J Paleolimnol (2012) 48:55–67 DOI 10.1007/s10933-012-9603-8

ORIGINAL PAPER

Late Holocene change in climate and atmospheric circulation inferred from geochemical records at Kepler Lake, south-central Alaska A. W. Gonyo • Zicheng Yu • G. E. Bebout

Received: 4 March 2010 / Accepted: 15 March 2012 / Published online: 24 April 2012 Ó Springer Science+Business Media B.V. 2012

Abstract Climate records during the last millennium are essential in placing recent anthropogenic-induced climate change into the context of natural climatic variability. However, detailed records are still sparse in Alaska, and these records would help elucidate climate patterns and possible forcing mechanisms. Here we present a multiple-proxy sedimentary record from Kepler Lake in south-central Alaska to reconstruct climatic and environmental changes over the last 800 years. Two short cores (85 and 101 cm long) from this groundwater-fed marl lake provide a detailed stable isotope and sediment lithological record with chronology based on four AMS 14C dates on terrestrial macrofossils and 210Pb analysis. The d18O values of inorganic calcite (CaCO3) range from -17.0 to -15.7 %, with the highest values during the period of 1450–1850 AD, coeval with the well-documented Little Ice Age (LIA) cold interval in Alaska. The high d18O values during the cold LIA are interpreted as

This is one of 18 papers published in a special issue edited by Darrell Kaufman, and dedicated to reconstructing Holocene climate and environmental change from Arctic lake sediments.

Electronic supplementary material The online version of this article (doi:10.1007/s10933-012-9603-8) contains supplementary material, which is available to authorized users.

reflecting shifts in atmospheric circulation. A weakening of the wintertime Aleutian low pressure system residing over the Gulf of Alaska during the LIA would have resulted in 18O-enriched winter precipitation as well as a colder and possibly drier winter climate in south-central Alaska. Also, elevated calcite contents of [80 % during the LIA reflect a lowering of lake level and/or enhanced seasonality (warmer summer and colder winter), as calcite precipitation in freshwater lakes is primarily a function of peak summer temperature and water depth. This interpretation is also supported by high d13C values, likely reflecting high aquatic productivity or increased residence times of the lake water during lower lake levels. The lower lake levels and warmer summers would have increased evaporative enrichment in 18O, also contributing to the high d18O values during the LIA. Our results indicate that changes in atmospheric circulation were an important component of climate change during the last millennium, exerting strong influence on regional climate in Alaska and the Arctic. Keywords Alaska  Late Holocene  Last millennium  Climate change  Stable isotopes  Atmospheric circulation

Introduction A. W. Gonyo  Z. Yu (&)  G. E. Bebout Department of Earth and Environmental Sciences, Lehigh University, Bethlehem, PA 18015, USA e-mail: [email protected]

Over the last 50 years Alaska has warmed at more than twice the average rate compared with the rest of the

123

56

United States, with the greatest warming occurring during the winter. Documenting climate variability in the recent geological past will help us understand the mode and cause of ongoing regional climate change in Alaska and beyond. South-central Alaska is of particular interest because of its proximity to, and sensitivity to changes in, the Aleutian Low (AL), an atmospheric circulation system that is thought to be a key driver or modulator of regional winter climate change in the northeastern Pacific region (Overland et al. 1998). Previous studies in the Yukon Territory, Canada, have suggested that the position and intensity of the AL have fluctuated significantly over the last 7,500 years (Anderson et al. 2005, 2006). Studies of glacier dynamics using dendrochronology have indicated that the Little Ice Age (LIA) period in the northeastern Pacific region was characterized by lower temperatures and, in most cases, drier conditions (Wiles et al. 2004; Daigle and Kaufman 2009). If the strength and position of the AL fluctuated during the LIA, it provides a possible explanation for the observed changes in regional climate. Recent regional studies have demonstrated the spatial complexity of climate change, in particular moisture variability, during the late Holocene. The heterogeneity of moisture variability was observed at sites across southern and central Alaska, with those in southwestern and northern Alaska experiencing wetter conditions during the LIA (Clegg et al. 2005, 2010), and those in southeastern Alaska exhibiting drier conditions (Clegg et al. 2005; Tinner et al. 2008). In their study at Kepler Lake, south-central Alaska, Forester et al. (1989) found higher lake-water salinity during the LIA based on ostracode assemblage data, also suggesting drier conditions at this time. Significant changes in climate occurred regionally and globally at the end of the LIA. Changes in atmospheric circulation have been interpreted from changes in sea-salt chemical species in ice cores from both Greenland and Antarctica (Kreutz et al. 1997). Also, there are observed teleconnections between the AL, the Pacific Decadal Oscillation (PDO) and modes of high-latitude circulation such as the Arctic Oscillation (AO) (Overland et al. 1998). Based on instrumental records, during periods of strengthened PacificNorth American (PNA) pattern and PDO, the AL is weak, and conversely, the AL is stronger during periods of weakened PNA and PDO (Overland et al. 1998). Zhao and Moore (2006) identified a reduction in

123

J Paleolimnol (2012) 48:55–67

Himalayan snow accumulation since the 1840s and, through analysis of sea-level pressure data, suggested that this was due to a weakening of the trade winds over the Pacific. Hendy et al. (2002) cited a decrease in sea surface salinity occurring at 1870, from an isotopic analysis of several coral cores, suggesting that the LIA was associated with strengthened trade winds, increased evaporation in the western tropical Pacific, and a steeper latitudinal temperature gradient. Following the LIA, trade winds weakened causing a freshening of tropical Pacific waters (Hendy et al. 2002). In this study, we analyzed two sediment cores for calcite stable isotopes and sediment composition from Kepler Lake in south-central Alaska to assess climate variability during the last millennium. Modern lake water was also analyzed to inform and constrain interpretations of the sedimentary proxies, in particular the calcite d18O data. The objectives of this study were to (1) derive a new high-resolution multiple proxy sedimentary record for the late Holocene from a marl lake in south-central Alaska, and (2) investigate possible large-scale shifts in atmospheric circulation, especially the position and intensity of the AL, associated with climate change during the LIA in the northern Pacific and elsewhere.

