Changes in glacial meltwater alter algal communities in lakes of Scoresby Sund,Renland,East Greenland throughout the Holocene: Abrupt reorganizations began 1000 years before present

Abstract

We investigated the response of lake algal communities to changes in glacial meltwater from the Renland Ice Cap (Greenland) through the Holocene to assess whether influxes always elicit consistent responses or novel responses. We measured sedimentary algal pigments in two proximal lakes, snow-fed Raven and glacier- and snow-fed Bunny Lake, and diatom community structure and turnover in Bunny Lake. Diatom data were not available in Raven Lake. We also modeled lake-level change in Bunny Lake to identify how glacial meltwater may have altered diatom habitat availability through time. Through a series of glacier advances and retreats over the Holocene, the algal response in Bunny Lake was relatively constant until approximately 1015 yr BP, after which there were major changes in sedimentary algal remains. Algal pigment concentrations sharply declined, and diatom species richness increased. Diatom community structure underwent three reorganizations. Until 1015 yr BP, assemblages were dominated by Pinnularia braunii and Aulacoseira pffaffiana. However, approximately 1015–480 yr BP, these species declined and Tabellaria flocculosa and Hannaea arcus became a significant component of the assemblage. Approximately 440 yr BP, A. pfaffiana increased along with species indicating elevated nitrogen. In contrast, the algal pigment records from nearby snow-fed Raven Lake showed different and minimal change through time. Our results suggest that changes in the magnitude and composition of meltwater in our two study lakes were unique over the last 1000 yr BP and elicited a non-linear threshold response absent during other periods of glacier advance and retreat. Deciphering the degree to which glaciers structure algal communities over time has strong implications for lakes as glaciers continue to recede.

Keywords

climate diatom East Greenland glacier Holocene lake level

Introduction

Glacial recession sets in motion a series of gradual geomorphic and biogeochemical responses across landscapes (Engstrom et al., 2000; Fritz and Anderson, 2013). However, changes in the extent and magnitude of glacial meltwater can also abruptly and dramatically alter lake communities, eliciting a non-linear, threshold response where small change can elicit rapid responses. Throughout the Holocene, glaciers have repeatedly expanded and receded in alpine and Arctic regions potentially altering downstream aquatic ecosystems. Evidence of abrupt or threshold effects on glacially fed alpine lakes is based on shifts in diatom assemblages in paleolimnological records (Slemmons et al., 2015) and in streams based on changes in water chemistry over short time scales (Perić et al., 2015), but an understanding of effects on Arctic lake biota over long time scales is limited.

Owing to low primary productivity rates and dilute water chemistry, high latitude lakes are particularly sensitive to environmental change. Given the oligotrophic nature of these lakes, and narrow thermal and nutrient tolerances of aquatic species, slight fluctuations in temperature and nutrient deposition can lead to marked change. Increasing air temperatures can lead to earlier ice off, later winter freezing (Magnuson et al., 2000), and altered lake mixing patterns (Straile et al., 2003), with subsequent increases in turnover in phytoplankton communities (Perren et al., 2009; Smol et al., 2005). In addition to temperature-related mechanisms, aquatic communities are shaped by the interaction of a wide range of chemical, physical, and ecological processes, ultimately influenced by landscape features, glacial history, disturbance events, and seasonality (Chapin and Körner, 1996; Jacobsen and Dangles, 2012).

Glaciers are landscape features, common in the Arctic, which may potentially amplify or dampen climate-driven change in aquatic ecosystems receiving glacier meltwater (Vincent et al., 2011). Glacier ablation rates, largely controlled by climatic factors such as summer temperatures and precipitation, influence the quantity of runoff to lakes. Comparisons of nitrogen (Robinson and Kawecka, 2005; Saros et al., 2010; Slemmons and Saros, 2012; Williams et al., 2007), dissolved organic carbon (Hood et al., 2009; Singer et al., 2012), persistent organic pollutants (Bizzotto et al., 2009; Blais et al., 2001; Bogdal et al., 2009), and eolian dust concentrations (Willemse et al., 2003) in glacier- and snow-fed (GSF) lakes reveal sharp differences between lake types. These differences are often dictated by atmospheric deposition rates, presence or absence of landscape features, microbial activity, and wind speed and direction; all potentially altering the particle and solute composition of runoff to these systems. As a result, lakes that are glacier-fed are often ecologically different from those that are fed by snow only (SF; Robinson and Kawecka, 2005; Slemmons and Saros, 2012; Slemmons et al., 2015), with climate change altering not only thermal-related processes within the lake, but also altering the quantity, duration, and composition of meltwater. Such changes create a high level of disturbance eliciting shifts in algal community structure (Karabanov et al., 2004). Increases in glacial meltwater may elicit shifts in the physical and chemical features of aquatic ecosystems, producing turbid, cold, and harsh environments relative to SF systems (Uehlinger et al., 2010), but the biological response to changes in glacial inputs over long temporal scales remains unclear. Furthermore, lake ontogeny may differ between GSF and SF lakes where primary production in non-glaciated lakes is controlled largely by terrestrial vegetation successional dynamics (Engstrom et al., 2000) compared with climate (e.g. ice cover) in glaciated regions (Michelutti et al., 2007).

In this study, we examined the response of algae to variations in glacial meltwater in Bunny Lake (GSF) in the Renland region of Scoresby Sund in East Greenland over the past 8000 years. Due to hydrologic and ice dynamics, Bunny Lake has experienced abrupt shifts in glacial meltwater input through the Holocene (Medford, 2013; Medford et al., in preparation), making this a model system for investigating the ecological effects of changing glacial meltwater inputs to an Arctic lake over the Holocene. We assessed sedimentary diatom assemblage structure and species turnover as well as fossil algal pigments in Bunny Lake, and compared this with a proximal, SF lake, Raven Lake. To assess habitat change driven by influx of glacial meltwater, we conducted lake-level modeling using fossil planktic to benthic diatom ratios.

Materials and methods

Study site

Situated along the coastline of East Greenland (~70.5–71.5°N, 24–28°W), Scoresby Sund (~69–72°N, 21–30°W), near the Renland Ice Cap, is the largest fjord system along East Greenland. Renland (~70.5–71.5°N, 24–28°W), west and inland of Scoresby Sund, contains numerous lakes, many of which are above ~1500 m in elevation (Figure 1). Evidence from ice-core, lacustrine, and marine sediment records suggests that this region of East Greenland exhibited an early-Holocene climatic optimum approximately 10,700 years ago (Funder, 1978; Hjort and Funder, 1974; Koc et al., 1993; Levy et al., 2014; Marienfeld, 1991; Nam, 1997; Notholt, 1998; Vinther et al., 2008). Based on inferred glacial history from lake sediment records in the Scoresby Sund region, there were minor fluctuations of glaciers periodically throughout the Holocene, with a greater extent in glaciation in the late-Holocene (Levy et al., 2014; Lowell et al., 2013; Medford, 2013).

