Holocene climatic change in the Alaskan Arctic as inferred from oxygen-isotope and lake-sediment analyses at Wahoo Lake

Abstract

Despite recent progress in understanding high-latitude climate variability, paleoclimate records are scarce from the Alaskan Arctic. We conducted isotopic and sediment analyses at Wahoo Lake to infer Holocene climate variability in northeastern Alaska. Water δ¹⁸O and δD values from the lake and its inlet/outlet streams suggest that winter precipitation dominates modern water inputs and that evaporation has limited influence on the lake’s hydrological budget. The isotopic composition of Pisidium exhibits marked variations during the past 11,500 years, with δ¹⁸O ranging between −18.7‰ and −16.2‰ and δ¹³C between −7.1‰ and −2.3‰ (Vienna Pee Dee Belemnite (VPDB)). Elevated δ¹⁸O and sediment composition from 11.5 to 8.9 kcal. BP suggest evaporative ¹⁸O enrichment and arid conditions. Rising lake levels are evidenced by the disappearance of Pisidium and a transition to low-carbonate gyttja ca. 6.3 kcal. BP and by the onset of sediment deposition on an adjacent shelf by 5.3 kcal. BP. These changes coincided with enhanced effective moisture in interior and southern Alaska as inferred from lake-level records and may be related to broad-scale atmospheric circulation changes. In the shelf sediments, carbonate abundance increases markedly at 3.5 kcal. BP, and δ¹⁸O increases from −18.0‰ to −16.5‰ at 2.1 kcal. BP, possibly resulting from increased temperature and/or summer precipitation. After 2.1 kcal. BP, δ¹⁸O fluctuates with an overall decreasing trend to −17.2‰ at 0.9 kcal. BP. Late-Holocene variations in our δ¹⁸O record display coherent patterns with regional glacier fluctuations at centennial to millennial scales, suggesting that δ¹⁸O minima were related to a combination of low temperatures and enhanced winter snowfall. Holocene variations in organic matter abundance at Wahoo Lake also show broad similarities to total solar irradiance, implying that suborbital solar variability played a role in modulating regional climate and aquatic productivity.

Keywords

Arctic carbonate isotopes Holocene solar irradiance

Introduction

Anthropogenic climate warming is amplified in the Arctic compared with the Northern Hemisphere average (IPCC, 2013; Serreze and Barry, 2011), and Arctic temperatures were higher in recent decades than in the previous two millennia (Kaufman et al., 2009). Paleoclimate records provide a long-term context to understand this sensitivity and assess the spatiotemporal patterns of climate change. Despite the recent acceleration of paleoclimate research, the spatial coverage of paleorecords and knowledge about the complex interplay of forcing mechanisms remain limited in many regions. A recent review of paleoclimate records from eastern Beringia highlights the spatial heterogeneity of Holocene climate change (Kaufman et al., 2016). Even some of the major features that were previously thought to be widespread, such as early-Holocene warmth and moisture deficit (Kaufman et al., 2004), do not appear to be ubiquitous in the proxy climate records from this vast region. For example, midge-based temperature reconstructions in Alaska revealed nonlinear temperature responses to the Holocene trend of decreasing insolation (Clegg et al., 2011), and oxygen-isotope records showed intra-regional variability of moisture history driven by changes in the strength and location of atmospheric circulation patterns (Anderson et al., 2005; Chipman et al., 2012). This spatial heterogeneity may be attributed to the influence of several synoptic-scale atmospheric features (Mock et al., 1998) and linkages with continental-scale ocean–atmosphere processes (L’Heureux et al., 2004). The relative paucity of paleoclimate records from remote areas of the Alaskan Arctic (Kaufman et al., 2016) hampers our understanding of the spatial patterns and associated drivers of Holocene climate variability and our assessment of the ecological impacts of recent warming (Chipman et al., 2016; Hu et al., 2015).

We analyzed sediment cores from Wahoo Lake (69.077°N, 146.928°W, 700 m a.s.l.), located in the northeastern Brooks Range, for lithological characteristics, including carbonate and organic matter (OM) percentages, as well as the isotopic composition of Pisidium. The calcification of freshwater mollusks occurs near equilibrium with the isotopic composition of lake water, with estimated disequilibrium offsets of 0.86‰ ± 0.17‰ for δ¹⁸O and −0.19‰ ± 0.63‰ for δ¹³C (Fritz and Poplawski, 1974; Von Grafenstein et al., 1999). Thus, the isotopic analysis of Pisidium circumvents the confounding influences of allochthonous detrital carbonate on the isotopic signal preserved in lake sediments (Leng and Marshall, 2004). In this paper, we report these results and interpret them in the context of modern isotopic composition of lake and stream water to infer Holocene climate variability. In particular, our isotope data document changes in effective moisture from the early Holocene to middle Holocene and fluctuations in the temperature and/or seasonality of precipitation during the late-Holocene. These results contribute to the growing body of evidence for Holocene climate change in northern Alaska and allow for comparisons with paleoclimate data from other regions to assess the large-scale climate forcings.

Study site

Wahoo Lake is located in the Echooka River Valley in the northeastern foothills of the Brooks Range (Figure 1a). Bedrock in the watershed consists of Permian and Triassic sandstone, siltstone, and shale of the Sadlerochit Group, with nearby outcrops of Pennsylvanian and Mississippian limestone of the Lisburne Group (Beikman, 1980). The valleys in this region are characterized by surficial deposits of colluvium and alluvial sediments originating from late-Wisconsinan glacial advances in the Brooks Range (Carson, 2006; Kaufman et al., 2011). Watershed soils are predominantly gelisols underlain by permafrost, and vegetation is dominated by shrub tundra (US Geological Survey, 2012). In July 2011, Wahoo Lake had a surface area of 0.28 km² and a maximum depth of 17.1 m. Although no bathymetric map is available for Wahoo Lake, our preliminary depth survey revealed four sub-basins, which are progressively deeper from east to west. The lake is a hydrologically open basin at present with a large watershed (2.7 km²), a single outlet, and two inlet streams (Figure 1b). The outlet stream currently incises a ridge to the west of the lake and drains into a tributary of the Echooka River. The meandering of this tributary stream has formed a broad valley floor, which intersects the ridge and strongly suggests that the river eroded the hillslope in the past. Thus, the lake levels at Wahoo Lake may have been influenced over time by variations in the steepness of the outflow channel in response to erosion of the intervening hillslope. The modern shoreline of Wahoo Lake is characterized by a prominent submerged shelf. Gastropod (subclasses Prosobranchia and Pulmonata) and bivalve (genus Pisidium) shells were abundant along these shallow shelves and the shoreline of the lake in the summer of 2011.

Figure 1.

(a) Northern Alaska (GoogleEarth^TM) with locations of Wahoo Lake (black), Barrow weather station (gray), and paleoclimate studies mentioned in text (white): (1) Qalluuraq Lake (Wooller et al., 2012); (2) Kurupa Lake (Boldt et al., 2015); (3) Takahula Lake (Clegg and Hu, 2010); (4) Tangled Up Lake (Anderson et al., 2001); (5) Meli Lake (Anderson et al., 2001); and (6) Trout Lake (Irvine et al., 2012). (b) Wahoo Lake watershed (shaded), with black dots marking approximate core sites and dashed arrows indicating inlets and outlet streams. Base map is USGS Mt. Michelson (A-5) quadrangle, 1:63,360 series, with 100-ft contours.

The ecoregions of northern Alaska experience below-freezing air temperatures and persistent snow cover for more than 8 months of the year (Zhang et al., 1996). Snow and rainfall in the northern Brooks Range originate from both the North Pacific and Arctic Ocean, with relative source amounts dependent on the location and strength of large-scale atmospheric circulation patterns (Klein et al., 2016). Snowfall accounts for ~40% of total annual precipitation in the northern foothills, although the contribution of winter precipitation north of the Brooks Range may be much higher because gauges likely underestimate snowfall (Zhang et al., 1996). Spring snowmelt results in a rapid decrease in albedo from 75% to 17% within the span of several days, and the subsequent increase of solar absorption ushers in the relatively short summer growing season (Stone et al., 2002). Mean temperature and total precipitation during the summer (June–August) at the Toolik Lake weather station, approximately 120 km southeast of Wahoo Lake, are 9.1°C and 193.1 mm, respectively (based on all available observations from 1988 to 2014; Environmental Data Center Team, 2015). Mean temperature and total precipitation during the winter (December–February) are −22.1°C and 288.9 mm, respectively.

Methods

Modern water analyses

In July 2011, temperature, dissolved oxygen concentration, and specific conductivity were measured with an YSI-85 probe at 1-m increments in the water column of the eastern basin of Wahoo Lake. Water samples were collected in 20-mL scintillation vials from inlet and outlet streams and at 1-m intervals in the water column of the lake using a Van Dorn sampler. All water samples were collected on days without rainfall events. These water samples were analyzed for δ¹⁸O and δD at the University of Arizona using a Finnigan Delta S mass spectrometer (precision <0.08‰ for δ¹⁸O and <0.9‰ for δD).

Modern isotopic values of precipitation from the Barrow weather station were obtained from the Global Network of Isotopes in Precipitation (GNIP) database (IAEA/WMO, 2014) and used to construct a Regional Meteoric Water Line (RMWL; Figure 2a, inset). The Barrow station is approximately 450 km northwest of Wahoo Lake (Figure 1a), and the GNIP isotope data from this site were obtained from irregularly sampled monthly precipitation from January 1962 to December 1969. We also construct a Regional Evaporation Line (REL) for the region from isotopic measurements of surface water from lakes in northern and interior Alaska (Anderson et al., 2013; Clegg and Hu, 2010) and northwestern Canada (Turner et al., 2010). To account for the effects of altitude and latitude on the isotopic values of precipitation at our site compared with values at Barrow, we modeled monthly isotopic values of precipitation at Wahoo Lake using the Online Isotopes in Precipitation Calculator (OIPC; Bowen, 2013; Bowen and Revenaugh, 2003; Bowen and Wilkinson, 2002). All measured and modeled water data are plotted against the Global Meteoric Water Line (GMWL; Figure 2a).