Study site Kepler Lake (61°330 1000 N, 149°110 4500 W) is located at an elevation of 26 m above sea level in the Matanuska Valley (Fig. 1a). During the last glaciation, the valley was overlain by the Matanuska Glacier, which retreated from the lake about 14,000 cal years BP (Kopczynski 2008). The valley is blanketed by glacial deposits in a crevasse-ridge complex composed of cross-bedded and planar-bedded sandy gravel (Larson et al. 2003). Mean annual precipitation is 422 mm, and mean annual temperature 2.7 °C, based on instrumental data from Matanuska AES weather station located approximately 4 km northwest of the lake (Fig. 2). Kepler Lake is a hydrologically-open, groundwater-fed lake that precipitates marl (unconsolidated calcium carbonate deposits). Three interconnected lakes, Kepler, Matanuska, and Bradley, have identical surface water elevation, so it is likely that Kepler Lake occasionally experiences exchanges with Matanuska Lake through groundwater and with Bradley Lake through a surface

J Paleolimnol (2012) 48:55–67

57

Fig. 1 Locations and setting. a Shaded relief map showing the locations of study site Kepler Lake (black dot) and other paleoclimate sites (white dots): Mica Lake, Grizzly Lake, Hallett Lake, Moose Lake, Mt. Logan, and Jellybean Lake. Other place names include Denali National Park weather station (AK03) and Anchorage, AK. Inset shows study region in Alaska. b Topographic map of the Kepler Lake area (map from US Geological Survey). c Kepler Lake bathymetric map redrawn from Alaska Department of Fish and Game (http://www.adfg. alaska.gov) (depths in meters)

slack-water channel. However, the dominant input to the lake is from shallow groundwater to the north (Fig. 1b). The lake has a maximum water depth of 22.5 m (Fig. 1c). Forester et al. (1989) analyzed two cores from Kepler Lake for ostracodes and pollen. The ostracode

assemblages display a decrease in the warm-water taxon Cyclocypris ovum and a decrease in total adult ostracode valves during the presumed LIA interval, which were interpreted to reflect lower water temperatures. However, the cold-water taxon Candona protzi is absent during the LIA, possibly indicating elevated

123

58

J Paleolimnol (2012) 48:55–67

lakes, wetlands, and rainwater in south-central Alaska (Table 1). Water samples were collected in 30 mL Nalgene bottles, and bottles were sealed tightly with caps underwater to remove air from the bottles. Two water samples from the 2008 collection were taken for dissolved inorganic carbon (DIC) analyses. These water samples were filtered through a 0.45 lm filter, placed inside 25 mL brown glass bottles with septum caps, and treated with CuSO4 to kill any microorganisms. Analyses of water samples were performed at the University of Arizona Isotope Lab. Elevation, pH, conductivity, and water temperature data were also recorded in the field using a GPS receiver and an Oakton pH and conductivity meter. Core collection and sampling

Fig. 2 Climate in south-central Alaska. a Precipitation (P) and temperature (T) data from the 1971–2000 climate normals for the Mt. McKinley weather station and isotope data from AK03 USNIP station in Denali National Park from 1989 to 1993 (IAEA/WMO 2001). No data available for April. The two stations are *1 km apart. b Precipitation and temperature data from the 1971–2000 climate normals for Matanuska AES station, located *1 km from Kepler Lake

solute concentration within the lake above the levels tolerated by this taxon. The pollen record varied little over the past 800 years, indicating either that the surrounding terrestrial vegetation was relatively insensitive to regional climate change, or that the sampling resolution was too coarse to detect short-term variations. The study by Forester et al. (1989) provides a record of major temperature and moisture changes throughout the last 2,500 years, but is limited by coarse age control, making it difficult to compare their results with regional climate records. We build on this study by analyzing new sediment cores from this lake, providing stronger age control, and augmenting the climatic interpretation with stable isotope proxies.

Methods Lake water sampling and isotope analysis A total of 20 water samples for water isotopic analyses were collected in August 2007 and July 2008 from

123

On 17 August 2007, two short surface cores (85-cmlong KL07-1, and 101-cm-long KL07-2) were collected from *4.5 m water depth on a boat using a plastic tube fit with a piston. The cores were kept upright and sampled at 1 cm intervals in the field using a Glew extrusion system (Glew et al. 2001) in order to preserve the sediment–water interface. Dating analysis and age models For 210Pb analyses, the sediments from the upper 30 cm of both cores were analyzed at MyCore, Inc., in Ontario, Canada. Using alpha spectrometry methods as described in Cornett et al. (1984), ten samples from core KL07-1 were analyzed from 0 to 15 cm, and ten samples from core KL07-2 were analyzed from 0 to 29 cm. The constant-rate-of-supply model was used for interpreting the 210Pb data (Appleby and Oldfield 1978). Radiocarbon dating was performed on five samples of terrestrial plant macrofossils from the two cores to establish the chronology of the older sediment sections (Table 2). The samples were sent to the Keck Carbon Cycle Accelerator Mass Spectrometry (AMS) Laboratory at the University of California-Irvine for 14C analysis. 14C dates were calibrated using the program CALIB 5.0.2 (Stuiver and Reimer 1993) with the INTCAL04 calibration data set (Reimer et al. 2004). The median of the probability density distribution of the calibrated age was used in all age model calculations (Table 2). Age models were established using a linear interpolation between the 14C ages for KL07-1