Figure 1.

Location of study lakes in Renland, Scoresby Sund region, East Greenland (black box). Gray box indicates the location of the Renland ice core.

Bunny Lake is a small lake (0.06 km²; Table 1) in a sequence of lakes receiving meltwater from a peripheral glacier (approximately 10 km away) of the Renland Ice Cap and an adjacent small ice cap (Figure 1). This lake has experienced abrupt shifts in glacial meltwater input through the Holocene (Medford, 2013). During glacier advance, an ice dam forms in the east end of Catalina Lake, causing glacial runoff to be diverted toward Bunny Lake. During periods of retreat, the ice dam is absent and, due to the topographic slope of the land, water flows away from Bunny Lake leaving it without glacier meltwater (Figure 2, see Medford, 2013 for a full description of this conceptual model). In addition to runoff from the Renland Ice Cap, Bunny Lake is surrounded by plateaus that although snow-free today in summer with approximately a 7-week ice-free period, may have supported small ice caps during cold periods and thus provided additional sources of meltwater and sediment. Glaciers in Scoresby Sund reached their maximum Holocene extent during the ‘Little Ice Age’ (LIA) (Hall et al., 2008; Kelly et al., 2008; Lowell et al., 2013) and are extensive enough at present to maintain an ice dam. Consequently, Bunny Lake currently receives glacial meltwater and is at its maximum possible lake level given the elevation of the sill that controls outflow. It is assumed that the current level is maintained by glacial meltwater input.

Table 1.

Select characteristics of the study lakes in Scoresby Sund region of East Greenland based on 2011 sampling.

	Bunny Lake (GSF)	Raven Lake (SF)
Latitude (°N)	71.03	71.07
Longitude (°W)	27.43	27.31
Elevation (m a.s.l.)	819	1054
Lake area (km²)	0.06	0.19
Z_max (m)	14	20
Conductivity (µS cm⁻¹)	4.6	2.9

GSF: glacier-fed, SF: snow-fed.

Figure 2.

Conceptual model of glacial meltwater flow to Bunny Lake. Today (and during periods of glacial advance), an ice dam causes overspill of Catalina Lake, to the west, through the Bunny Lake chain. Dashed lines are estimates of the maximum LIA-GA extent which causes an overflow of meltwater from Catalina Lake into Bunny and represents the current flow of meltwater; black circles represent position of inferred location of plateau glaciers, which may have been an additional source of meltwater during the LIA-GA. Black arrow represents the direction of meltwater flow during the LIA-GA.

Glacial advance (GA) inferred from the abundance of rock flour (silt- and clay-rich inorganic sediment) began shortly after 1015 yr BP; this sediment persisted until ~480 yr BP when it was interrupted by an influx of sand (Medford, 2013; Medford et al., in preparation). The precise origin of the sand is uncertain, but it may relate to intensification of glaciation and the development of local ice caps on the plateaus immediately above the lake (Medford et al., in preparation). A return to sediment dominated by rock flour occurred after 265 yr BP, but the exact time of this transition is not well-dated. For the purposes of this study, we refer to the earlier part of the sediment core record, from 1015 to 480 yr BP, as a significant GA. Part of the sediment core record (post-480 yr BP) is dominated by sand influx and broadly overlaps with the ‘Little Ice Age’. We refer to this period as the ‘Little Ice Age’ Glacial Advance (LIA-GA). The nearby groundwater- and snow-fed lake Raven Lake (Table 1) serves as a glacier-free reference lake (see Medford, 2013 for a full description of this reference lake).

Core collection, dating, and characterization

To assess how algal communities have responded to Holocene changes in glacial meltwater in this region, sediment cores were extracted from the deepest area of each lake using a piston coring system during August, 2011. For Bunny Lake, we analyzed two overlapping sediment cores (BNL11-1A-1: ~125 cm length and BNL11-1B-1: ~78 cm length) from the south basin, which together span more than 8000 years. For Raven Lake, we analyzed one sediment core (RAV11-1A-1: ~153 cm length) from the deepest basin. This core spans at least 11,000 years. Core stratigraphy, grain size analysis, magnetic susceptibility, percentage organic composition, and age model results can be found in Medford (2013), with pertinent details on Figure 3a. Periods of glacial/non-glacial activity were identified based on the sedimentary record (Medford, 2013; Medford et al., in preparation).

Figure 3.

(a) Biogeochemical stratigraphy of Bunny Lake. Diatom abundance is measured as valves × 10⁶ per gram of sediment. Species richness values are the deviation from the average richness. Detrended correspondence analysis (DCA) of the entire diatom assemblage. Planktic to benthic ratio includes meroplanktic species such as Aulacoseira sp. in the planktic calculation. For both Figure (a) and (b) and henceforth, the LIA-GA is represented by the light gray bar (50–480 yr BP) and the GA by the dark gray bar (480–1015 yr BP). (b) Diatom stratigraphy of Bunny Lake, a glacially fed lake, ordered by appearance in the record and showing the relative frequencies of species composing at least 10% of the assemblage at a given depth or the relative frequencies of species of ecological interest. The brackets indicate the presence of Fragilaria tenera (morphotype 2); all remaining data points represent Fragilaria tenera (morphotype 1).

Cores were dated using macrovegetation (BNL11-1A-1: 37.4 cm; BNL11-1B-1: 77 cm; RAV11-1A-1: 56 cm) and sieved organic matter (BNL11-1A-1: 23, 61, 107 cm; BNL11-1B-1: 56 cm; RAV11-1A-1: 9, 24, 111, 125.5 cm) (Medford, 2013; Medford et al., in preparation) analyzed at the National Ocean Sciences Accelerator Mass Spectrometry facility. All radiocarbon dates, herein as calibrated ages in yr BP, were calibrated to calendar years using CALIB 6.0 (Stuiver et al., 2005). The fraction of organic sediments was calculated in 2-cm increments by weight of loss on ignition (LOI) at 550°C for 2 h (Heiri et al., 2001).