Figure 2.

Isotope composition of modern water. (a) OIPC-modeled seasonal precipitation values for Wahoo Lake (black circles; Bowen, 2013; Bowen and Revenaugh, 2003; Bowen and Wilkinson, 2002); Barrow GNIP data (white circles; IAEA/WMO, 2014) used to construct RMWL (inset); modern lake water samples (gray diamonds; Anderson et al., 2013; Clegg and Hu, 2010; Turner et al., 2010) used to construct a Regional Evaporation Line (REL; dashed line); and modern isotope values from the inlets (diamonds), outlet (square), and surface water (gray triangles) of Wahoo Lake. All data are plotted with the Global Meteoric Water Line (GMWL). (b) Comparison of δ¹⁸O values from Wahoo Lake and monthly OIPC-modeled precipitation. (c) Lake water δ¹⁸O values plotted by depth in the water column.

Sediment analyses

In July 2011, we obtained sediment cores from the eastern basin (8.1 m depth) and an adjacent shelf (2.8 m depth) using a modified Livingstone piston corer. We did not obtain sediment from the deepest basin (17.1 m depth) because of our interest in obtaining carbonate shells for isotopic analysis. To obtain the sediment–water interface at each core site, we used a polycarbonate tube fitted with a piston, and the unconsolidated sediments of the uppermost 20–30 cm were subsampled at contiguous 0.5-cm intervals in the field. Overlapping cores from each site were split lengthwise and correlated using stratigraphic markers to ensure a continuous sedimentary sequence. Loss on ignition (LOI) was performed on 0.5-cm³ subsamples at contiguous 1-cm intervals. These subsamples were weighed before and after drying at 120°C overnight and following combustion at 550°C for 2 h and at 950°C for 4 h in a muffle furnace. Sample weights were used to calculate sediment bulk density (BD; g dry sediment cm⁻³) and percentages of OM (% LOI at 550°C), calcium carbonate (CaCO₃% = % LOI at 950°C × (100/44)), and residual lithic and ash (LA% = 100 − OM% − CaCO₃%).

To isolate carbonate material for isotopic analysis, we washed subsamples at contiguous 1-cm intervals from the cores through a 125-µm mesh with distilled water. Carbonate shells were identified to genus and subclass level using a dissection microscope and the taxonomic keys of Thorp and Covich (2009). All identifiable Pisidium shells within each sample were combined for isotopic analyses, and the number of shells per sample ranged between 1 and 7 (mean = 2). Shell sizes (diameter of longest axis) ranged between 1 and 9 mm (mean = 3.7 mm). Pisidium shells were isolated and prepared for δ¹⁸O and δ¹³C analyses using a protocol adapted from Ito (2001). Briefly, shells were rinsed with deionized H₂O to remove sediment particles, transferred to a clean glass vial using 95% ethanol, immersed overnight in 2.5% NaOCl, triple-rinsed with deionized H₂O, air-dried, and ground to a fine powder with a mortar and pestle. Isotopic analyses were conducted at the Stable Isotope Laboratory of the Illinois State Geological Survey using a Kiel III carbonate device and Finnigan MAT252 isotope ratio mass spectrometer. Precision on laboratory standards is ±0.1‰ and ±0.15‰ for δ¹³C and δ¹⁸O, respectively.

We measured ²¹⁰Pb activity on the uppermost sediments from both of our core sites. Samples were prepared following the protocol of Eakins and Morrison (1978), and activity was measured with an Ortec OctêtePlus alpha spectrometer. The age of each sample from the basin core was calculated with a constant rate of supply model (Binford, 1990; Oldfield et al., 1978). The ²¹⁰Pb profile from the shelf sediments did not follow an exponential decay pattern, which may have been caused by faulty spectrometer readings due to a black precipitate that formed on the planchettes. This issue precluded the use of ²¹⁰Pb dating for the shelf core. For accelerator mass spectrometry (AMS) ¹⁴C dating, we submitted 10 terrestrial-plant macrofossils to the Lawrence Livermore National Laboratory. Prior to submission, we followed the acid–base–acid procedure described in Oswald et al. (2005) to prepare the terrestrial macrofossils. Radiocarbon ages were calibrated to years before CE 1950 (cal. BP) using the IntCal13 dataset in CALIB version 7.0.4 (Reimer et al., 2013; Stuiver and Reimer, 1993). We developed an age–depth relationship with a 95% confidence interval for each core using a cubic spline function with 10,000 bootstrapped iterations in the CLAM age-modeling routine (Blaauw, 2010; Figure 3).

Figure 3.

Age–depth relationships for the (a) basin and (b) shelf sediment cores from Wahoo Lake. Cubic spline model, radiocarbon ages, and ²¹⁰Pb ages plotted with 2σ ranges.

Results

Isotopic composition of modern water

In July 2011, the temperature of the water column at Wahoo Lake ranged from 16.3°C at the surface to 3.5°C in the hypolimnion, with a well-defined thermocline at 4.0–5.0 m below the surface. Dissolved oxygen in the epilimnion (0–4.0 m) was 9.7 ppm, declining to 0.0 ppm in the bottom 3 m of the water column. Lake water δ¹⁸O values were generally lower (<−19.9‰ Vienna Standard Mean Ocean Water (VSMOW)) in the upper 5 m of the water column, with the exception of a peak at 2 m, and higher (>−19.9‰) between 6 and 7 m water depth (Figure 2c). The isotopic values of the outlet (δD = −162.0‰, δ¹⁸O = −19.8‰) and surface water (δD = −162.5‰ and −162.9‰, δ¹⁸O = −20.0‰ and −20.1‰) samples were higher than those of the inlet streams (δD = −164.5‰ and −165.4‰, δ¹⁸O = −20.8‰ and −21.1‰).

At the Barrow weather station, precipitation δD values range from −223.3‰ to −18.2‰, and δ¹⁸O values from −29.0‰ to −4.2‰. The slope of the RMWL based on these data is 7.12 δD/δ¹⁸O (Figure 2a, inset). The OIPC-modeled δD values for Wahoo Lake range from −219‰ to −118‰, and δ¹⁸O values range from −29.0‰ to −15.2‰. These values fall along the RMWL, with a very similar slope (7.10 δD/δ¹⁸O). The slopes of the GNIP-based RMWL and monthly precipitation values are lower than those of the GMWL (δD = 8.0 × δ¹⁸O + 10.0) and the surface water line for the Alaska-Yukon territory (δD = 8.0 × δ¹⁸O + 6.4; Lachniet et al., 2016). We plot our data against the GMWL to facilitate comparisons of lake water isotopic values between our site and other sites in Alaska and Canada (Figure 2a). The isotopic values of inlet and outlet water samples fall near the April, October, January, and February OIPC values and plot along the REL (slope = 4.67 δD/δ¹⁸O), near the intercept between the REL and GMWL relative to most of the enriched regional lake water values. The isotopic difference between inlet and outlet streams at Wahoo Lake falls within the variability of both the OIPC and Barrow station values (Figure 2a and b).

Sediment-core chronology

The six ¹⁴C ages for the sediment sequence of the Wahoo Lake basin are all in stratigraphic order (Table 1; Figure 3a). The base of the sediment core (264 cm) ended in unconsolidated gravel, and sedimentation in the eastern basin began in the early Holocene (11.5 kcal. BP). In this basin, the mean sediment accumulation rate is low (0.02 cm yr⁻¹) from 11.5 to 1.4 kcal. BP, increasing to peak values of ~0.12 cm yr⁻¹ between 1.4 and 1.0 kcal. BP and decreasing to 0.06 cm yr⁻¹ from 1 ka to present. The shelf sediment core also ended in unconsolidated gravel. The core spans from 5.3 kcal. BP to present, with all four ¹⁴C ages in stratigraphic order (Figure 3b). The mean sediment accumulation rate of the shelf core is 0.03 cm yr⁻¹ from 5.3 to 3.0 kcal. BP, increasing to 0.18 cm yr⁻¹ between 3.0 and 2.5 kcal. BP and decreasing to 0.05 cm yr⁻¹ from 2.5 kcal. BP to present.

Table 1.

Radiocarbon ages, ²¹⁰Pb activity, and modeled or calibrated dates from Wahoo Lake sediments.