Long Lake

8/18/2007

Kepler Lake

Kepler Lake Rain

Hamilton Well

7/27/2008

7/27/2008

8/22/2007

Kepler Lake Spring 08

Anchorage Rain

8/21/2007

7/27/2008

Wasilla Lake

8/19/2007

7/27/2008

Hundred Mile Lake

Matanuska Glacier State Park

8/18/2007

Long Lake (Mile 86)

Irene Lake

8/18/2007

Canoe Lake

8/18/2007

8/18/2007

Crofby well

Echo Lake Kepler Lake Spring

8/18/2007 8/18/2007

Fish Lake

Matanuska Lake

8/18/2007

8/18/2007

Lake

Edmonds Lake

8/18/2007

8/18/2007

Lake Wetland

Mirror Lake

8/18/2007

Well

Rain

Wetland

Lake

Rain

Lake

Well

Lake

Lake

Lake

Well

Lake

Lake

Lake

Lake

Lake

Lake

Kepler Lake

8/17/2007

Water type

Site name

Collection date

61.663

61.555

61.555

61.555

61.218

61.587

61.8

61.806

61.805

61.785

61.558

61.561

61.559

61.557

61.552 61.556

61.553

61.434

61.428

61.555

Latitude (°N)

149.316

149.205

149.205

149.205

149.874

149.394

147.815

147.846

148.236

148.562

149.186

149.201

149.194

149.196

149.208 149.206

149.228

149.394

149.414

149.205

Longitude (°W)

Table 1 Information for water isotope samples from south-central Alaska

176

40

40

40

44

96

549

543

485

310

77

50

47

44

30 36

26

100

103

40

Elevation (m asl)

-16.7

-19.4

-18.2

-16.8

-13.9

-7.6

-18.9

-15.4

-14.0

-16.7

-19.1

-13.4

-13.0

-15.0

-14.3 -17.6

-12.7

-13.3

-13.2

-16.5

d18O (%, VSMOW)

-129.9

-166.7

-144.7

-136.1

-117.7

-77.2

-152.5

-138.6

-124.6

-135.2

-147.2

-114.6

-114.4

-126.1

-122.0 -137.6

-110.8

-112.0

-113.4

-131.2

dD (%, VSMOW)

-9.0

-9.5

d13CDIC (%, VPDB)

7.95

8.16

8.21

8.66

8.5

7.3

8.22

8.25

8.2

8.17 7.36

8.36

8.99

8.54

7.83

pH

219

553

419

490

321

557

353

326

396

343 510

235

190.6

239

418

Conductivity (lS/cm)

17.9

6.4

19.4

17.1

19.4

19.3

19.2

19.3

18.8 19.1

19.1

18.9

19.1

19.2

Temp. (°C)

J Paleolimnol (2012) 48:55–67 59

123

60

J Paleolimnol (2012) 48:55–67

Table 2 AMS radiocarbon (14C) dates from cores KL07-1 and KL07-2, Kepler Lake, south-central Alaska Core

Depth (cm)

Material dated

AMS lab no.a

14

C dates (year BP ± SE)

KL07-1

45–46

Leaf fragments, birch scales

42048

190 ± 15

KL07-1

54–55

Leaf fragments

42049

-520 ± 20

KL07-1

66–67

Leaf fragments, birch scales

42050

340 ± 20

Calibrated age (2r range in year AD)b

Median probability age (AD)

Sample size (mg C)

1663–1951

1775

0.18

1565

0.14

Rejected (modern carbon) 1473–1635

0.05

KL07-1

84–85

Leaf stem, birch scales

42051

540 ± 20

1324–1431

1408

0.10

KL07-2

99–100

Birch seeds, Birch scales

54644

790 ± 15

1220–1267

1243

0.13

a

Samples analyzed by the Keck Carbon Cycle Accelerator Mass Spectrometry Laboratory at the University of California Irvine

b

Calibrated by the program CALIB Rev 5.0.2 (Stuiver and Reimer 1993) using INTCAL04 data set (Reimer et al. 2004)

and the sediment water interface as 2007 AD. Ages were transferred to core KL07-2 on the basis of distinct stratigraphic intervals through visual comparison of loss-on-ignition results.

ratio mass spectrometer. Oxygen and carbon isotope ratios of calcite (18O/16O, 13C/12C) are reported relative to the Vienna Pee Dee Belemnite (VPDB) standard. The results were presented in d notation, which is defined as:

Loss-on-ignition analysis

d ¼ ðRsample =Rstandard  1Þ  1000 ð&Þ

Samples of 0.7 mL wet sediment were taken at 1 cm intervals for loss-on-ignition (LOI) analyses (Dean 1974). The samples were dried in an oven overnight at 100 °C to determine moisture content from the weight loss. The samples were then combusted at 550 °C for 1 h to estimate the organic-matter (OM) content, and then at 1,000 °C for 1 h to estimate the amount of carbonate (CaCO3). CO2 evolves from carbonate minerals beginning at *800 °C and should be completely evolved by 850 °C (Dean 1974).

where R is the molar ratio of the heavy isotope over the light isotope, 18O/16O or 13C/12C. Analyses of lab standards, in both laboratories, show an analytical precision better than 0.3 % for d18O and 0.2 % for d13C (2r for both).