Diatom enumeration

Samples for diatom analysis were collected from every 0.5-cm increment, treated with 10% HCl to remove carbonate material and then 30% H₂O₂ to remove organic matter. Diatom slides were prepared according to standard procedures (Battarbee, 1986) and analyzed at 600× magnification, under oil immersion, using an Olympus BX51 microscope with differential interference contrast. A minimum of 300 diatom valves were identified and enumerated from random transects from each slide. Diatoms were identified to species using Krammer and Lange-Bertalot (1986–1991) and Camburn and Charles (2000). To infer overall diatom concentration, diatom slurries were spiked with 400 µL of 5-µm Duke Standards™ uniform polymer microspheres to achieve a concentration of 1.868 × 10⁶ spheres mL⁻¹. The total diatom concentration was based on the following equation: diatom concentration = (microspheres introduced × diatoms counted)/microspheres counted, according to the protocol described by Battarbee et al. (2001).

Pigment analysis

Pigments originating from photosynthetic organisms can be used to approximate past lake primary production and depositional conditions and, since pigments have some taxonomic specificity, can provide information regarding past algal community structure (McGowan, 2007). In order to compare algal production prior with and during the most recent GA, sedimentary pigments were quantified from ~2500 yr BP to present in Bunny and Raven Lake using standard high performance liquid chromatographic separations of chlorophylls (chls), carotenoids, and associated derivatives (Leavitt and Findlay, 1994). Pigments were extracted, filtered, and dried under N₂ gas. Extracts were separated in an Agilent 1200 series separation module with quaternary pump following methods adapted from Chen et al. (2001). Pigments were classified by spectral characteristics and chromatographic mobility and compared with unialgal cultures (Chen et al., 2001). Analysis covered pigments from all algae and higher plants (chlorophyll a and derivatives, β-carotene), chlorophytes (chlorophyll b and derivatives, lutein), total cyanobacteria (zeaxanthin), colonial cyanobacteria (canthaxanthin), diatoms (diatoxanthin, fucoxanthin), cryptophytes (alloxanthin), chrysophytes (fucoxanthin, diatoxanthin), Chlorobiaceae (green sulfur bacteria; Chl e), and dinoflagellates (diatoxanthin). Ultraviolet (UV)-absorbing pigments, produced primarily by cyanobacteria as photoprotectants (McGowan, 2007), were also measured. The sum of two chemically stable indicators, β-carotene and pheophytin a, was used to estimate total algal biomass (Leavitt and Carpenter, 1990). Pigment concentrations were normalized to sediment organic matter, and quantified at ~2.0-cm intervals (slightly shorter than centennial resolution).

Three-dimensional depth-habitat modeling

Depth-habitat modeling permits the calculations of lake volume, planar surface area, and basin surface areas at different lake elevations. These measurements determine how planktic and benthic habitat areas change by manipulating lake level (Stone and Fritz, 2004). This model can be used to interpret the planktic:benthic (P:B_core) from a core and infer lake level in the past. In order to assess whether glacial meltwater altered lake habitat conditions, the average P:B_core of diatom assemblages was calculated for each depth in the cores based on ecology of the species (P:B_core; Krammer and Lange-Bertalot, 1986–1991). Tychoplanktic and meroplanktic species were included in the planktic component of assemblages. Comparisons were made excluding tychoplanktic and meroplanktic species from the analysis to determine whether this yielded similar patterns and it was found that this did not alter the result.

Because lake level alters the availability of open water habitat for planktic diatoms, the photic zone depth, and available substrates for benthic diatoms (Hobbs et al., 2011; Stone and Fritz, 2004; Wigdahl et al., 2014), a model of the inferred available planktic and benthic habitat (P:B_model) can be created by manipulating lake level. A bathymetric map (Medford, 2013) was imported into Surfer 7.0 (Golden Software, 1999) to create a three-dimensional graphical image. This image was digitized and used to calculate volume, lake surface area, planar surface areas, and available benthic and planktic habitat areas as described in Stone and Fritz (2004; Figure 7). Decreasing water level was modeled by lowering the lake level in 1-m increments and calculating the available planktic habitat. This process was continued until the lake level was 2 m in depth. For available benthic habitat area calculations, maximum photic penetration depths were selected as 2.0 and 4.0 m. The available benthic habitat was modeled at these different photic zones for each 1-m drop in lake level. Given the lack of limnological measurements of light for this lake and given the turbidity present, photic penetration zones were predicted to be somewhat shallow. The model assumed a 2-m depth was necessary for planktic habitat. Changing this value by ±1 m does slightly alter the values for P:B_core but the general trend of the curves remain consistent. We modeled the available benthic diatom habitats for Bunny Lake, using two photic penetration scenarios; a low-light penetration scenario applied a 2.0-m average photic penetration and under a second scenario we doubled this range. Because we lacked annual photic penetration data, we used these scenarios as conservative estimates of the range in light environment based on the premise that Bunny Lake likely experiences frequent pulses of glacial flour.

Statistical analyses

Principal components analysis (PCA) was conducted on pigment concentrations from each lake to express dominant trends or patterns of variation over time using R (version 2.12.2; Legendre and Birks, 2012). Detrended correspondence analysis (DCA) in R (version 2.12.2) was used to summarize the prevailing trend in diatom community structure (Hill and Gauch, 1980) and included the percentage relative abundance of all diatom species, square root transformed, with down-weighting of rare taxa. Plotted through time, the DCA Axis 1 scores are an indication of the degree of community turnover between samples in time units of standard deviation (Hobbs et al., 2011). To identify the influence of variable meltwater input through time on species richness, average diatom species richness was calculated for each assemblage using a consistent sample size of 300 individuals with rarefaction analysis (Analytic Rarefaction 1.3, Steven M Holland). Relative species richness, hereafter referred to as species richness unless otherwise indicated, was normalized to the average deviation of the mean.

Results

Diatom stratigraphy and community change

Diatoms were not preserved in the sediment of Raven Lake. In Bunny Lake, diatom concentrations were highest from 8080 to 7900 yr BP and then decreased (Figure 3a). Diatom diversity was generally low, with a total of 65 diatom taxa found in the sediment core (Figure 3b). The diatom profile of Bunny Lake exhibited abrupt and substantial shifts in the dominant species throughout the core, some of which correspond with inferred periods of glacial advance and retreat (Figure 3b; Medford, 2013; Medford et al., in preparation). From 7200 yr BP, the diatom community was primarily benthic and was largely dominated by Pinnularia braunii. The relative abundance of Achnanthales was also relatively high (>15% of the assemblage) during this period but declined at ~7500 yr BP. While consistently present throughout the core, Fragilariales peaked in relative abundance at ~7415–7375 yr BP. Approximately 7250 yr BP, P. braunii declined slightly and the relative abundance of the meroplanktic species Aulacoseira pfaffiana increased to compose 20–60% of the assemblage until approximately 1015 yr BP. Fragilaria tenera morphotype 2 (with bent frustules compared with the standard form of this species; Cremer and Wagner, 2004) appeared in the record between 7050 and 6750 yr BP. Cavinula pseudoscutiformis increased in relative abundance from ~3900 to 2100 yr BP.