Sample depth (cm)	Material	Laboratory ID	²¹⁰Pb activity (dpm g⁻¹) or ¹⁴C age (yr BP)^a	Modeled or calibrated date (cal. yr BP) with 2σ range
Basin
0.0–1.0	Bulk sediment	UIUC R5-01	31.14 ± 2.84	−61
1.0–2.0	Bulk sediment	UIUC R5-02	30.05 ± 2.53	−55	−56– −55
2.0–3.0	Bulk sediment	UIUC R5-03	22.20 ± 1.91	−47	−48– −46
3.0–4.0	Bulk sediment	UIUC R5-04	17.62 ± 1.42	−40	−42– −38
4.0–5.0	Bulk sediment	UIUC R5-05	19.92 ± 1.78	−34	−36– −31
5.0–6.0	Bulk sediment	UIUC R5-06	13.05 ± 1.17	−22	−26– −18
6.0–7.0	Bulk sediment	UIUC R5-07	11.58 ± 1.07	−11	−16– −5
7.0–8.0	Bulk sediment	UIUC R5-08	9.22 ± 0.82	3	−5–11
8.0–9.0	Bulk sediment	UIUC R5-09	7.85 ± 0.71	16	4–27
9.0–10.0	Bulk sediment	UIUC R5-11	5.81 ± 0.49	29	12–46
10.0–11.0	Bulk sediment	UIUC R5-12	5.68 ± 0.5	38	17–60
11.0–12.0	Bulk sediment	UIUC R5-13	4.14 ± 0.37	53	19–87
12.0–13.0	Bulk sediment	UIUC R5-14	3.80 ± 0.35	67	16–118
14.0–15.0	Bulk sediment	UIUC R5-16	3.05 ± 0.27	N/A
16.0–17.0	Bulk sediment	UIUC R5-18	2.91 ± 0.26	N/A
18.0–19.0	Bulk sediment	UIUC R5-21	3.05 ± 0.29	N/A
20.0–21.0	Bulk sediment	UIUC R5-23	2.18 ± 0.19	N/A
22.0–23.0	Bulk sediment	UIUC R5-25	2.45 ± 0.25	N/A
73.0–74.0	Twigs and rootlets	CAMS 160624	1480 ± 50	1372	1296–1521
112.5–113.0	Leaf fragment	CAMS 158762	1830 ± 40	1767	1629–1870
134.5–135.0	Leaf/wood/rootlets	CAMS 158763	3375 ± 50	3619	3476–3818
159.0–159.5	Twig w/ bark	CAMS 157403	5375 ± 35	6193	6012–6281
185.5–186.0	Wood fragment	CAMS 158764	7170 ± 35	7985	7936–8035
217.0–217.5	Wood and bark	CAMS 157404	8410 ± 170	9373	8991–9887
Shelf
41.0–41.5	Twigs and rootlets	CAMS 167928	1190 ± 70	1118	969–1268
122.0–123.0	Leaf and wood	CAMS 160625	2555 ± 40	2648	2491–2757
199.0–200.5	Leaf/wood/rootlets	CAMS 160626	3115 ± 35	3331	3229–3437
231.0–232.5	Two small twigs	CAMS 157405	4400 ± 40	4970	4857–5267

UIUC: University of Illinois, Urbana-Champaign; CAMS: Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory, Livermore, CA.

Lead-210 activity or conventional radiocarbon years before present (CE 1950), with standard deviation.

Core lithology and sediment composition

The early-Holocene sediments (264–162 cm; 11.5–6.3 kcal. BP) of the basin core are thinly (<1 cm) interbedded fine silts and clays (Figure 4a). Residual lithic and ash abundance (LA%) is generally high in this laminated interval (mean = 57.0%). Pisidium shells are abundant in some intervals of the early-Holocene sediments; however, they are not continuously present, resulting in the relatively coarse and uneven resolution of our isotope analysis for the early Holocene. CaCO₃% ranges between 8.1% and 50.9% (mean = 28.0%), and OM% varies between 3.7% and 38.0% (mean = 15.3%; Figure 4a). At 162 cm (6.3 kcal. BP), the sediments transition to black gyttja with low CaCO₃% (2.9–25.3%, mean = 7.6%) and high OM% (11.7–65.3%, mean = 40.1%). Pisidium shells are not present in this sedimentary unit. At 96 cm (1.4 kcal. BP), the sediments transition to silt-rich gyttja with lower OM% (15.1–58.3%, mean = 36.6%), and highly variable CaCO₃% (1.4–53.5%, mean = 11.8%). Only one sample (at 96 cm; 1.4 kcal. BP) from this interval had enough identifiable Pisidium material for isotopic analysis.

Figure 4.

Lithological descriptions of (a) basin and (b) shelf sediments and sediment composition based on loss on ignition, Wahoo Lake. (c) Pisidium δ¹⁸O and δ¹³C values from shelf (circles) and basin (triangles) cores.

The basal sediments of the shelf core are predominantly gyttja with abundant moss fragments. These sediments have high OM% (39.6–79.3%, mean = 55.9%) and low CaCO₃% (1.2–13.5%, mean = 5.8%; Figure 4b). At 200.5 cm (3.5 kcal. BP), sediments transition to mud rich in carbonate, with 24.8–74.2% (mean = 49.0%) CaCO₃ and 12.5–50.1% (mean = 28.6%) OM. Pisidium and gastropod shells are present from this transition until 34 cm (0.9 kcal. BP), above which carbonate shells are no longer preserved in the sediments.

Oxygen-isotope and carbon-isotope compositions of Pisidium carbonate

The δ¹⁸O and δ¹³C values of Pisidium at Wahoo Lake range from −18.7‰ to −16.2‰ (VPDB) and from −7.1‰ to −2.3‰ (VPDB), respectively (Figure 4c). The sample sizes for our Pisidium isotope measurements were 1–7 shells per sample. The ranges of isotopic values for single-shell samples (−18.5‰ to −16.2‰ for δ¹⁸O and −7.1‰ to −3.0‰ for δ¹³C) are similar to those for multiple-shell samples (−18.7‰ to 16.5‰ for δ¹⁸O and −6.7‰ to −2.3‰ for δ¹³C). Isotopic values from 11.4 to 7.2 kcal. BP were obtained from Pisidium remains in the basin core, but no Pisidium shells are available for isotopic analysis in the remainder of this core, except the single sample at 1.4 kcal. BP. Thus, the isotopic record from 3.5 to 0.9 kcal. BP was derived primarily from Pisidium remains in the shelf core. The mean δ¹⁸O value from the early-Holocene basin samples (−17.2‰, standard deviation (SD) = 0.40‰, n = 18) is higher than that of the late-Holocene shelf samples (−17.6‰, SD = 0.46‰, n = 63; t-test, p = 0.0003), although the ranges of δ¹⁸O values from the basin (−18.0‰ to −16.4‰) and the shelf (−18.7‰ to −16.2‰) overlap. The δ¹³C values show a greater offset, with a mean (range, SD) of −6.1‰ (−6.7‰ to −4.1‰, 0.45‰) for the basin core and −4.4‰ (−7.1‰ to −2.3‰, 0.81‰) for the shelf core (t-test, p < 0.0001).

The δ¹⁸O value of the oldest sample (11.4 kcal. BP) is −17.5‰, after which values increase and remain relatively high (−17.2‰ to −16.4‰) from 11.3 to 8.9 kcal. BP. These early-Holocene δ¹⁸O values are followed by a marked decrease from −17.2‰ at 8.9 kcal. BP to a minimum of −18.0‰ at 7.4 kcal. BP and a subsequent increase to −17.0‰ at 7.2 kcal. BP. The δ¹³C values in the early-Holocene vary between −6.5‰ and −4.6‰ and generally decline through time.

The high abundance of Pisidium remains and relatively high sedimentation rate of the shelf core resulted in high-resolution isotopic analysis for the late-Holocene compared with the early-Holocene. From 3.5 to 0.9 kcal. BP, both δ¹⁸O and δ¹³C fluctuate markedly. δ¹⁸O increases from −18.0‰ to −16.5‰ with high variability from 3.5 to 2.1 kcal. BP and subsequently declines overall to a minimum of −18.5‰ at 1.8 kcal. BP. δ¹⁸O increases to −17.1‰ between 1.8 to 1.5 kcal. BP, followed by a decrease to −17.7‰ at 1.1 kcal. BP and then by an increase to −17.2‰ at 0.9 kcal. BP. From 3.5 to 2.8 kcal. BP, δ¹³C decreases from −3.8‰ to −7.1‰, followed by an increasing trend until 2.0 kcal. BP with a range of −7.1‰ to −2.3‰. From 2.0 to 0.9 kcal. BP, δ¹³C ranges from −5.8‰ to −2.3‰ with an overall decreasing trend.

The δ¹⁸O and δ¹³C values of Pisidium are not significantly correlated (r = 0.13, p = 0.25, n = 82), although the correlation coefficient was higher for the early-Holocene (r = 0.41, p = 0.09, n =18) than for the late-Holocene (r = 0.19, p = 0.12, n = 64; Figure 5). The δ¹⁸O and δ¹³C records show some similarity at millennial timescales in the early-Holocene, with a general pattern of decreasing values from 11.4 to 7.4 kcal. BP and an increase from 7.4 to 7.2 kcal. BP (Figure 4c). In the late-Holocene, both δ¹⁸O and δ¹³C increase overall from 3.5 to 2.1 kcal. BP and decrease from 2.1 to 0.9 kcal. BP, although their centennial-scale fluctuations are not consistent.

Figure 5.

δ¹⁸O versus δ¹³C values of Pisidium shells preserved in Wahoo Lake sediments.

Discussion

Controls on water isotopic composition

Modern precipitation at Wahoo Lake, based on the OIPC model, is relatively depleted in ¹⁸O and ²H compared with that at the Barrow weather station (Figure 2a). This pattern reflects inland movement of moisture relative to the Barrow weather station, which is ~440 km northwest of Wahoo Lake (Figure 1a), as well as the effects of continentality and altitude on water isotopes derived from both Arctic and North Pacific moisture sources (Lachniet et al., 2016). The δ¹⁸O and δD values of water samples from Wahoo Lake are near the OIPC-modeled values of winter and spring precipitation (Figure 2a and b). This suggests that the modern water budget of Wahoo Lake is dominated by winter and spring snowfall, a pattern that has been documented at lakes from the southern Brooks Range (Clegg and Hu, 2010) and streams throughout Alaska and the Yukon Territory (Lachniet et al., 2016). Spring snowmelt occurs when the ground is still frozen, and meltwater reaches lakes via overland flow and ephemeral streams with no or minimal isotopic fractionation. The inlet-water isotopic values from Wahoo Lake obtained during the summer are more depleted in ¹⁸O and ²H than summer precipitation (Figure 2a), which reinforces the interpretation that input water to the lake is largely derived from depleted snowmelt. In comparison, summer precipitation is more likely to be absorbed by plants and soils in the watershed and thus likely contributes less overall to the lake water isotopic signal. Nonetheless, the intersection of the GMWL and lake-specific evaporation line based on the input and outflow water samples from Wahoo Lake is slightly higher (δ¹⁸O = −22.34‰ and δD = −168.77‰) than the intersection of the GMWL and the REL (δ¹⁸O = −23.82‰ and δD = −180.56‰; Figure 2), which is an estimate of the mean annual weighted isotopic composition of precipitation (Turner et al., 2010). This pattern suggests some influence of summer rainfall and/or groundwater inputs in addition to snowfall (Anderson et al., 2013; Turner et al., 2010). The relatively depleted isotopic values of the lake water at Wahoo Lake may also reflect inputs from permafrost thaw and/or supra-permafrost groundwater flow. Despite these potentially interacting influences on the hydrological budget and isotopic composition of lake water, the isotopic values of modern surface waters in Alaska are most strongly influenced by climatic and physiographic parameters and are thus considered appropriate proxies for δ¹⁸O and δD of meteoric precipitation (Lachniet et al., 2016).