Stable isotope (d18O and d13C) analysis of bulk calcite Samples were taken from the two cores at 1 cm intervals and analyzed for oxygen and carbon isotopes. Bulk samples were freeze-dried, and shell and plant fragments were removed manually under a stereo microscope and discarded. For core KL07-1, the treated samples of 4.7–8.8 mg were reacted overnight at 25 °C with 100 % phosphoric acid. The CO2 released by this reaction was cryogenically purified and measured in dual-inlet mode using a Finnigan MAT 252 mass spectrometer at Lehigh University. Samples from core KL07-2 were analyzed at the University of Wyoming stable isotope laboratory using a Finnigan gas bench coupled to a Thermo Finnigan DeltaPLUS XP continuous flow stable isotope

123

Results1 Lake water isotopes Water isotopes from 13 lake samples collected in this study from the Kepler Lake region define a local evaporation line (LEL; Table 1, Fig. 3). Precipitation isotope data from Denali National Park weather station (AK03; IAEA/WMO 2001) and our own rain sample measurements plot near the global meteoric water line (GMWL; Fig. 3). The Kepler Lake water samples also plotted near the GMWL and are located near the intercept of the LEL and GMWL, indicating that Kepler Lake is relatively insensitive to evaporative effects (Craig 1961).

1

All of the data from Kepler Lake (Alaska) presented in this study are available on-line through the World Data Center for Paleoclimatology (http://www.ncdc.noaa.gov/paleo/pubs/ jopl2012arctic/jopl2012arctic.html).

J Paleolimnol (2012) 48:55–67 Table 3 210Pb activity and dates from cores KL07-1 and KL07-2, Kepler Lake, south-central Alaska

61

Core

Depth (cm)

210

210

Unable to reach background

Pb activity (dpm/g)

Pb date (AD)

1r error (years)

KL07-1

0–1

2.52

KL07-1

1–2

2.43

KL07-1

2–3

2.13

KL07-1

3–4

1.62

KL07-1

4–5

1.86

KL07-1

6–7

1.48

KL07-1

8–9

1.14

KL07-1 KL07-1

11–12 13–14

2.10 1.66

KL07-1

15–16

2.68

KL07-2

0–1

3.90

2007.6

0

KL07-2

2–3

3.76

2002.6

0.7

KL07-2

5–6

3.46

1995.4

1.4

KL07-2

8–9

3.09

1986.9

3.0

KL07-2

11–12

2.91

1974.7

4.4

KL07-2

14–15

2.31

1958.8

9.2

KL07-2

19–20

1.41

1915.6

37.2

KL07-2

22–23

1.10

Estimated background 0.985 dpm/g

KL07-2

24–25

0.96

KL07-2

29–30

0.90

Fig. 3 Isotope plot of water samples from the Kepler Lake region. AK03 precipitation data from USNIP, part of the IAEA/ WMO (2001) and other lake data (open circles) from Schiff et al. (2009). The global meteoric water line (GMWL: d2H = 8d18O ? 10) and local evaporation line (LEL) are labeled

small size (Table 2). 210Pb results from KL07-1 yielded anomalously low concentrations ranging from 1.14 to 2.68 dpm/g. Our 210Pb analysis was unable to reach background level and includes several reversals (that is, higher concentrations at increased depth). This indicates possible disturbance within the upper section of the core and, as such, the 210Pb chronology cannot be constructed for core KL07-1 (Table 3). One AMS 14C date from core KL07-2 produced a calibrated age of 1240 AD. 210Pb dates from KL07-2 on the basis of the constant-rate-of-supply model provided a chronology for the last 94 years (Table 3; Fig. 4b; Cornett et al. 1984). Chronology for core KL07-2 was generated using a linear interpolation between the 210Pb dates, one calibrated radiocarbon age at the base of the core (Fig. 4), and ages of four tie points based on correlations of LOI results with ones from core KL07-1 (Fig. 5; next subsection).

AMS 14C and 210Pb Loss-on-ignition Three out of four calibrated AMS 14C ages from core KL07-1 were used to establish chronology since 1400 AD. One sample from KL07-1 (at 55 cm) was rejected as it produced a post-bomb age, likely caused by contamination by modern carbon due to its extremely

In core KL07-1, OM values range from 3.1 to 6.4 % by weight (mean = 4.4 %), CaCO3 values range from 68.4 to 88.2 % (mean = 81 %), and residual silicate values range from 8.6 to 26.0 % (mean = 14.6 %)

123

62

Fig. 4 Age models for Kepler Lake cores. a Core KL07-1, based on three calibrated AMS 14C ages. b Core KL07-2, based on one calibrated AMS 14C age, four tie points between the two cores on the basis of loss-on-ignition results (Fig. 5), and a 210Pb

J Paleolimnol (2012) 48:55–67

profile from 10 analyses using the constant-rate-of-supply model. Error bars are the 2r ranges, and the sediment–water interface is 2007 AD. Both age models constructed using a linear interpolation

72.4 %, displaying an increase of *20 % from 1600 to 1950 AD, and silicate values range 8.6 to 26.0 %, with a mean of 14.6 % (Figs. 5, 6). Possible causes for the observed differences in two cores include different sedimentation rates between the two coring sites, disturbance possibly at core KL07-1, and loss of sediment during coring. Calcite d18O and d13C