Over the last 1000 years, Bunny Lake experienced substantial shifts in diatom communities with three different novel species assemblages occurring during this time. The relative abundance of Achnanthales increased (>15% of the assemblage) and remained high from ~1015 yr BP to present. During the start of the GA (~1015 yr BP) there was a major shift to a novel diatom community as P. braunnii and A. pfaffiana declined and Tabellaria flocculosa, F. tenera morphotype 1 (with standard appearance of valves; Cremer and Wagner, 2004), Surirella minuta, and Fragilaria crotonensis increased. T. flocculosa flourished during the GA period, while the other three taxa that initially increased in this period declined after these spikes (Figure 3b). Hannaea arcus also increased midway through this time. Shortly after the start of the LIA-GA (~410 yr BP), A. pfaffiana and Fragilaria capucina increased and dominated, while T. flocculosa and H. arcus declined. These changes reversed to some extent when the sediment changed from sand to clay and persist in the modern assemblage. F. tenera morphotype 1 also increased again in the modern assemblage (50 yr BP to present).

Diatom species richness was relatively stable from 8000 to 975 yr BP with slight variations around a mean of 17 ± 5. Species richness then shifted to more variable but higher values starting with the GA to present with a mean of 27 ± 7 (Figure 3a). This shift was synchronous with a drop in percentage organic material. There was a sharp decline in richness at approximately 375–340 yr BP, after which richness values increased again. DCA Axis 1 scores were fairly good estimators of the variation in diatom communities, explaining 29% of the variance and demonstrating a shift toward greater turnover occurring around the GA (1015 yr BP; Figure 3a).

Planktic to benthic ratios indicate that the community was dominated by benthic taxa (average P:B_core through entire core = 0.96), largely driven by P. braunii, from the beginning of the record until approximately 6000 yr BP, as well as from 3580 to 2850 and 1375 to 870 yr BP. Diatom communities switched to planktic domination from ~2500 to 1450 yr BP, with high abundances of the meroplanktic species A. pfaffiana. Around 760 yr BP, planktic taxa primarily dominated the remainder of the GA and the LIA-GA, with a substantial increase in the relative abundance of T. flocculosa. This period contained several planktic peaks in the P:B_core record.

Pigments

Pigments in Bunny Lake showed a clear shift in concentrations through the course of the record, with a substantial decline occurring at approximately 1000 yr BP that has persisted to the present (Figure 4). Sedimentary pigment profiles of alloxanthin, diatoxanthin, canthaxanthin, and lutein–zeanthin showed similar trends throughout the early part of the record. Most pigment concentrations were relatively high from ~2200 to 1700 yr BP, except total chlorophylls (Σ chlorophyll a, pheophytin a and b, and pheophorbide a) which peaked at approximately 1050 and 2200 yr BP (Figure 4). Pigments were essentially absent from ~1015 yr BP to present compared with 2500–1000 yr BP. The key patterns observed in the pigment analyses for Bunny Lake (Figure 4) are summarized by PCA Axis 1 scores which were strong estimators of the major patterns in algal change with Axis 1 explaining 59% of the variance in Bunny Lake.

Figure 4.

Algal pigment analysis of Bunny Lake includes alloxanthin, diatoxanthin, canthaxanthin, lutein–zeaxanthin, total algal biomass (sum of β-carotene and pheophytin a), and total chlorophyll pigments (TChl; a sum of chlorophyll a, pheophytin a and b and pheophorbide a). Principal Components Analysis (PCA) includes analysis of entire pigment analysis.

In Raven Lake, pigment concentrations were variable through the ~2500-year record, showed no similarity in trends among pigments, and did not abruptly shift as in Bunny Lake (Figure 5). Concentrations of UV-absorbing compound, total algal abundance (Σ pheophytin a and β-carotene), and diatoxanthin peaked at ~1920 yr BP. During the LIA-GA, TChl and canthaxanthin peaked at ~1000 yr BP and alloxanthin, lutein–zeaxanthin, and bacterial Chl e all peaked at 1100 yr BP. The key patterns observed in the pigment analyses for Raven Lake (Figure 5) are summarized by PCA Axis 1 scores which explained 29% of the variance. PCA Axis 1 scores for Raven Lake showed no directional change throughout the 40-cm increments but revealed variation at about 1360, 1070, and 875 yr BP, as well as during the GA (~500 yr BP). Pigments in Raven Lake were not present/preserved from 831 yr BP to present.

Figure 5.

Algal pigment analysis of Raven Lake includes alloxanthin, diatoxanthin, canthaxanthin, lutein–zeaxanthin, total algal biomass (sum of β-carotene and pheophytin a), and total chlorophyll pigments (TChl; a sum of chlorophyll a, pheophytin a and b and pheophorbide a). Principal Components Analysis (PCA) includes analysis of entire pigment analysis.

Three-dimensional depth-habitat modeling

Our three-dimensional model examines available habitat for planktic and benthic diatoms based upon the changes in basin surface area exposed to light and planar surface area deep enough to sustain abundant plankton (~1 m). The model explores how these available habitats would change under different lake surface elevation settings, from maximum potential depth to minimum depth, thus providing a selection of available planktic and benthic ratios to compare with our diatom assemblage data (Stone and Fritz, 2004; Wigdahl et al., 2014). Based upon changes in basin morphometry, our model predicted that available benthic diatom habitat would surpass available planktic diatom habitat when the lake elevation was <1 m (4.0-m photic zone) to 1 m (2.0-m photic zone) below the current lake level. Additionally, the model predicts that the maximum available planktic:benthic habitat occurs under modern lake elevations (both photic penetration scenarios; Figures 6 and 7). The model also explains that any depths lower than 6 m shallower than the current lake level would substantially diminish available planktic habitat and essentially split the lake basin into two benthic-habitat-dominate basins (Figure 7).

Figure 6.

Comparison of P:B_core and P:B_model for Bunny Lake. P:B_model indicated the available planktic to benthic habitat for lake elevations scenarios varying from 827 to 824 m and under two different photic penetration regimes (2.0 and 4.0 m). The dashed line on the P:B_core indicates the point at which a lake switches from benthic dominated to planktic dominated.

