: Quaternary Science Reviews,
v. 18, p. 789–810.
Overpeck, J.T., Hughen, K.A., Hardy, D., Bradley,
R.S., Case, R., Douglas, M.S.V., Finney, B.,
Gajewski, K., Jacoby, G., Jennings, A.E., Lamoureux, S., Lasca, A., MacDonald, G.M.,
Moore, J.J., Retelle, M., Wolfe, A.P., and Zielinski, G., 1997, Arctic environmental change
of the last four centuries: Science, v. 278,
p. 1251–1256.
Paterson, W.S.B., Koerner, R.M., Fisher, D., Johnsen, S.J., Clausen, H.B., Dansgaard, W., Bucher, P., and Oeschger, H., 1977, An oxygen isotope climatic record from the Devon Ice Cap,
Arctic Canada: Nature, v. 266, p. 508–511.
Psenner, R., and Schmidt, R., 1992, Climate driven
pH control of remote alpine lakes and effects
of acid deposition: Nature, v. 356, p. 781–783.
Smol, J.P., and Cumming, B.F., 2000, Tracking longterm changes in climate using algal indicators
in lake sediments: Journal of Phycology,
v. 36, p. 986–1011.
Sommaruga-Wögrath, S., Koinig, K., Schmidt, R.,
Sommaruga, R., Tessandri, R., and Psenner,
R., 1997, Temperature effects on the acidity of
remote alpine lakes: Nature, v. 387, p. 64–67.
Steig, E.J., Wolfe, A.P., and Miller, G.H., 1998, Wisconsinan glacial refugia on eastern Baffin Island: Coupled evidence from cosmogenic isotopes and lake sediments: Geology, v. 26,
p. 835–838.
Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr,
G.S., Hughen, K.A., Kromer, B., McCormac,
F.G., van der Plicht, J., and Spurk, M., 1998,
INTCAL98 radiocarbon age calibration
24 000–0 cal BP: Radiocarbon, v. 40,
p. 1041–1083.
ter Braak, C.J.F, 1987, Ordination, in Jongman,
R.H.G, et al., eds., Data analysis in community and landscape ecology: Wageningen,
Center for Agricultural Publishing and Documentation, p. 91–173.
Wolfe, A.P., 1997, On diatom concentrations in lake
sediments: Results of an inter-laboratory comparison and other experiments performed on a
uniform sample: Journal of Paleolimnology,
v. 18, p. 61–66.
Wolfe, A.P., Fréchette, B., Richard, P.J.H., Miller,
G.H., and Forman, S.L., 2000, Paleoecological
assessment of a .90 000-year lacustrine sequence from Fog Lake, Baffin Island, Arctic
Canada: Quaternary Science Reviews, v. 19,
p. 1677–1699.
Zdanowicz, C.M., Zielinski, G.A., Wake, C.P., Fisher, D.A., and Koerner, R.M., 2000, A Holocene record of atmospheric dust deposition on
the Penny Ice Cap, Baffin Island, Canada:
Quaternary Research, v. 53, p. 62–69.
Manuscript received April 13, 2001
Revised manuscript received October 30, 2001
Manuscript accepted November 26, 2001
Printed in USA
GEOLOGY, March 2002