Fig. 5 Loss-on-ignition results from sediment cores from Kepler Lake plotted on depth scale, also showing four correlation tie points for assigning ages to core KL07-2

(Figs. 5, 6). In core KL07-2, OM values (by weight) range from 5.0 to 9.4 %, with a mean of 9.4 %, CaCO3 values range from 55.1 to 82.3 %, with a mean of Fig. 6 Loss-on-ignition and isotopes results from sediment cores from Kepler Lake on age scale. d18O and d13C values are relative to VPDB

123

In core KL07-1, d18O values range from -17.0 to -15.8 % and are elevated by between 0.5 and 1.0 % during the LIA (1450–1850 AD). The d13C values range from -8.8 to -6.9 % and display a positive 1.5 % shift from -8.0 to -6.5 % during the period of 1600–1870 AD, after which they remain relatively constant at -8.0 % (Fig. 6). d18O values in core KL07-2 range from -17.4 to -15.2 % and increase for approximately 1.0 % during the period of

J Paleolimnol (2012) 48:55–67

1300–1700 AD. After 1700 AD, values decline by approximately 0.5 %, and then remain steady until the present. d13C values range from -9.6 to -6.1 % for the period of 1600–1870 AD, but then display a negative 1.5 % shift at 1870 AD (Fig. 6). In general, the d18O values from both cores show similar first order pattern, with higher values during the LIA.

Discussion Oxygen isotopes as an indicator of atmospheric circulation change To understand the first-order long-term change of 0.6–1.0 % increase in d18O values during the LIA (1450–1840 AD), we need to consider the relevant controls on lake water and carbonate oxygen isotopes, namely the lake water d18O values, and temperature of the lake water. The d18O of lake water is directly affected by the d18O of meteoric precipitation, the proportion and d18O value of groundwater input, and evaporative enrichment. In a hydrologically-closed lake, d18O values can be strongly modified by evaporation due to long water residence time, leading to an enrichment of the heavier 18O isotope in the lake water. Water samples from Kepler Lake taken during peak evaporation season in August 2007 and July 2008 plot very near the GMWL and near local groundwater values on dD versus d18O plot (Fig. 3), indicating minimal evaporative enrichment in 18O in this lake at that time. This is likely caused by a relatively short residence time, due to the large input of shallow groundwater with d18O that reflects long-term averaging of the d18O of the precipitation (Fig. 3). The groundwater input to Kepler Lake is likely shallow based on the similar d18O values for groundwater and the lake water (-17.6 and -16.5 %, respectively). It is possible that lake hydrology has changed in the past, but we argue it is unlikely that such a change is a dominant control of the observed d18O changes in the last 800 years. Lake water isotopes can also be affected by air temperature, through its influence on meteoric precipitation d18O with a relationship of 0.6 % per °C (Dansgaard 1964). The temperature-dependent O-isotope fractionation between calcite and water is -0.24 % per °C (Friedman and O’Neil 1977; Leng and Marshall 2004). Water temperature in the

63

littoral zones and at the surface of lakes is primarily controlled by air temperature; as such, calcite d18O has often been used as a proxy of changes in air temperature. The relevant literature for this region indicates that the LIA was a period of decreased temperature based on evidence including narrow tree-ring widths, glacial advance, fossil midge assemblages, and geochemical records (Wiles et al. 2004; Daigle and Kaufman 2009; Clegg et al. 2010; Hu et al. 2001). Recent synthesis of 23 proxy temperature records indicates that Arctic temperature shows a gradual cooling trend over the last 2,000 years until approximately 1900 AD (Kaufman et al. 2009; Fig. 6). Lake water temperature closely tracks air temperature in shallow littoral zones, with the strongest relationship between air and water temperature occurring during July (Livingstone and Lotter 1998). The cooler temperatures during the LIA would be expected to cause lower d18O values; thus temperature cannot be the primary factor for the positive shift in d18O observed during the LIA (Fig. 7). If the effects of evaporation and temperature are not dominant at Kepler Lake, a change in moisture source, resulting from a shift in atmospheric circulation, is more likely the primary control of d18O at Kepler Lake. The Aleutian Low (AL), a lowpressure system residing in the Gulf of Alaska during the winter season, is the dominant circulation control in south-central Alaska (Overland et al. 1998). Fisher et al. (2004) used simulations to explain how changes in moisture source controlled by the size and position of the AL could affect the d18O in precipitation at Mt. Logan and Jellybean Lake, both about 500 km east of Kepler Lake (Fig. 1). The CaCO3 at Jellybean Lake and glacier ice at Mt. Logan both display shifts to more positive d18O values during the LIA (Fig. 7), which has been interpreted as decreased isotopic distillation of air masses during an interval of weakened AL (Anderson et al. 2005). These results are consistent with those from Mica Lake where Schiff et al. (2009) interpreted the negative shift in d18O values during the LIA as a more zonal flow regime, also indicative of a weakened AL. This is thought to be due to the coastal location of Mica Lake, which during zonal flow regimes receives moisture experiencing larger amounts of rainout as the storm trajectories pass over south-western Alaska. During meriodinal flow

123

64

J Paleolimnol (2012) 48:55–67

Fig. 7 Regional correlation of d18O and temperature records. a Kepler Lake d18O record from bulk calcite b Jellybean Lake d18O record from bulk calcite (Anderson et al. 2005). c Mica Lake d18O record from diatom silica (Schiff et al. 2009).