Figure 7.

Modeled available planktic area under three different lake elevation scenarios (modern – 837, 835, and 831 m), benthic area not shown and isopleths represent lake depths. Available benthic area, shaded area, under two different photic regimes (2.0 and 4.0 m) and under three different lake elevation scenarios (modern – 837, 835, and 831 m). Black dot represents the location of the core.

When we compare the model results with the P:B_core, the high proportion of benthic diatoms in the early part of the record is most reasonably explained by lake setting where the water was at least 1 m lower than the modern lake (depending upon light penetration scenarios), since this is the depth where available benthic habitat area is greater than available planktic habitat area. Above ~40-cm depth in the core, the diatom assemblages shift toward dominance by planktic diatoms, suggesting more modern lake elevations, with a few pronounced reversions to lower lake conditions. An alternative, but equally possible, explanation is that lake levels remained relatively stable, but the light environment changed substantially through time, with a substantial decrease in the average light environment in the upper part of the core. This would reduce the available benthic habitat area without substantially reducing the available planktic habitat, but would likely result in a significantly shallower thermocline.

Discussion

The algal community in GSF Bunny Lake was relatively stable for most of the Holocene (prior to 1015 yr BP), with inputs of glacial meltwater having only minor and temporary effects on these communities. However, at the start of the GA (1015 yr BP), there was a shift in diatom species assemblages, an increase in diatom species richness, and a decline in algal pigments in this lake. Our results reveal that the onset of the GA and subsequent changes in glacier meltwater influx altered lake habitat conditions and triggered the development of a novel diatom assemblage. The start of the LIA-GA reversed many of these changes in diatom communities, while at the end of this period there was a partial return to GA assemblages that have persisted to the present day. In contrast, the algal pigment record in Raven Lake, a proximal snow-fed lake, did not show the pigment changes observed in Bunny Lake. In Bunny Lake, we attribute these changes in algal assemblages to three key effects of glacier meltwater that merged during the GA and subsequent LIA-GA: a greater influx of inorganic sediment, an increase in lake level, and a rise in nutrient subsidies in glacial meltwater. In particular, the presence of H. arcus indicates the presence of inflowing glacial meltwater as this species is a lotic indicator (Ludlam et al., 1996). The combination of these factors resulted in conditions that were novel in this ecosystem compared with the rest of the Holocene, and hence, elicited major shifts in algal assemblages.

Shifts in diatom communities occurred following an influx of sand in Bunny Lake approximately 480 yr BP. The relative abundance of T. flocculosa and P:B_core declined during this period although the assemblage remained pelagic. During this period of sand influx, the relative abundance of A. pfaffiana increased and may be the result of declining light availability, as A. pfaffiana is able to adjust its position in the water column when light availability is diminished (Camburn and Kingston, 2000). However, A. pfaffiana decreases when clay concentrations increase indicating that the population change in this species may have less to do with light availability and more to do with other undetermined mechanisms. Additionally, the increases in Achnantheales, particularly those diatoms that grow on sand grains, indicate the possible colonization of these species from an inflowing stream. The subsequent decline in sand input led to a rebound in the P:B_core and a rise in diatom abundance as inferred from the concentration of diatom valves. Algal pigments, however, remained low. Raven Lake, situated at a higher elevation and not connected to glacial meltwater channels, showed no evidence of sand or diminished phytoplankton pigments during this time.

With the LIA considered to be the maximum extent of the glaciers during the Holocene for the Scoresby Sund region (Levy et al., 2014; Lowell et al., 2013; Medford, 2013) and with the Bunny Lake cores showing evidence of glacial meltwater input, we infer that the quantity of meltwater was high during the LIA-GA resulting in an increase in lake level. Changes in lake level such as those elicited by increased meltwater input can cause an immediate habitat change for lake organisms, particularly benthic algae which have a restricted capacity to regulate to lake-level flux (Hoagland and Peterson, 1990; Stone and Fritz, 2004). Lake-level modeling indicates that the P:B in Bunny Lake is fairly sensitive to lake-level change and that only a modest change in lake level (drop of 0.5 m from present level) can elicit a shift to a greater availability of benthic habitat given the morphometry of the basin. Based on a comparison of P:B_core to P:B_model, the P:B_core most closely matches the P:B_model for a majority of the Holocene where the lake level was at least 1 m lower than present day levels (~100 yr BP–present; 837 m a.s.l.). However, during 350–270 and 850–560 yr BP, the P:B_core was greater than 1. With a relatively narrow range of lake level flux (0–1 m) yielding a switch from planktic to benthic diatoms, the P:B_core and P:B_model indicate that lake level was at least 836–837 m throughout periods of the Holocene. The model also predicts that lake levels lower than 6 m substantially reduce available planktic habitat and split the lake into two basins dominated by benthic habitat.

The P:B_core supports the assertion that there was an increase in lake level at ~1000 yr BP; however, as discussed above, this shift in P:B_core is not unique to the GA and the LIA-GA, with other periods showing elevated P:B. An influx of sediment at ~975 yr BP may have also altered water clarity, decreasing the available benthic habitat area without substantially altering planktic habitat and thereby leading to a shift in the P:B. The change in hydrology and subsequent change in lake level is likely one of the factors that promoted an abrupt change in lake biota, as indicated by the shift from the large, heavily silicified species, P. braunii (benthic) and A. pfaffiana, (meroplanktic) to the planktic diatom taxa, T. flocculosa and H. arcus; the latter two species have depth requirements and respond to an increase in water depth respectively (Schmidt et al., 2004) and are common in fast moving mountain waters (Kociolek, 2010). A. pfaffiana, capable of adjusting position in the water column (Camburn and Kingston, 2000), may thrive in variable water level conditions. Species richness increase may also be linked to the rise in lake level, since during this period benthic diversity was still prevalent (albeit diminished in numbers) and planktic species diversity increased.