Evaporation in lakes with a long water residence time can result in post-input isotopic fractionation of lake water. We do not have the limno-hydrological data needed to determine the water residence time of Wahoo Lake. The lake is a topographically open basin at present, and the lake water residence time is probably short given the high relief (>400 m from highest ridge to outflow channel) and relatively short flowpaths (<1.5 km) in the catchment (McGuire et al., 2005). The isotopic composition of the modern water from Wahoo Lake plot along the REL defined by regional lake water samples (Figure 2a) and the differences in the δ¹⁸O and δD values of the inlet, outlet, and lake water samples from Wahoo Lake (Figure 2a) indicate isotopic enrichment of lake water during its residence in the lake basin in July 2011. However, these differences are small (~1‰ for δ¹⁸O and ~4‰ for δD), suggesting that evaporative influence on the lake water isotopic composition is limited at present. In addition, the Wahoo Lake water values plot near the intersection of the GMWL and REL (Figure 2a), indicating that the degree of evaporative enrichment is small relative to the other lake surface waters in the region. Furthermore, the δ¹⁸O of the epilimnion water samples are depleted in ¹⁸O relative to deeper waters (Figure 2c), a pattern opposite to that expected if lighter isotopes were being preferentially removed from the surface water via evaporation.

The extent of evaporation at Wahoo Lake can be assessed by calculating the evaporation-to-inflow ratio (E/I; Turner et al., 2010; Yi et al., 2008). An evaporation-dominated lake would have an E/I of >1. In lieu of an index lake or evaporation pan experiment, we use the modern water δ¹⁸O values at Wahoo Lake to estimate the equilibrium and kinetic separation terms (Gibson and Edwards, 2002; Horita and Wesolowski, 1994) and the evaporation rate between the inlet and outlet (Gibson and Edwards, 2002), assuming relative humidity, temperature, and atmospheric water vapor δ¹⁸O values of 0.74, 9.3°C, and −28.2‰, respectively, based on measurements from Toolik Lake in the same region (Environmental Data Center Team, 2015; Klein et al., 2016). The E/I ratio at Wahoo Lake is low (0.25), which suggests that evaporation losses are only 25% of the inflow water (Anderson et al., 2013). Together, these data suggest that evaporation plays a relatively minor role in determining the isotopic composition of modern lake water at our site.

Inferring Holocene climate change from isotopic and lithological data

At Wahoo Lake, δ¹⁸O values are overall higher in the early-Holocene sediments from the basin core (mean = −17.2‰) than in the late-Holocene samples from the shelf core (mean = −17.6‰; Figure 4c). It is difficult to determine whether these differences are related solely to temporal changes in the isotopic composition of the lake water or whether they reflect isotopic differences in the water depth at which the shells formed. Samples from the modern water column show that the δ¹⁸O offset between shallow and deep water is minimal (~0.2‰; Figure 2c), and thus, it is unlikely that differences in water depth (i.e. Pisidium habitat) would dramatically affect the isotopic values. In addition, our sedimentary indicators suggest that Wahoo Lake in the early Holocene was shallow compared with modern (see below), and thus, the water depth at the time when the Pisidium shells were deposited in the basin may have been comparable with the modern water depth at the shelf site. Furthermore, the ranges of δ¹⁸O values between the core sites overlap, and the single late-Holocene Pisidium sample from the basin core has a δ¹⁸O value of −17.6‰ and a δ¹³C value of 4.1‰, which lie between the isotopic values of surrounding contemporaneous samples from the shelf core (Figure 4c). These data together suggest that the difference in isotopic composition of the water between deep and shallow habitats is minimal and that the isotopic composition of Pisidium shells that formed simultaneously in different areas of the lake reflect the same lake water isotopic composition. Thus, we interpret the δ¹⁸O difference between the early-Holocene and late-Holocene samples at Wahoo Lake as temporal changes in the isotopic composition of the lake water.

In the summer of 2011, Pisidium were only observed on the shallow shelves near the shoreline of Wahoo Lake, and we thus assume that the shell remains preserved in our lake-sediment cores either formed in situ (shelf core) or near the core site when lake levels were lower (basin core). Freshwater Pisidium are infaunal filter feeders that calcify near equilibrium with the δ¹⁸O and δ¹³C of lake water. The vital effects (i.e. disequilibrium offsets) for Pisidium spp. are estimated to be 0.86‰ ± 0.17‰ for δ¹⁸O and −0.19‰ ± 0.63‰ for δ¹³C and are generally constant through time and at different lake depths and temperatures (Apolinarska and Hammarlund, 2009; Von Grafenstein et al., 1999). Furthermore, the ranges of isotopic values for modern Pisidium spp. are comparable with the ranges within individual species of other taxa (ostracods) and similar to or smaller than the annual range of calcite precipitated in equilibrium with lake water (Von Grafenstein et al., 1999). In light of these studies, which document that Pisidium shells closely reflect water isotopic composition in freshwater lakes, we assume that the variability associated with species-specific and disequilibrium offsets are minimal in our record. Although we do not have duplicates for our isotopic samples because of the lack of available material in our core, the ranges of isotopic values of samples are similar for single-shell versus multiple-shell samples (see section ‘Results’), suggesting that variations in our record are also not influenced by small sample sizes. The lifespan of individual Pisidium bivalves ranges from ca. 4 months to 4 years (Holopainen and Hanski, 1986), and Pisidium precipitates carbonate continuously to form their shells over their lifespan. However, Pisidium produce the majority of their carbonate during the ice-free growing season (Apolinarska and Hammarlund, 2009; Hammarlund et al., 1999; Von Grafenstein et al., 1999), and the isotopic values of their shells are close to lake water values from April to September (Von Grafenstein et al., 1999). Thus, the Pisidium shells from the Wahoo Lake sediments likely capture the isotopic signal of summer lake water integrated over several years, which is well within the chronological uncertainty associated with our samples.

The isotopic composition of carbonate is controlled by two major factors (assuming constant vital effects and minimal influences of habitat and species differences): the temperature at which carbonate formation occurs within the lake, and the isotopic composition of the water from which the carbonate is precipitated. The Pisidium carbonate δ¹⁸O record from Wahoo Lake displays a range of 2.5‰, with a minimum of −18.7‰ and maximum of −16.2‰ (Figure 4c). The temperature-dependent fractionation of the δ¹⁸O composition of carbonate is approximately −0.24‰ °C⁻¹, assuming chemical equilibrium and subsequent equilibrium isotope exchange (Craig, 1965). Combining this with the isotopic effect of meteoric precipitation established for northern Alaska (0.36‰ °C⁻¹; Klein et al., 2016) results in a net temperature effect of 0.12‰ °C⁻¹ (Leng and Marshall, 2004). Thus, the δ¹⁸O range of 2.5‰ at Wahoo Lake is equivalent to temperature variation of up to 20.8°C if temperature-dependent fractionation was the sole factor affecting our δ¹⁸O record. Holocene temperature changes of this magnitude are very unlikely to have occurred at Wahoo Lake. Several factors may lead to large differences in the isotopic composition of lake water. Evaporative enrichment alters the composition of water by several per mille, with the magnitude depending on the humidity and temperature of the atmosphere under which it occurs (Dansgaard, 1964). For example, Clegg and Hu (2010) found that δ¹⁸O values of Chara stem encrustations in Takahula Lake, located in the southern Brooks Range, varied by as much as 5‰, with winter precipitation inputs and post-input evaporative enrichment interacting as key controls of the δ¹⁸O values. Similarly, the relative seasonality of the meteoric precipitation contributing to the lake water budget can also result in several per mille deviations in lake water isotopic values (Leng and Marshall, 2004). For example, the expected Pisidium δ¹⁸O value of modern water at Wahoo Lake (collected in early July) is −19.4‰ VPDB, accounting for the equilibrium fractionation of calcite at the mean epilimnion temperature (Leng and Marshall, 2004) and the Pisidium disequilibrium offset (Von Grafenstein et al., 1999). This value is 0.7‰ lower than the minimum δ¹⁸O value of our Pisidium samples. This offset is not surprising, as we expect higher Pisidium δ¹⁸O values in shells that calcify throughout the ice-free season, thus incorporating lake water isotopic values that are influenced by precipitation in late summer and early fall, as well as some degree of evaporative enrichment. The relatively depleted values at present compared with the rest of the record may also reflect inputs from recent permafrost thaw in the region (Chipman et al., 2016; Osterkamp, 2007), although we do not have the data necessary to confirm this aspect of the hydrological budget at Wahoo Lake.