d Mt. Logan d18O ice core with a 5-point moving average applied (Fisher et al. 2008), and e Arctic temperature synthesis (Kaufman et al. 2009)

regime, Mica Lake receives moisture experiencing smaller amounts of rainout due to its location in the Gulf of Alaska (Schiff et al. 2009). For Kepler Lake, the LIA period is also interpreted as a period of enhanced zonal flow due to a weakened and/or westward AL. We suggest that enhanced zonal flow would have caused the d18O of the precipitation at Kepler Lake to shift to more positive values caused by a decrease in isotopic distillation when the AL was weakened. Yu et al. (2008) used the observed spatial pattern of precipitation isotope values in western Alaska and southern Yukon to interpret a large isotopic shift during the early Holocene at Hundred Mile Lake in the Matanuska Valley as caused by change in moisture source. The AL is a principally a winter phenomenon, and winter precipitation accounts for only about 25 % of the annual precipitation in the Matanuska Valley. However, it is possible that the groundwater and lake are mostly and selectively recharged by winter snowfall during peak snowmelt and discharge in the spring. The dominant winter precipitation recharge of groundwater and lake water balance has been documented elsewhere (e.g. Shuman and Donnelly 2006). Thus, the lake water budget and water isotopic signatures are dominated by spring snowmelt and winter precipitation, likely reflecting the dynamics of the AL.

Possible changes in effective moisture and seasonality during the LIA

123

The LIA in Alaska is considered to be a period of lower temperatures, as documented from lake, treering, and glacier records. However, the record from Kepler Lake displays an increase in CaCO3 content and d13C during the LIA. CaCO3 production is a function of several factors: peak summer temperature, Ca2? input from groundwater into the lake, and biological productivity causing calcite supersaturation through the drawdown of CO2 during photosynthesis (Kalff 2002). d13C values depend on (1) groundwater input that affects the d13C of the lake DIC, (2) in-lake productivity, and (3) the mass balance of terrestrial versus aquatic organic matter deposited in the lake (Duston et al. 1986; Langmuir 1997; Kalff 2002; Leng and Marshall 2004). Carbonate mineral solubility decreases with decreasing pCO2, increasing temperature, and increasing pH (Langmuir 1997). One possible explanation for the observed shifts in carbonate content and d13C values is an increase in groundwater input to Kepler Lake; however, the evidence from Kepler Lake does not support this hypothesis. First, increases in groundwater input to the lake system would cause d13C values to shift toward DIC values of -9.0 to -9.5 % (Table 1). Second, an

J Paleolimnol (2012) 48:55–67

increase in groundwater input would cause a corresponding shift in d18O to more negative values, based on the indication from the water isotopic dataset that groundwater d18O values are more negative than those of the lake water. However, during this period, there is no corresponding shift in the d18O data for Kepler Lake, and values during the LIA period (1450–1840 AD) are more positive, rather than more negative. The second possible explanation for the observed shifts at Kepler Lake is a change in primary productivity, thermal stratification of lake water, and CO2 drawdown in the water column due in part to a change in temperature. Under conditions of increased primary productivity, CO2 is removed and pH increases, causing lake water to become supersaturated with respect to calcite. The photosynthetic activity within the lake preferentially removes the lighter (12C) isotope, causing a positive shift in the d13C values. Conditions of enhanced thermal stratification within the water column create conditions of greater CO2 depletion during periods of enhanced primary productivity, shifting the water toward calcite saturation and leading to depletion of 12C within the epilimnion. Temperature is closely linked to the aforementioned controls of freshwater calcite precipitation: during higher temperature periods, calcite solubility decreases, causing water to become saturated with calcite. Periods of higher temperature also typically cause higher primary productivity and enhanced thermal stratification, making the direct effect of an increase in temperature on carbonate precipitation difficult to measure. The two most plausible explanations for the changes observed at Kepler Lake during the LIA are a change in seasonality characterized by a shorter growing season but higher peak summer temperatures, and a change in effective moisture to drier conditions. Enhanced seasonality during the LIA would be expected to cause an increase in CaCO3 content and more positive d13C values. Shorter but more intense summers, rather than a decrease in summer temperature, would lead to a greater number of days during which lake waters reached calcite supersaturation due to photosynthetic drawdown of CO2, increasing CaCO3 content and producing more positive values of d13C (Fig. 7). In this scenario, the observed narrow tree-ring widths and glacial advances during the LIA could be attributed to a shorter growing/ablation season. Following the LIA, CaCO3 content in the

65

sediments returned to values similar to those prior to the LIA, at *65 %, and calcite d13C values declined to ca. -9 %. These calcite d13C values are similar to values of -9.5 to -9.0 % measured for d13CDIC at Kepler Lake, indicating a return to a condition of lower primary productivity (Fig. 7). The proposed change in seasonality could be linked to the shift in the strength and/or position of the AL discussed earlier, as the onset of the LIA at *1600 AD matches well with the maximum glacial extent of the LIA at Goat Lake, located in the Kenai Mountains north of the Gulf of Alaska, at 1660 AD (Daigle and Kaufman 2009). Further research into the effects of the AL on seasonal temperature and precipitation in south-central Alaska is required to fully examine this question. In the second scenario, reductions in effective moisture similar to those noted in other studies cause lowering of lake levels (Clegg et al. 2005; Tinner et al. 2008). A lower lake level would be expected to cause an increase in CaCO3 content owing to the increased production in shallower waters. Lowered lake levels would also lead to increase in d13C, due to longer residence times of the lake water and higher d13C of the DIC, resulting from enhanced exchange with atmospheric CO2 that is comparatively enriched in 13C (Fig. 7). Changes in effective moisture would also be expected to cause a positive shift in d18O, as evaporative enrichment of lake water leads to higher values (Fig. 7). This hypothesis is supported by an earlier study conducted on Kepler Lake that noted a decrease or absence, during the colder LIA period, of an ostracode taxon (Candona protzi) that requires cold conditions but cannot tolerate increased salinity levels (Forester et al. 1989). Although the exact mechanism responsible for the observed patterns in our proxy data sets remains elusive, both of the scenarios (seasonality and effective moisture) discussed here are consistent with expected responses to a westerly/weakened AL, underscoring its importance to regional climate in south-central Alaska. Conclusions (1)