The diatom community structure in Bunny Lake suggests that mild nitrogen enrichment from glacial meltwater began at 1000 yr BP, and persisted since that time. The increases in relative abundances of F. crotonensis and F. tenera, two species with moderate nitrogen requirements (Das et al., 2005; Saros et al., 2005), suggest that meltwater became a source of moderate nitrogen enrichment during the GA. This phenomenon is also evident in other remote systems, where nitrogen-indicating species also increased in relative abundance in a GSF lake in the central Rocky Mountains at ~1000 yr BP (Slemmons et al., 2015). Experiments with populations of F. crotonensis and F. tenera from remote alpine lakes in the US revealed that these species have moderate N requirements (half-saturation constant (K_s) for nitrate of 0.028 µM (Michel et al., 2006)) and high N requirements (K_s for nitrate of 0.397 µM (Williams et al., 2016), respectively. As K_s values indicate a lower requirement and better competitive ability for a given nutrient compared with other phytoplankton species, the increases in the relative abundances of these two species suggest increasing availability of nitrate in the water column. Such increases in these two taxa in sediment records have occurred with mild nitrate enrichment via atmospheric deposition in remote alpine lakes of the US Rocky Mountains (Saros et al., 2005; Wolfe et al., 2001) and in Olympic National Park (Sheibley et al., 2014) over the 20th century. Furthermore, increases in T. flocculosa in Bunny Lake during the start of the GA may also be due, in part, to increasing N availability. In the US Rocky Mountains, this species has a similar distribution pattern to F. crotonensis (Arnett et al., 2012). While we recognize the issues with optima inferred from weighted averaging, we note that in the US Rocky Mountains, T. flocculosa has a moderate nitrate optimum (34 µg L⁻¹) that is similar to that of F. crotonensis (29 µg L⁻¹; calibration set from Arnett et al., 2012 but nitrate optima not provided; Saros, unpublished).

Evidence to support nitrogen enrichment is further gleaned from nitrate concentrations in Greenland ice cores. Hansson (1994) demonstrated that nitrate concentrations in the Renland ice core doubled from the Younger Dryas into the Holocene, and were sustained at that higher concentration over the Holocene. Linkages between glacier nitrate concentration and the amount and timing of nitrate release are not clear; however, evidence suggests numerous processes are at work (as reviewed in Slemmons et al., 2013) including concentration of atmospheric nitrogen deposition within glacier ice (Lafreniére and Sharp, 2005), enhanced soil microbial processes (Williams et al., 2007), and changes in landscape features (i.e. degree of soil and vegetative cover; Sickman et al., 2001), making it difficult to constrain the timing of elevated nitrates in glacier-fed lakes. Our evidence suggests that glacier extent increased during and following the GA, with enhanced inputs of glacial meltwater to Bunny Lake. Based on knowledge of the ecology of these two diatoms, the increase in the relative abundances of F. crotonensis and F. tenera morphotype 1 in Bunny Lake at this time (and not during other periods of GA) suggests release of nitrate from glacier advance and melting. Other factors may also have influenced the relative abundance of these species including enhanced phosphorus concentration, which has been observed in some Greenland GF rivers (Hawkings et al., 2016). F. crotonensis has both moderate nitrate and total phosphorus optima (23.3 µL⁻¹; Miettinen, 2003); measurements of current lakewater nutrient conditions would assist with deciphering these effects but was not possible in this study. Alternatively, Rühland et al. (2015) have suggested that these taxa are indicators of warming and enhanced lake stratification. However, evidence from lake core sedimentology indicates that this was a time of GA, indicating that these taxa increased during a cooler period.

There was also a spike in the relative abundance of F. tenera (morphotype 2) at 6730 yr BP. However, this species was a different morphotype compared with the taxon present at the start of the GA and is characterized by centrally bent valves. This morphotype has been suggested by Gold et al. (2003) and McFarland et al. (1997) to occur as the result of an increase in concentrations of dissolved metals (i.e. cadmium, copper, iron, and zinc). However, Cremer and Wagner (2004) found this morphotype exists in non-polluted lakes of Store Koldewey, Northeast Greenland. The presence of this variant of F. tenera (morphotype 2) and the lack of other nitrogen-indicating taxa prior to the GA suggests that glacier meltwater was not present and/or sufficiently enriched in nitrogen to elicit an ecological change in this lake during that time. With atmospheric deposition as the primary source of nitrogen to glaciers and glacial meltwater (Wynn et al., 2007), the lack of nitrogen effects on Bunny Lake during periods of glacier advance prior to 1000 yr BP indicates that deposition was not high enough to result in nitrogen-enriched meltwater.

The pigment profile of Raven Lake was quite different compared with neighboring Bunny Lake. Raven Lake showed little directional change over the 2500-year period in terms of core stratigraphy and organic matter (Medford, 2013; Medford et al., in preparation) providing evidence that this lake was not influenced by glacial meltwater during the record. Pigments were present in variable quantities in Raven Lake from 831 yr BP to present, during which time they were absent from the Bunny Lake record. Raven Lake was also dominated by non-siliceous algae. Cremer and Wagner (2004) found the absence of diatoms in Fox Lake in the neighboring region of Store Koldewey in Northeast Greenland and attributed this to possibly low ion concentrations, particularly silica. This may be a possible explanation for why diatoms were not present in Raven Lake; however, without additional lake chemistry data, this conclusion is limited.

The possibility of pigment degradation does exist in this study as pigments can degrade rapidly and selectively (Furlong and Carpenter, 1988; Hurley and Armstrong, 1990) and may be the reason for minimal algal pigments post 1000 yr BP in Bunny Lake. However, it is also possible that light attenuation conditions may have changed in Bunny Lake during times of glacier advance with the influx of glacial flour and at times, sand (~450–200 yr BP), thereby reducing algal production. As pigments were preserved during 2500–1000 yr BP in the Bunny Lake record, and there was a potential increase in lake production from the start of GA to present as is evident by an increase in nitrogen-indicating taxa, the extremely low pigment concentrations from 1000 to present were likely due to ineffective pigment deposition and/or preservation resulting from increased turbidity and lower light availability as glacial flour can negatively affect the preservation of sedimentary algal records (Karabanov et al., 2004).

Our results indicate that a shift in the quality, quantity, and duration of glacial meltwater from the Renland Ice Cap starting about 1000 yr BP altered the algal communities in Bunny Lake. We attribute these changes to the synergistic effects of multiple alterations caused by increased influx of glacial meltwater, including elevated lake level, increased nitrogen, and increased turbidity (caused in part by higher sand input). The convergence of these mechanisms created an environment unique to other periods in the Holocene. While glaciers set in motion a series of biogeochemical lake responses following deglaciation (Engstrom et al., 2000; Fritz and Anderson, 2013) that occur gradually over long time scales, they can also abruptly and dramatically alter lake communities, eliciting a non-linear, threshold response. As a result, the nature of the meltwater from the start of the GA and LIA-GA to present was fundamentally unique in duration and quantity in comparison to that of other periods of glacier advance. This study supports the assertion that climate-driven change in glaciers can amplify the effects of climate on lake ecosystems, and the synergy of these three factors during the LIA-GA elicited a tipping point in ecosystem dynamics resulting in unprecedented change relative to other periods of glacial advance and retreat.