We cannot unambiguously tease apart the various factors that may have contributed to the variations in the Pisidium δ¹⁸O record from Wahoo Lake. Nonetheless, past changes in the hydrologic setting of the lake can help assess the influence of relative controls on the isotopic composition of the lake water. Small, closed-basin lakes are strongly influenced by evaporative enrichment compared with large, open-basin lakes, which are more sensitive to temperature and seasonality of precipitation (Leng and Marshall, 2004). The lithologic changes in the Wahoo Lake sediment cores provide qualitative constraints for interpreting changes in the hydrological setting and controls on carbonate δ¹⁸O through time. From 11.5 to 6.3 kcal. BP, the basin-core sediments have low OM% and high lithic abundance (Figure 4a), suggesting relatively high allochthonous inputs to the lake and/or low productivity. Additionally, high CaCO₃% and the presence of Pisidium shells, which requires high Ca²⁺ concentrations (Hunter, 1964), suggest that the lake water was supersaturated with respect to carbonate, although the bulk-sediment CaCO₃ likely also includes allochthonous sources. These data suggest that unlike today, Wahoo Lake may have been a topographically closed basin in the early Holocene with reduced subsurface throughflow and that the isotopic values of Pisidium during the early Holocene were likely influenced by evaporative enrichment. Thus, we interpret δ¹⁸O fluctuations in the early Holocene primarily as a proxy for changes in effective moisture.

The δ¹⁸O record from Wahoo Lake shows relatively high values (−17.2‰ to −16.4‰) from 11.3 to 8.9 kcal. BP with marked variations at millennial timescales (Figure 4c), suggesting that evaporative ¹⁸O enrichment was high but variable. δ¹⁸O values display a decreasing trend from 8.9 to 7.4 kcal. BP, implying increasing effective moisture. The abrupt increase in δ¹⁸O from 7.4 to 7.2 kcal. BP suggests a short interval of enhanced aridity prior to the middle-Holocene hiatus in Pisidium shell preservation. Although factors other than aridity may have contributed to the marked δ¹⁸O fluctuations within the early Holocene, we do not attempt to interpret these fluctuations in detail given the coarse temporal resolution of our early-Holocene δ¹⁸O data.

Strong covariance of carbonate δ¹³C and δ¹⁸O is a signal of hydrologically closed basins (Talbot, 1990). At Wahoo Lake, Pisidium δ¹³C and δ¹⁸O from the early Holocene are not significantly correlated, but they show stronger covariance than during the late Holocene (Figure 5). The lack of a strong δ¹³C–δ¹⁸O relationship may reflect insufficient duration of closed-basin conditions (Leng and Marshall, 2004; Li and Ku, 1997) and/or the influence of subsurface water flow despite topographically closed conditions. In addition, the relationship may be convoluted by factors such as aquatic productivity, respiration of OM near the sediment–water interface, and the presence of methane in pore waters (e.g. Apolinarska and Hammarlund, 2009).

A prominent submerged shelf characterizes most of the shoreline of Wahoo Lake at present, indicating that an increase in lake level occurred sometime in the past. The lithological transition to black gyttja and increased OM% around 6.3 kcal. BP (Figure 4a) suggests that lake level rose in the middle-Holocene. Increased water depth in the basin would inhibit wind-induced mixing and facilitate lake stratification in summer, as observed in July 2011, which could, in turn, promote anoxic conditions in the eastern basin and the preservation of OM in the sediments. Increased water depth and stratification of the lake water would also explain the decrease in CaCO₃% to low values and the disappearance of Pisidium shells in the basin core. The basal age of the shelf core at 5.3 kcal. BP (Figure 4b) and the unconsolidated gravel immediately beneath the shelf sediments suggest that lake level continued to rise in the middle Holocene and resulted in sediment deposition at the core site by this time. Water levels at the shelf site probably remained low, as inferred from the abundance of moss remains in the sediments. The late-Holocene sediments of the shelf core do not show any indication of a sedimentary hiatus. As at present, the basin was probably topographically and hydrologically open throughout the middle Holocene and late Holocene, and the isotopic values of the inlet and water column samples suggest limited influence of evaporative enrichment.

The change in the hydrological setting at Wahoo Lake (from a closed basin to open basin) suggests that the early-Holocene and late-Holocene δ¹⁸O values may not be directly comparable, as different controls likely influenced the isotopic composition of the lake water. Variations in Pisidium δ¹⁸O values during the late Holocene, when the lake transitioned to an open basin, likely reflect changes in the annual temperature and the seasonality of precipitation. Evaporation may have also played a minor role in the late-Holocene δ¹⁸O values, given the possibility that lake levels fluctuated in response to changes in the steepness of the outflow channel. It is also possible that permafrost thaw and/or supra-permafrost runoff may have also influenced isotopic variability of the late Holocene, although we consider those inputs negligible given the strong link between climate and surface water isotopes in Alaska at present (Lachniet et al., 2016). At 3.5 kcal. BP, CaCO₃% increased, and Pisidium shells were abundant between 3.5 and 0.9 kcal. BP in the shelf core, although no Pisidium shells were present after 0.9 kcal. BP. Shallow water on the shelf likely provided habitats for Pisidium growth as well as minimal carbonate dissolution. The relatively large δ¹⁸O variations (ranging from −18.7‰ to −16.2‰) from 3.5 to 0.9 kcal. BP suggest that the late Holocene was characterized by marked fluctuations in temperature and/or seasonality of precipitation. The increasing trend of δ¹⁸O values from 3.5 to 2.1 kcal. BP, and the δ¹⁸O maxima centered at 0.9 and 1.5 kcal. BP, suggest increased temperatures or enhanced summer relative to winter precipitation. The minima in δ¹⁸O at 1.8 and 1.1 kcal. BP are indicative of periods of cooler temperatures and/or increased winter precipitation.

Open lake basins are generally characterized by lower lake water δ¹⁸O and less pronounced shifts in carbonate δ¹⁸O (Leng and Marshall, 2004) than closed-basin sites in the same area. At Wahoo Lake, the mean δ¹⁸O value of the late Holocene is only 0.4‰ lower than the early-Holocene mean (Figure 4c), despite our interpretation that the lake may have been a topographically closed basin before transitioning to an open basin around 6.3 kcal. BP. Other factors may have differed between these time periods in addition to hydrological setting. For example, Arctic sea-ice extent was reduced in the early Holocene compared with the late Holocene (Dyke and Savelle, 2001; Kaufman et al., 2004), which may have enhanced the influx of ¹⁸O-depleted moisture from the cold Arctic Ocean relative to warmer North Pacific moisture sources (e.g. Klein et al., 2016). Reduced sea ice may have also increased the amount of isotopically lighter winter relative to summer precipitation in the northern foothills of the Brooks Range. Enhanced seasonality could also reduce the mean carbonate δ¹⁸O of the early Holocene through temperature effects, with isotopically lighter meteoric inputs to the basin during colder winters (Dansgaard, 1964) combined with additional ¹⁸O depletion during carbonate precipitation in warmer summers (temperature-dependent fractionation; Craig, 1965). Although we cannot rigorously evaluate the relative roles of these factors in driving the overall δ¹⁸O difference between the early Holocene and late Holocene, our sedimentological indicators provide a framework for interpreting the likely drivers of isotopic variations within each time period.

Comparisons with other records of Holocene climate

A recent synthesis of paleoclimate records from Alaska and adjacent Canada highlights spatial heterogeneity in effective moisture and temperature anomalies during the Holocene (Kaufman et al., 2016). Some of the paleorecords from the region suggest a Holocene Thermal Maximum (HTM) associated with peak summer insolation and pronounced aridity during the early Holocene (Kaufman et al., 2004). However, the signatures of the HTM and early Holocene moisture deficit are not ubiquitous (Kaufman et al., 2016). For example, lake levels were lower during the early Holocene than during the middle Holocene and late Holocene at a number of sites in interior (Abbott et al., 2000; Barber and Finney, 2000; Finkenbinder et al., 2014) and southern Alaska (Anderson et al., 2006; Kaufman et al., 2010), suggesting marked decreases in effective moisture during the HTM. In contrast, other records do not show evidence of the HTM (Clegg et al., 2011; Kaufman et al., 2016) and the timing of early-Holocene aridity was spatially variable (Kaufman et al., 2016). Paleoclimate records from the Alaskan Arctic are scarce (Kaufman et al., 2016), hampering the assessment of the characteristics and drivers of this spatial variability.

At Wahoo Lake, the arid interval with low lake levels between 11.5 and 8.9 kcal. BP, as inferred from lithological and isotope indicators, is broadly consistent with early-Holocene aridity in interior and southern Alaska. This arid interval was followed by a gradual increase in effective moisture from 8.9 to 7.4 kcal. BP (Figure 4), which is also supported by evidence for soil paludification and expansion of dwarf-shrub tundra in the eastern Arctic Foothills ca. 10.0–7.5 kcal. BP (Oswald et al., 2003) and in the northeastern Brooks Range after 8.5 kcal. BP (Oswald et al., 2012). Other published records from the Alaskan Arctic and adjacent Canada also provide evidence of moisture and possibly temperature variations during the early Holocene, although the nature and timing of these fluctuations differ among the records. For example, analyses of floodplain, fluvial, peat, and lacustrine deposits from western and central portions of the North Slope suggest elevated effective moisture and temperature between ca. 11.5 and 9.5 kcal. BP (Mann et al., 2002, 2010). Sedimentary indicators suggest a reduction in lake depth ca. 10.2–9.1 kcal. BP in the western Arctic Foothills (Gaglioti et al., 2014), and chironomid assemblages from Qalluuraq Lake suggest that some of the warmest summer temperatures of the early-Holocene occurred after 10.0 kcal. BP on the western North Slope (Wooller et al., 2012; Figures 1a and 6). In the northern Yukon, chironomid data from Trout Lake suggest elevated summer temperatures from 10.8 to 9.8 kcal. BP (Figures 1a and 6), and shoreline sediments indicate that the lake level was lower by several meters prior to 9.9 kcal. BP (Irvine et al., 2012). Taken together, these records provide evidence for early-Holocene aridity that may have been accompanied by warmer conditions in the Alaskan Arctic.

Figure 6.