The calcite d18O record from Kepler Lake in south-central Alaska exhibits elevated values during the LIA, suggesting a weakened and/or westward shift of the Aleutian Low and enhanced zonal flow or decreased distillation of

123

66

(2)

(3)

J Paleolimnol (2012) 48:55–67

air masses. This conclusion is consistent with other records of isotopic and climatic change in the region. During the LIA in south-central Alaska, effective moisture was likely lower, with higher peak summer temperatures and enhanced seasonality (shorter summers with greater high temperature extremes). Both phenomena are consistent with the records of narrow tree-ring widths and glacial advance during the LIA. The changes in effective moisture or seasonality during the LIA directly affected the aquatic productivity at Kepler Lake, as indicated by increases in calcite d13C values and in carbonate sedimentation. These changes are most likely due to changes in the position (westward) or intensity (weak) of the AL during the LIA. Our study at Kepler Lake demonstrates the importance of using multiple proxies to fully understand the inherently complex nature of climatic change and forcing mechanisms.

Acknowledgments We thank Bob Booth for his field coring assistance and discussion; Kristi Wallace of the USGS Volcano Observatory in Anchorage, Alaska for logistical support and boat access; Sarah Kopcyznski for discussion on the glacial history of the study region; and two anonymous reviewers and Darrell Kaufman for constructive comments and suggestions. We also acknowledge UC-Irvine Keck AMS lab and MyCore Lab for dating analysis. Funding for this research was provided by NSF grants (ATM-0628455; EAR-0711355), and a graduate research grant from the Department of Earth and Environmental Sciences at Lehigh University.

References Anderson L, Abbott MB, Finney BP, Burns SJ (2005) Regional atmospheric circulation change in the North Pacific during the Holocene inferred from lacustrine carbonate oxygen isotopes, Yukon Territory, Canada. Quat Res 64:21–35 Anderson L, Abbott MB, Finney BP, Burns SJ (2006) Late Holocene moisture balance variability in the southwest Yukon Territory, Canada. Quat Sci Rev 26:130–141 Appleby PG, Oldfield F (1978) The calculation of 210Pb dates assuming a constant rate of supply of unsupported 210Pb to the sediment. Catena 5:1–8 Clegg BF, Tinner W, Henderson A, Bigler C, Hu FS (2005) Spatial manifestation of the Little Ice Age in Alaska. Eos Transactions of the American Geophysical Union 86, abstract PP31A-1513 Clegg BF, Clark GH, Chipman ML, Chou M, Ian WR, Tinner W, Hu FS (2010) Six millennia of summer temperature

123

variation based on midge analysis of lake sediments from Alaska. Quat Sci Rev 29:3308–3316 Cornett RJ, Chant LA, Link D (1984) Sedimentation of 210Pb in Laurentian shield lakes. Water Poll Res J Can 19:97–109 Craig H (1961) Isotopic variations in meteoric waters. Science 133:1702–1703. doi:10.1126/science/133.3465.1702 Daigle TA, Kaufman DS (2009) Holocene climate inferred from glacier extent, lake sediment and tree rings at Goat Lake, Kenai Mountains, Alaska, USA. J Quat Sci 24:33–45 Dansgaard W (1964) Stable isotopes in precipitation. Tellus 16:436–468 Dean WE (1974) Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: comparison with other methods. J Sed Petrol 44:242–248 Duston NM, Owen RM, Wilkinson BH (1986) Water chemistry and sedimentological observations in Littlefield Lake, Michigan: implications for lacustrine marl deposition. Environ Geol Water Sci 8(4):229–236 Fisher DA, Wake C, Kreutz K, Yalcin K, Steig E, Mayewski P, Anderson L, Zheng J, Rupper S, Zdanowicz C, Demuth M, Waszkiewizc M, Dahl-Jensen D, Goto-Azuma K, Bourgeois JB, Koerner RM, Sekerka J, Osterberg E, Abbott MB, Finney BP, Burns SJ (2004) Stable isotope records from Mount Logan, Eclipse ice cores and nearby Jellybean Lake. Water cycle of the North Pacific over 2000 years and over five vertical kilometres: sudden shifts and tropical connections. Geographie physique et Quaternaire 58:9033– 9048 Fisher D, Osterberg E, Dyke A, Dahl-Jensen D, Demuth M, Zdanowicz C, Bourgeois J, Koerner RM, Mayewski P, Wake C, Kreutz K, Steig E, Zheng J, Yalcin K, GotoAzuma K, Luckman B, Rupper S (2008) The Mt Logan Holocene—late Wisconsinan isotope record: tropical Pacific—Yukon connections. The Holocene 18:667– 677 Forester RM, Delorme DL, Ager TA (1989) A lacustrine record of late Holocene climate change from south-central Alaska. Geophysical Monograph Series 55, American Geophysical Union, pp 33–40 Friedman I, O’Neil JR (1977) Stable isotope fractionation factors of geochemical interest. In: Friedman I, O’Neil JR (eds), Data of geochemistry, 6th edn. Geological Survey Professional Paper, United States Government Printing Office Glew JR, Smol JP, Last WM (2001) Sediment core collection and extrusion. In: Last WM, Smol JP (eds), Tracking environmental change using lake sediments, vol 1. Basin analysis, coring, and chronological techniques. Kluwer, Dordrecht, pp 73–105 Hendy EJ, Gagan MK, Alibert CA, McCulloch MT, Lough JM, Isdale PJ (2002) Abrupt decrease in tropical Pacific sea surface salinity at end of Little Ice Age. Science 295:1511– 1514 Hu FS, Ito E, Brown TA, Curry BB, Engstrom DR (2001) Pronounced climatic variations in Alaska during the last two millennia. Proc Nat Acad Sci 98:10552–10556 IAEA/WMO (2001) United States network for isotopes in precipitation (USNIP) as part of the global network of isotopes in precipitation (GNIP) and isotope hydrology information system (ISOHIS). International Atomic Energy Agency