Footnotes

Acknowledgements

The authors would like to express their gratitude to Laura Levy, Paul Wilcox, and Yarrow Axford for field work in Greenland.

Funding

Funding for this work was provided by the National Science Foundation Arctic Natural Sciences, National Science Foundation ADVANCE program, and a University of Maine Correll Fellowship.

References

Arnett

Saros

Mast

(2012) A caveat regarding diatom-inferred nitrogen concentrations in oligotrophic lakes. Journal of Paleolimnology 47: 277–291.

Battarbee

(1986) Diatom analysis. In: Berglund

(ed.) Handbook of Holocene Palaeoecology and Palaeohydrology. Chichester: Wiley, pp. 527–570.

Battarbee

Jones

Flower

. (2001) Diatoms. In: Smol

Birks

HJB

Last

(eds) Tracking Environmental Change Using Lake Sediments. Volume 3: Terrestrial, Algal and Siliceous Indicators. Dordrecht: Kluwer Academic Publishers, pp. 175–180.

Bizzotto

Villa

Vaj

. (2009) Comparison of glacial and non-glacial-fed streams to evaluate the loading of persistent organic pollutants through seasonal snow/ice melt. Chemosphere 74: 924–930.

Blais

Schindler

Sharp

. (2001) Fluxes of semivolatile organochlorine compounds in Bow Lake, a high-altitude, glacier-fed, subalpine lake in the Canadian Rocky Mountains. Limnology and Oceanography 46: 2019–2031.

Bogdal

Schmid

Zennegg

. (2009) Blast from the past: Melting glaciers as a relevant source for persistent organic pollutants. Environmental Science & Technology 43: 8173–8177.

Camburn

Charles

(2000) Diatoms of Low-Alkalinity Lakes in the Northeastern United States. Special Publications 18. Philadelphia, PA: Academy of Natural Sciences of Philadelphia.

Chapin

Körner

(eds) (1996) Arctic and Alpine Biodiversity: Patterns, Causes and Ecosystem Consequences. Berlin: Springer-Verlag.

Chen

Bianchi

McKee

. (2001) Historical trends of hypoxia on the Louisiana shelf: Application of pigments as biomarkers. Organic Geochemistry 32: 543–561.

10.

Cremer

Wagner

(2004) Planktonic diatom communities in High Arctic lakes (Store Koldewey, Northeast Greenland). Canadian Journal of Botany 82: 1744–1757.

11.

Das

Vinebrooke

Sanchez-Azofeifa

. (2005) Inferring sedimentary chlorophyll concentrations with reflectance spectroscopy: A novel approach to reconstructing historical changes in the trophic status of mountain lakes. Canadian Journal of Fisheries and Aquatic Sciences 62: 1067–1078.

12.

Engstrom

Fritz

Almendinger

. (2000) Chemical and biological trends during lake evolution in recently deglaciated terrain. Nature 408: 161–166.

13.

Fritz

Anderson

(2013) The relative influences of climate and catchment processes on Holocene lake development in glaciated regions. Journal of Paleolimnology 49: 349–362.

14.

Funder

(1978) Holocene stratigraphy and vegetation history in the Scoresby Sund area, East Greenland. Grønlands Geologiske Undersøgelse Bulletin 129: 1–66.

15.

Furlong

Carpenter

(1988) Pigment preservation and remineralization in oxic coastal marine sediments. Geochimica et Cosmochimica Acta 52: 87–99.

16.

Gold

Feurtet-Mazel

Coste

. (2003) Impacts of Cd and Zn on the development of periphytic diatom communities in artificial streams located along a river pollution gradient. Archives of Environmental Contamination and Toxicology 44: 189–197.

17.

Hall

Baroni

Denton

(2008) The most extensive Holocene advance in East Greenland occurred in the Little Ice Age. Polar Research 27: 128–134.

18.

Hansson

(1994) The Renland ice core. A Northern Hemisphere record of aerosol composition over 120,000 years. Tellus 46: 390–418.

19.

Hawkings

Wadham

Tranter

. (2016) The Greenland Ice Sheet as a hot spot of phosphorus weathering and export in the Arctic. Global Biogeochemical Cycles 30: 191–210.

20.

Heiri

Lotter

Lemcke

(2001) Loss on ignition as a method for estimating organic and carbonate content in sediments: Reproducibility and comparability of results. Journal of Paleolimnology 25: 101–110.

21.

Hill

Gauch

Jr (1980) Detrended correspondence analysis: an improved ordination technique. Vegetatio 42: 47–58.

22.

Hjort

Funder

(1974) The subfossil occurrence of Mytilus edulis L. in central East Greenland. Boreas 3: 23–33.

23.

Hoagland

Peterson

(1990) Effects of light and wave disturbance on vertical zonation of attached microalgage in a large reservoir. Journal of Phycology 26: 450–457.

24.

Hobbs

Vinebrooke

Wolfe

(2011) Biogeochemical responses of two alpine lakes to climate change and atmospheric deposition, Jasper and Banff National parks, Canadian Rocky Mountains. Canadian Journal of Fisheries and Aquatic Sciences 68: 1480–1494.

25.

Hood

Fellman

Spencer

RGM

. (2009) Glaciers as a source of ancient and labile organic matter to the marine environment. Nature 462: 1044–1048.

26.

Hurley

Armstrong

(1990) Fluxes and transformations of aquatic pigments in Lake Mendota, Wisconsin. Limnology and Oceanography 35: 384–398.

27.

Jacobsen

Dangles

(2012) Environmental harshness and global richness patterns in glacier-fed streams. Global Ecology and Biogeography 21: 647–656.

28.

Karabanov

Williams

Kuzmin

. (2004) Ecological collapse of Lake Baikal and Lake Hovsgol ecosystems during the Last Glacial and consequences for aquatic species diversity. Palaeogeography, Palaeoclimatology, Palaeoecology 209: 227–243.

29.

Kelly

Lowell

Hall

. (2008) A ¹⁰Be chronology of lateglacial and Holocene mountain glaciation in the Scoresby Sund region, east Greenland: Implications for seasonality during lateglacial time. Quaternary Science Reviews 27: 2273–2282.

30.

Koc

Jansen

Haflidason

(1993) Paleoceanographic reconstructions of surface ocean conditions in the Greenland, Iceland and Norwegian Seas through the last 14 ka based on diatoms. Quaternary Science Reviews 12: 115–140.