Comparison of Pisidium δ¹⁸O values from Wahoo Lake with published paleoclimate records from northern Alaska. Wahoo data plotted at top, with inset to show high-resolution trends and a 500-year loess smooth (bold) for the late-Holocene samples. Black circles and gray squares show data interpreted by original authors as primarily temperature and moisture proxies, respectively. From top to bottom: Pisidium δ¹⁸O (Wahoo Lake, this study), chironomid-based July temperature (Trout Lake; Irvine et al., 2012), Chara δ¹⁸O (Tangled Up Lake; Anderson et al., 2001), June–September temperature anomalies with 1σ range (Kurupa Lake; Boldt et al., 2015), chironomid-based July temperature (Qalluuraq Lake; Wooller et al., 2012), cellulose δ¹⁸O (Meli Lake; Anderson et al., 2001), Chara δ¹⁸O (Takahula Lake; Clegg and Hu, 2010), and Chara δ¹⁸O (Qalluuraq Lake; Wooller et al., 2012).

The sedimentary change to organic-rich gyttja in the basin core from Wahoo Lake, as well as the onset of sediment deposition on the nearby shelf by 5.3 kcal. BP, suggests rising lake levels during the middle Holocene. Similar to our interpretation, rising lake levels during the middle Holocene have been identified in a number of records in other areas of Alaska (Abbott et al., 2000; Barber and Finney, 2000; Edwards et al., 2000; Finkenbinder et al., 2014; Finney et al., 2012; Mann et al., 2002). In addition, decreased evaporation under high moisture conditions were inferred from low δ¹⁸O values in several isotope records (Anderson et al., 2001; Clegg and Hu, 2010; Wooller et al., 2012) from the foothills of the Brooks Range (Figures 1a and 6; Meli and Takahula lakes) and western North Slope (Figures 1a and 6; Qalluuraq Lake), although the timing varies among sites (Figure 6). Evidence of glacial advances in the Brooks Range ca. 5.0 kcal. BP also indicates increased moisture and decreased summer temperature during the middle Holocene (Ellis and Calkin, 1984). Increased effective moisture during the middle Holocene and late Holocene in Alaska has been attributed to increased prevalence of a westerly Aleutian Low, which enhanced winter moisture delivery to the interior (e.g. Anderson et al., 2005; Chipman et al., 2012; Rodionov et al., 2007). Although the Aleutian Low may not directly influence moisture delivery to the northern Brooks Range, atmospheric teleconnections could have indirectly affected climate in this region, given that positive precipitation anomalies in the interior of Alaska are often accompanied by similarly increased precipitation in northern Alaska (Mock et al., 1998). For example, the strengthening of the Aleutian Low in southern Alaska generally coincides with enhanced moisture delivery from the North Pacific relative to the Arctic Ocean in northern Alaska (Klein et al., 2016).

Within the range of chronological uncertainty, the late-Holocene fluctuations in δ¹⁸O at Wahoo Lake display coherent patterns with other late-Holocene reconstructions from northern Alaska. For example, δ¹⁸O increased between 3.5 and 2.1 kcal. BP with generally lower values thereafter at Wahoo Lake (Figure 6). Similarly, elevated summer temperatures from 3.0 to 2.0 kcal. BP, followed by a sharp decline, were inferred from sediment chlorophyll content from Kurupa Lake (Boldt et al., 2015; Figure 6), located in the northern Brooks Range ~320 km west of Wahoo Lake (Figure 1a). In addition, carbonate δ¹⁸O values were elevated between 3.2 and 2.0 kcal. BP at Tangled Up Lake in the southeastern foothills of the Brooks Range (Anderson et al., 2001; Figures 1a and 6), suggesting warmer summer temperatures followed by cooling until ca. 1.0 kcal. BP. The late-Holocene intervals of low δ¹⁸O at Wahoo Lake also coincide with periods of glacial advance in Alaska, implying that the late-Holocene δ¹⁸O minima were related to a combination of low temperatures and enhanced winter relative to summer precipitation. For example, the overall low δ¹⁸O values at Wahoo Lake ca. 3.5–2.5 kcal. BP (Figure 6, inset) coincide with episodes of widespread mountain glacier expansion in Alaska ca. 3.3–2.9 kcal. BP (Solomina et al., 2015). Several distinct δ¹⁸O troughs at Wahoo Lake also broadly correspond to glacial advances in the Brooks Range ca. 3.2–3.3, 2.6–2.7, and 1.0 kcal. BP (Badding et al., 2013; Evison et al., 1996; Pendleton, 2013; Solomina and Calkin, 2003), based on lichen and ¹⁰Be-derived moraine ages. These records together suggest marked late-Holocene variations in temperature, precipitation seasonality, runoff budgets, and/or permafrost thaw in the Alaskan Arctic.

Solar irradiance as a possible driver of Holocene climate variability

Suborbital-scale variations in solar irradiance have been linked to Holocene climatic variability in paleorecords from high-latitude regions of the North Pacific (Anchukaitis et al., 2013; Clegg et al., 2011; Hu et al., 2003; Tinner et al., 2015) and North Atlantic (Bond et al., 2001; Jiang et al., 2015; Moffa-Sánchez et al., 2014), as well as low-latitude to mid-latitude regions around the world (Fleitmann et al., 2003; Gupta et al., 2005; Hodell et al., 2001; Poore et al., 2004; Wang et al., 2005). We compare our data from Wahoo Lake with a Holocene time series of total solar irradiance (TSI) reconstructed from ¹⁰Be in Antarctic and Greenland ice cores as well as ¹⁴C from tree rings (Steinhilber et al., 2012). The δ¹⁸O record from Wahoo Lake is not correlated with this TSI record, which is not surprising given that the environmental controls of δ¹⁸O at this site are nonstationary and that our δ¹⁸O record is discontinuous and coarsely resolved in the early Holocene (Figure 6). However, temporal variations in OM% in the Holocene sediments of Wahoo Lake display broad similarities with those of the TSI record (Figure 7). Notably, both TSI and OM% are generally lower in the early Holocene and middle Holocene than in the late Holocene, and some of the individual minima and maxima align between the two time series, with a moderate correlation (r = 0.55, p < 0.001) at multi-centennial timescales.

Figure 7.

Organic matter percentage (OM%; black curve) of the basin-core sediments from Wahoo Lake in comparison with total solar irradiance anomalies (Steinhilber et al., 2012; gray curve). Both records were smoothed with a 500-year loess window; circles show raw irradiance data.

The paleoenvironmental significance of this correlation is difficult to decipher, in part because sedimentary OM% has numerous controls, including aquatic and terrestrial productivity, OM preservation in sediment, the relative amount of terrestrial mineral input, and in-lake production and preservation of carbonate. However, several previous studies have documented broad coherency in the temporal patterns of solar variability and inferred lake productivity. For example, Hu et al. (2003) and Gavin et al. (2011) identified relationships between biogenic silica (a proxy of lake productivity) and solar forcing at high-latitude lakes in North America. These records suggest that variability in solar irradiance indirectly affects lake productivity through climatic impacts, such as temperature-driven and moisture-driven changes in nutrient delivery from the watershed and/or variations in the duration of seasonal ice cover. Our OM% record from Wahoo Lake may reflect similar responses of aquatic productivity to solar-induced climate change.

The linkage of Holocene climate change with solar irradiance remains a controversial topic. Nonetheless, possible evidence has emerged from a number of paleorecords from Alaska (e.g. Anchukaitis et al., 2013; Barclay et al., 2013; Davi et al., 2003; Hu et al., 2003; Tinner et al., 2008, 2015; Wiles et al., 2004). For example, neoglacial advances corresponded to multi-decadal periods of low solar activity, suggesting that variability of solar irradiance exerted a primary control on glacial advance and retreat during the late Holocene (Solomina et al., 2015; Wiles et al., 2004). Pronounced shifts in aquatic and terrestrial ecosystems appear to coincide with fluctuations in solar irradiance during the Holocene (Douarin et al., 2016; Hu et al., 2003; Katsuki et al., 2014; Schwörer et al., 2016; Tinner et al., 2015). Although the mechanism of such linkages remains unclear (Gray et al., 2010; Wanner et al., 2008), both modeling and paleodata have identified suborbital solar variability as a key influence on large-scale atmospheric circulation patterns, such as Arctic Oscillation/North Atlantic Oscillation (AO/NAO) and El Niño-Southern Oscillation (ENSO; Gray et al., 2010; Shindell et al., 2001, 2003). Given the complexity and nonlinearity of climate responses to radiative forcing (Clegg et al., 2011; Hansen et al., 1997; Meehl et al., 2003), the accumulation of evidence linking high-latitude paleoclimate records to solar variation is intriguing. These potential linkages warrant further research in order to understand this component of natural climate variability, which is essential in anticipating how anthropogenic and natural forcings will interact to affect the stability and sensitivity of the high-latitude climate system.

Footnotes

Acknowledgements

We thank T Brown, S Fanta, and C Eastoe for radiocarbon, carbonate-isotope, and water-isotope analyses, respectively. We also thank P Higuera, R Kelly, and A Young for field assistance, and B Clegg, D Devotta, M Fernandez, and J Napier for thoughtful comments on the manuscript. RSV thanks K Howell, T Kamp, M Marti, D Reinhard, C Stewart, and E Sharton-Bierig for support.

Funding

Funding for this research was provided by NSF grant ARC-1023477 to FSH, an EPA STAR Fellowship to MLC, and a UIUC Department of Geology Undergraduate Research Award to RSV. All data from this study are available from the NOAA’s National Centers for Environmental Information ().

References

Abbott

Finney

Edwards

, et al. (2000) Lake-level reconstruction and paleohydrology of Birch Lake, central Alaska, based on seismic reflection profiles and core transects. Quaternary Research 53(2): 154–166.

Anchukaitis

D’Arrigo

Andreu-Hayles

, et al. (2013) Tree-ring-reconstructed summer temperatures from northwestern North America during the last nine centuries. Journal of Climate 26(10): 3001–3012.

Anderson

Abbott

Finney

(2001) Holocene climate inferred from oxygen isotope ratios in lake sediments, central Brooks Range, Alaska. Quaternary Research 55(3): 313–321.