J Paleolimnol (2012) 48:55–67 and World Meteorological Organization. http://www.nrel. colostate.edu/projects/usnip/ Kalff J (2002) Inorganic carbon and pH. In: Limnology. Prentice Hall Publishing, New Jersey, pp 222–223 Kaufman DS, Schneider DP, McKay NP, Ammann CM, Bradley RS, Briffa KR, Miller GH, Otto-Bliesner BL, Overpeck JT, Vinther BM (2009) Arctic Lakes 2 k Project Members (Abbott M, Axford Y, Bird B, Birks HJB, Bjune AE, Briner J, Cook T, Chipman M, Francus P, Gajewski K, Geirsdottir A, Hu FS, Kutchko B, Lameoureux S, Loso M, MacDonald G, Peros M, Porinchu D, Schiff C, Seppa H, Thomas E) Recent warming reverses long-term Arctic cooling. Science 325:1236–1239 Kopczynski SE (2008) Remote Sensing of alpine glacial snow melt flooding, and radiometric constraints on the final ice collapse chronology of the Upper Cook Inlet, Alaska. PhD dissertation, Lehigh University, 2008 Kreutz KJ, Mayewski PA, Meeker LD, Twickler MS, Whitlow SI, Pittalwala II (1997) Bipolar changes in atmospheric circulation during the Little Ice Age. Science 277:1294– 1296 Langmuir D (1997) Carbonate chemistry. In: McConnin R (eds), Aqueous environmental geochemistry. Prentice-Hall, New Jersey Larson GJ, Evenson EB, Lawson DE, Ensminger SL, Baker G, Alley RB (2003) Glacial geology of upper Cook Inlet, Matanuska Glacier and Denali Highway, Alaska. In: Easterbrook DJ (eds), Quaternary geology of the United States, INQUA 2003 Field Guide Volume, Desert Research Institute, Reno, NV. pp 245–264 Leng MJ, Marshall JD (2004) Paleoclimate interpretation of stable isotope data from lake sediment archives. Quat Sci Rev 23:811–831 Livingstone DM, Lotter AF (1998) The relationship between air and water temperatures in lakes of the Swiss Plateau: a case study with palaeolimnological implications. J Paleolimnol 19:181–198

67 Overland JE, Adams JM, Bond NA (1998) Decadal variability of the Aleutian Low and its relation to high-latitude circulation. J Clim 12:1542–1548 Reimer PJ, Baillie MGL, Bard E, Bayliss A, Beck JW, Bertrand CJH, Blackwell PG, Buck CE, Burr GS, Cutler KB, Damon PE, Edwards RL, Fairbanks RG, Friedrich M, Guilderson TP, Hogg AG, Hughen KA, Kromer B, McCormac FG, Manning SW, Ramsey CB, Reimer RW, Remmele S, Southon JR, Stuiver M, Talamo S, Taylor FW, van der Plicht J, Weyhenmeyer CE (2004) IntCal04 terrestrial radiocarbon age calibration, 26–0 ka BP. Radiocarbon 46:1029–1058 Schiff CJ, Kaufman DS, Wolfe AP, Dodd J, Sharp Z (2009) Late Holocene storm-trajectory changes inferred from the oxygen isotope composition of lake diatoms, south Alaska. J Paleolimnol 41:189–208 Shuman B, Donnelly JP (2006) The influence of seasonal precipitation and temperature regimes on lake levels in the northeastern United States during the Holocene. Quat Res 65:44–56 Stuiver M, Reimer PJ (1993) Extended 14C database and revised CALIB radiocarbon calibration program. Radiocarbon 35:215–230 Tinner W, Bigler C, Gedye S, Gregory-Eaves I, Jones R, Kaltendrieder P, Krahenbuhl U, Hu FS (2008) A 700-year paleoecological record of ecosystem responses to climatic variation from Alaska. Ecology 89:729–743 Wiles GC, D’Arrigo RD, Villalba R, Calkin PE, Barclay DJ (2004) Century-scale solar variability and Alaskan temperature change over the past millennium. Geophys Res Lett 31:L15203 Yu ZC, Walker KN, Evenson EB, Hadjas I (2008) Late glacial and early Holocene climate oscillations in the Matanuska Valley, south-central Alaska. Quat Sci Rev 27:148–161 Zhao H, Moore GWK (2006) Reduction in Himalayan snow accumulation and weakening of the trade winds over the Pacific since the 1840s. Geophysl Res Lett 33:L17709

123