Anderson

Abbott

Finney

, et al. (2005) Regional atmospheric circulation change in the North Pacific during the Holocene inferred from lacustrine carbonate oxygen isotopes, Yukon Territory, Canada. Quaternary Research 64(1): 21–35.

Anderson

Birks

Rover

, et al. (2013) Controls on recent Alaskan lake changes identified from water isotopes and remote sensing. Geophysical Research Letters 40(13): 3413–3418.

Anderson

Hallet

Berg

, et al. (2006) Holocene development of boreal forests and fire regimes on the Kenai Lowlands of Alaska. The Holocene 16(6): 791–803.

Apolinarska

Hammarlund

(2009) Multi-component stable isotope records from Late Weichselian and early Holocene lake sediments at Imiołki, Poland: Palaeoclimatic and methodological implications. Journal of Quaternary Science 24(8): 948–959.

Badding

Briner

Kaufman

(2013) ¹⁰Be ages of late Pleistocene deglaciation and Neoglaciation in the north-central Brooks Range, Arctic Alaska. Journal of Quaternary Science 28(1): 95–102.

Barber

Finney

(2000) Late Quaternary paleoclimatic reconstructions for interior Alaska based on paleolake-level data and hydrologic models. Journal of Paleolimnology 24(1): 29–41.

10.

Barclay

Yager

Graves

, et al. (2013) Late-Holocene glacial history of the Copper River Delta, coastal south-central Alaska, and controls on valley glacier fluctuations. Quaternary Science Reviews 81: 74–89.

11.

Beikman

(1980) Geological map of Alaska. US Geological Survey Map, scale 1:2,500,000. Available at: https://mrdata.usgs.gov/ak/geology/.

12.

Binford

(1990) Calculation and uncertainty analysis of ²¹⁰Pb dates for PIRLA project lake sediment cores. Journal of Paleolimnology 3(3): 253–267.

13.

Blaauw

(2010) Methods and code for ‘classical’ age-modelling of radiocarbon sequences. Quaternary Geochronology 5(5): 512–518.

14.

Boldt

Kaufman

McKay

, et al. (2015) Holocene summer temperature reconstruction from sedimentary chlorophyll content, with treatment of age uncertainties, Kurupa Lake, Arctic Alaska. The Holocene 25: 641–650.

15.

Bond

Kromer

Beer

, et al. (2001) Persistent solar influence on North Atlantic climate during the Holocene. Science 294(5549): 2130–2136.

16.

Bowen

(2013) The online isotopes in precipitation calculator, version 2.2. Available at: http://www.waterisotopes.org.

17.

Bowen

Revenaugh

(2003) Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39(10). DOI: 10.1029/2003WR002086.

18.

Bowen

Wilkinson

(2002) Spatial distribution of δ¹⁸O in meteoric precipitation. Geology 30(4): 315–318.

19.

Carson

(2006) Surficial-Geologic Map of the Kavik River Area, West-Central Mount Michelson Quadrangle, Northeastern Brooks Range, Alaska. Report of investigation RI 2009-3. Fairbanks, AK: Alaska Division of Geological & Geophysical Surveys.

20.

Chipman

Clegg

(2012) Variation in the moisture regime of northeastern interior Alaska and possible linkages to the Aleutian Low: Inferences from a late-Holocene δ¹⁸O record. Journal of Paleolimnology 48(1): 69–81.

21.

Chipman

Kling

Lundstrom

, et al. (2016) Multiple thermo-erosional episodes during the past six millennia: Implications for the response of Arctic permafrost to climate change. Geology 44(6): 439–442.

22.

Clegg

(2010) An oxygen-isotope record of Holocene climate change in the south-central Brooks Range, Alaska. Quaternary Science Reviews 29(7): 928–939.

23.

Clegg

Kelly

Clarke

, et al. (2011) Nonlinear response of summer temperature to Holocene insolation forcing in Alaska. Proceedings of the National Academy of Sciences 108(48): 19299–19304.

24.

Craig

(1965) The measurement of oxygen isotope paleotemperatures. In: Tongiori

(ed.) Stable Isotopes in Oceanographic Studies and Paleotemperatures. Pisa: Laboratorio di Geologia Nucleare, pp. 26–27.

25.

Dansgaard

(1964) Stable isotopes in precipitation. Tellus 16(4): 436–468.

26.

Davi

Jacoby

Wiles

(2003) Boreal temperature variability inferred from maximum latewood density and tree-ring width data, Wrangell Mountain region, Alaska. Quaternary Research 60(3): 252–262.

27.

Douarin

Elliot

Noble

, et al. (2016) North Atlantic ecosystem sensitivity to Holocene shifts in meridional overturning circulation. Geophysical Research Letters 43(1): 291–298.

28.

Dyke

Savelle

(2001) Holocene history of the Bering Sea bowhead whale (Balaena mysticetus) in its Beaufort Sea summer grounds off southwestern Victoria Island, western Canadian Arctic. Quaternary Research 55: 371–379.

29.

Eakins

Morrison

(1978) A new procedure for the determination of lead-210 in lake and marine sediments. The International Journal of Applied Radiation and Isotopes 29(9): 531–536.

30.

Edwards

Bigelow

Finney

, et al. (2000) Records of aquatic pollen and sediment properties as indicators of late-Quaternary Alaskan lake levels. Journal of Paleolimnology 24(1): 55–68.

31.

Ellis

Calkin

(1984) Chronology of Holocene glaciation, central Brooks range, Alaska. The Geological Society of America Bulletin 95(8): 897–912.

32.

Environmental Data Center Team (2015) Meteorological Monitoring Program at Toolik, Alaska. Fairbanks, AK: Toolik Field Station, Institute of Arctic Biology, University of Alaska Fairbanks. Available at: http://toolik.alaska.edu/edc/abiotic_monitoring/data_query.php.

33.

Evison

Calkin

Ellis

(1996) Late-Holocene glaciation and twentieth-century retreat, northeastern Brooks Range, Alaska. The Holocene 6(1): 17–24.

34.

Finkenbinder

Abbott

Edwards

, et al. (2014) A 31,000 year record of paleoenvironmental and lake-level change from Harding Lake, Alaska, USA. Quaternary Science Reviews 87: 98–113.

35.

Finney

Bigelow

Barber

, et al. (2012) Holocene climate change and carbon cycling in a groundwater-fed, boreal forest lake: Dune Lake, Alaska. Journal of Paleolimnology 48(1): 43–54.

36.

Fleitmann

Burns

Mudelsee

, et al. (2003) Holocene forcing of the Indian monsoon recorded in a stalagmite from southern Oman. Science 300(5626): 1737–1739.

37.

Fritz

Poplawski

(1974) ¹⁸O and ¹³C in the shells of freshwater molluscs and their environments. Earth and Planetary Science Letters 24(1): 91–98.

38.

Gaglioti

Mann

Jones

, et al. (2014) Radiocarbon age-offsets in an arctic lake reveal the long-term response of permafrost carbon to climate change. Journal of Geophysical Research: Biogeosciences 119(8): 1630–1651.

39.

Gavin

Henderson

Westover

, et al. (2011) Abrupt Holocene climate change and potential response to solar forcing in western Canada. Quaternary Science Reviews 30(9): 1243–1255.

40.

Gibson

Edward

TWD

(2002) Regional water balance trends and evaporation-transpiration partitioning from a stable isotope survey of lakes in northern Canada. Global Biogeochemical Cycles 16(2): 10-1–10-14.

41.

Gray

Beer

Geller

, et al. (2010) Solar influences on climate. Reviews of Geophysics 48(4). DOI: 10.1029/2009RG000282.

42.

Gupta

Das

Anderson

(2005) Solar influence on the Indian summer monsoon during the Holocene. Geophysical Research Letters 32(17). DOI: 10.1029/2005GL022685.

43.

Hammarlund

Edwards

Björck

, et al. (1999) Climate and environment during the Younger Dryas (GS-1) as reflected by composite stable isotope records of lacustrine carbonates at Torreberga, southern Sweden. Journal of Quaternary Science 14: 17–28.

44.

Hansen

Sato

Ruedy

(1997) Radiative forcing and climate response. Journal of Geophysical Research: Atmospheres 102(D6): 6831–6864.

45.

Hodell

Brenner

Curtis

, et al. (2001) Solar forcing of drought frequency in the Maya lowlands. Science 292(5520): 1367–1370.

46.

Holopainen

Hanski

(1986) Life history variation in Pisidium (Bivalvia: Pisidiidae). Ecography 9(2): 85–98.

47.

Horita

Wesolowski

(1994) Liquid-vapor fractionation of oxygen and hydrogen isotopes of water from the freezing to the critical temperature. Geochimica et Cosmochimica Acta 58(16): 3425–3437.

48.

Higuera

Duffy

, et al. (2015) Arctic tundra fires: Natural variability and responses to climate change. Frontiers in Ecology and the Environment 13(7): 369–377.

49.

Kaufman

Yoneji

, et al. (2003) Cyclic variation and solar forcing of Holocene climate in the Alaskan subarctic. Science 301(5641): 1890–1893.

50.

Hunter

(1964) Physiological aspects of ecology in non-marine molluscs. In: Wilbur

Yonge

(eds) Physiology of Mollusca, Pt. 1. New York: Academic Press, pp. 83–116.

51.

IAEA/WMO (2014) Global network of isotopes in precipitation. The GNIP database. Available at: http://www.iaea.org/water.

52.

IPCC (2013) Climate Change 2013 Working Group 1: The Physical Science Basis – Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (ed TF Stocker, D Qin, GK Plattner et al , et al.). Cambridge: Cambridge University Press.

53.

Irvine

Cwynar

Vermaire

, et al. (2012) Midge-inferred temperature reconstructions and vegetation change over the last~ 15,000 years from Trout Lake, northern Yukon Territory, eastern Beringia. Journal of Paleolimnology 48(1): 133–146.

54.

Ito

(2001) Application of stable isotope techniques to inorganic and biogenic carbonates. In: Last

Smol

(eds) Tracking Environmental Change Using Lake Sediments. Dordrecht: Springer, pp. 351–371.

55.

Jiang

Muscheler

Björck

, et al. (2015) Solar forcing of Holocene summer sea-surface temperatures in the northern North Atlantic. Geology 43(3): 203–206.

56.

Katsuki

Itaki

Khim

, et al. (2014) Response of the Bering Sea to 11-year solar irradiance cycles during the Bølling-Allerød. Geophysical Research Letters 41(8): 2892–2898.

57.

Kaufman

Ager

Anderson

, et al. (2004) Holocene thermal maximum in the western Arctic (0–180°W). Quaternary Science Reviews 23(5): 529–560.

58.

Kaufman

Anderson

, et al. (2010) Evidence for a variable and wet Younger Dryas in southern Alaska. Quaternary Science Reviews 29(11–12): 1445–1452.

59.

Kaufman

Axford

Henderson

, et al. (2016) Holocene climate changes in eastern Beringia (NW North America) – A systematic review of multi-proxy evidence. Quaternary Science Reviews 147: 312–339.

60.

Kaufman

Schneider

McKay

, et al. (2009) Recent warming reverses long-term Arctic cooling. Science 325(5945): 1236–1239.

61.

Kaufman

Young

Briner

, et al. (2011) Alaska Paleo glacier Atlas (version 2). Developments in Quaternary Sciences 15: 427–445.

62.

Klein

Nolan

McConnell

, et al. (2016) McCall Glacier record of Arctic climate change: Interpreting a northern Alaska ice core with regional water isotopes. Quaternary Science Reviews 131: 274–284.

63.

Lachniet

Lawson

Stephen

, et al. (2016) Isoscapes of δ¹⁸O and δ²H reveal climatic forcings on Alaska and Yukon precipitation. Water Resources Research 52: 6575–6586.

64.

Leng

Marshall

(2004) Palaeoclimate interpretation of stable isotope data from lake sediment archives. Quaternary Science Reviews 23(7): 811–831.

65.

L’Heureux

Mann

Cook

, et al. (2004) Atmospheric circulation influences on seasonal precipitation patterns in Alaska during the latter 20th century. Journal of Geophysical Research: Atmospheres 109(D6). DOI: 10.1029/2003JD003845.

66.

(1997) δ¹³C–δ¹⁸C covariance as a paleohydrological indicator for closed-basin lakes. Palaeogeography, Palaeoclimatology, Palaeoecology 133(1): 69–80.

67.

McGuire

McDonnell

Weiler

, et al. (2005) The role of topography on catchment-scale water residence time. Water Resources Research 41(5). DOI: 10.1029/2004WR003657.

68.

Mann

Groves

Reanier

, et al. (2010) Floodplains, permafrost, cottonwood trees, and peat: What happened the last time climate warmed suddenly in arctic Alaska? Quaternary Science Reviews 29(27): 3812–3830.

69.

Mann

Peteet

Reanier

, et al. (2002) Responses of an arctic landscape to Lateglacial and early Holocene climatic changes: The importance of moisture. Quaternary Science Reviews 21(8): 997–1021.

70.

Meehl

Washington

Wigley

TML

, et al. (2003) Solar and greenhouse gas forcing and climate response in the twentieth century. Journal of Climate 16(3): 426–444.

71.

Mock

Bartlein

Anderson

(1998) Atmospheric circulation patterns and spatial climatic variations in Beringia. International Journal of Climatology 18(10): 1085–1104.

72.

Moffa-Sánchez

Born

Hall

, et al. (2014) Solar forcing of North Atlantic surface temperature and salinity over the past millennium. Nature Geoscience 7(4): 275–278.

73.

Oldfield

Appleby

Battarbee

(1978) Alternative ²¹⁰Pb dating: Results from the New Guinea highlands and Lough Erne. Nature 271(5643): 339–342.

74.

Osterkamp

(2007) Characteristics of the recent warming of permafrost in Alaska. Journal of Geophysical Research 112: F02S02.

75.

Oswald

Anderson

Brown

, et al. (2005) Effects of sample mass and macrofossil type on radiocarbon dating of arctic and boreal lake sediments. The Holocene 15(5): 758–767.

76.

Oswald

Brubaker

, et al. (2003) Holocene pollen records from the central Arctic Foothills, northern Alaska: Testing the role of substrate in the response of tundra to climate change. Journal of Ecology 91(6): 1034–1048.

77.

Oswald

Gavin

Anderson

, et al. (2012) A 14,500-year record of landscape change from Okpilak Lake, northeastern Brooks Range, northern Alaska. Journal of Paleolimnology 48(1): 101–113.

78.

Pendleton

(2013) 10Be chronology of late Pleistocene and Holocene glaciation of the Alapah Mountain and Arrigetch Peaks areas, central Brooks Range, Alaska. MS Thesis, University at Buffalo.

79.

Poore

Quinn

Verardo

(2004) Century-scale movement of the Atlantic Intertropical Convergence Zone linked to solar variability. Geophysical Research Letters 31(12). DOI: 10.1029/2004GL019940.

80.

Reimer

Bard

Bayliss

, et al. (2013) IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55(4): 1869–1887.

81.

Rodionov

Bond

Overland

(2007) The Aleutian Low, storm tracks, and winter climate variability in the Bering Sea. Deep Sea Research Part II: Topical Studies in Oceanography 54(23): 2560–2577.

82.

Schwörer

Gavin

Walker

, et al. (2016) Holocene tree line changes in the Canadian Cordillera are controlled by climate and topography. Journal of Biogeography. Epub ahead of print 2 December. DOI: 10.1111/jbi.12904.

83.

Serreze

Barry

(2011) Processes and impacts of Arctic amplification: A research synthesis. Global and Planetary Change 77(1): 85–96.

84.

Shindell

Schmidt

Miller

, et al. (2001) Northern Hemisphere Winter Climate Response to Greenhouse Gas, Ozone, Solar and Volcanic Forcing. Report, NASA Goddard Institute for Space Studies, New York, 1 August.

85.

Shindell

Schmidt

Miller

, et al. (2003) Volcanic and solar forcing of climate change during the preindustrial era. Journal of Climate 16(24): 4094–4107.

86.

Solomina

Calkin

(2003) Lichenometry as applied to moraines in Alaska. Arctic, Antarctic, and Alpine Research 35: 129–143.

87.

Solomina

Bradley

Hodgson

, et al. (2015) Holocene glacier fluctuations. Quaternary Science Reviews 111: 9–34.

88.

Steinhilber

Abreu

Beer

(2012) 9,400 years of cosmic radiation and solar activity from ice cores and tree rings. Proceedings of the National Academy of Sciences 109(16): 5967–5971.

89.

Stone

Dutton

Harris

(2002) Earlier spring snowmelt in northern Alaska as an indicator of climate change. Journal of Geophysical Research: Atmospheres 107(D10): ACL 10-1–ACL 10-13.

90.

Stuiver

Reimer

(1993) Extended ¹⁴C database and revised CALIB radiocarbon calibration program. Radiocarbon 35(1): 215–230.

91.

Talbot

(1990) A review of the palaeohydrological interpretation of carbon and oxygen isotopic ratios in primary lacustrine carbonates. Chemical Geology: Isotope Geoscience Section 80(4): 261–279.

92.

Thorp

Covich

(2009) Ecology and Classification of North American Freshwater Invertebrates. San Diego, CA: Academic Press.

93.

Tinner

Beer

Bigler

, et al. (2015) Late-Holocene climate variability and ecosystem responses in Alaska inferred from high-resolution multiproxy sediment analyses at Grizzly Lake. Quaternary Science Reviews 126: 41–56.

94.

Tinner

Bigler

Gedye

, et al. (2008) A 700-year paleoecological record of boreal ecosystem responses to climatic variation from Alaska. Ecology 89(3): 729–743.

95.

Turner

Wolfe

Edwards

(2010) Characterizing the role of hydrological processes on lake water balances in the Old Crow Flats, Yukon Territory, Canada, using water isotope tracers. Journal of Hydrology 386(1): 103–117.

96.

US Geological Survey (2012) National Land Cover Alaska – National Geospatial Data Asset (NGDA) Land Use Land Cover. Sioux Falls, SD: US Geological Survey.

97.

Von Grafenstein

Erlernkeuser

Trimborn

(1999) Oxygen and carbon isotopes in modern fresh-water ostracod valves: Assessing vital offsets and autecological effects of interest for palaeoclimate studies. Palaeogeography, Palaeoclimatology, Palaeoecology 148(1): 133–152.

98.

Wang

Cheng

Edwards

, et al. (2005) The Holocene Asian monsoon: Links to solar changes and North Atlantic climate. Science 308(5723): 854–857.

99.

Wanner

Beer

Bütikofer

, et al. (2008) Mid- to late-Holocene climate change: An overview. Quaternary Science Reviews 27(19): 1791–1828.

100.

Wiles

D’Arrigo

Villalba

, et al. (2004) Century-scale solar variability and Alaskan temperature change over the past millennium. Geophysical Research Letters 31(15).DOI: 10.1029/2004GL020050.

101.

Wooller

Pohlman

Gaglioti

, et al. (2012) Reconstruction of past methane availability in an Arctic Alaska wetland indicates climate influenced methane release during the past~ 12,000 years. Journal of Paleolimnology 48(1): 27–42.

102.

Brock

Falcone

, et al. (2008) A coupled isotope tracer method to characterize input water to lakes. Journal of Hydrology 350(1): 1–13.

103.

Zhang

Osterkamp

Stamnes

(1996) Some characteristics of the climate in northern Alaska, USA. Arctic and Alpine Research 28(4): 509–518.