31.

Kociolek

(2010) Hannaea arcus. In: Diatoms of the United States. Available at: http://westerndiatoms.colorado.edu/taxa/species/hannaea_arcus.

32.

Krammer

Lange-Bertalot

(1986–1991) Bacillariophyceae. In: Ettl

Gartner

Gerloff

. (eds) Sußwasserflora von Mitteleuropa [Freshwater flora of central Europe.], volumes 2, no. 1–4. Stuttgart; Jena: Gustav Fischer Verlag.

33.

Lafreniére

Sharp

(2005) A comparison of solute fluxes and sources from glacial and non-glacial catchments over contrasting melt seasons. Hydrological Processes 19: 2991–3012.

34.

Leavitt

Carpenter

(1990) Regulation of pigment sedimentation by photo-oxidation and herbivore grazing. Canadian Journal of Fisheries and Aquatic Sciences 47: 1166–1176.

35.

Leavitt

Findlay

(1994) Comparison of fossil pigments with 20 years of phytoplankton data from eutrophic Lake 227, Experimental Lakes Area, Ontario. Canadian Journal of Fisheries and Aquatic Sciences 51: 2286–2299.

36.

Legendre

Birks

HJB

(2012) From classical to canonical ordination. In: Birks

JBH

Lotter

Juggins

. (eds) Tracking Environmental Change Using Lake Sediments: Data Handling and Numerical Techniques. Dordrecht: Springer, pp. 201–248.

37.

Levy

Kelly

Lowell

. (2014) Holocene fluctuations of Bregne Ice Cap, Scoresby Sund, East Greenland: A proxy for climate along the Greenland Ice Sheet margin. Quaternary Science Reviews 92: 357–368.

38.

Lowell

Hall

Kelly

. (2013) Late-Holocene expansion of Istorvet Ice Cap, Liverpool Land, east Greenland. Quaternary Science Reviews 63: 128–140.

39.

Ludlam

Feeney

Douglas

MSV

(1996) Changes in the importance of lotic and littoral diatoms in a high arctic lake over the last 191 years. Journal of Paleolimnology 16: 187–204.

40.

McFarland

Hill

Willingham

(1997) Abnormal Fragilaria spp. (Bacillariophyceae) in streams impacted by mine drainage. Journal of Freshwater Ecology 12: 141–149.

41.

McGowan

(2007) Pigments in sediments of aquatic environments. In: Elias

(ed.) Encyclopedia of Quaternary Sciences. Amsterdam: Elsevier Science, pp. 2062–2074.

42.

Magnuson

Robertson

Benson

. (2000) Historical trends in lake and river ice cover in the northern hemisphere. Science 289: 1743–1746.

43.

Marienfeld

(1991) Holozäne Sedimentationsentwicklung im Scoresby Sund, Ost-Grönland. Reports on Polar Research 96. Bremerhaven: Alfred Wegener Institute for Polar and Marine Research, 162 pp.

44.

Medford

(1934) Holocene glacial history of Renland, East Greenland reconstructed from lake sediments. Electronic Theses and Dissertations. Orono, ME: University of Maine.

45.

Medford

Hall

Lowell

. (in preparation) Holocene glacial history of Renland, East Greenland, reconstructed from lake sediments. Quaternary Science Reviews.

46.

Michel

Saros

Interlandi

. (2006) Resource requirements of four freshwater diatom taxa determined by in situ growth bioassays using natural populations from alpine lakes. Hydrobiologia 568: 235–243.

47.

Michelutti

Wolfe

Briner

. (2007) Climatically controlled chemical and biological development in Arctic lakes. Journal of Geophysical Research: Biogeosciences 112: G03002.

48.

Miettinen

(2003) A diatom-total phosphorus transfer function for freshwater lakes in southeastern Finland, including cross-validation with independent test lakes. Boreal Environment Research 8: 215–228.

49.

Nam

(1997) Late quaternary glacial history and paleoceanographic reconstructions along the East Greenland continental margin: Evidence from high-resolution records of stable isotopes and ice-rafted debris. Reports on Polar Research 241. Bremerhaven: Alfred Wegener Institute for Polar and Marine Research, 157 pp.

50.

Notholt

(1998) Die Auswirkungen der ‘North East Water’ Polynia auf die Sedimentation vor NO-Grönland und Untersuchungen zur Paläo-Ozeanographie seit dem Mittel weichsel. Reports on Polar Research 275. Bremerhaven: Alfred Wegener Institute for Polar and Marine Research, 183 pp.

51.

Perić

Jolidon

Uehlinger

. (2015) Long-term ecological patterns of alpine streams: An imprint of glacial legacies. Limnology and Oceanography 60(3) 992–1007.

52.

Perren

Douglas

MSV

Anderson

(2009) Diatoms reveal complex spatial and temporal patterns of recent limnological change in West Greenland. Journal of Paleolimnology 42: 233–247.

53.

Robinson

Kawecka

(2005) Benthic diatoms of an Alpine stream/lake network in Switzerland. Aquatic Sciences 67: 492–506.

54.

Rühland

Paterson

Smol

(2015) Lake diatom responses to warming: Reviewing the evidence. Journal of Paleolimnology 54: 1–35.

55.

Saros

Interlandi

Doyle

. (2005) Are the deep chlorophyll maxima in alpine lakes primarily induced by nutrient availability, not UV avoidance? Arctic, Antarctica and Alpine Research 37: 557–563.

56.

Saros

Rose

Clow

. (2010) Melting alpine glaciers enrich high-elevation lakes with reactive nitrogen. Environmental Science & Technology 44: 4891–4896.

57.

Schmidt

Kamenik

Lange-Bertalot

. (2004) Fragilaria and Staurosira (Bacillariophyceae) from sediment surfaces of 40 lakes in the Austrian Alps in relation to environmental variables, and their potential for palaeoclimatology. Journal of Limnology 63: 171–189.

58.

Sheibley

Enache

Swarzenski

. (2014) Nitrogen deposition effects on diatom communities in lakes from three national parks in Washington State. Water, Air & Soil Pollution 225: 1857.

59.

Sickman

Leydecker

Melack

(2001) Nitrogen mass balances and abiotic controls on N retention and yield in high-elevation catchments of the Sierra Nevada, California, United States. Water Resources Research 37: 1445–1461.

60.

Singer

Fasching

Wilhelm

. (2012) Biogeochemically diverse organic matter in Alpine glaciers and its downstream fate. Nature Geoscience 5: 710–714.

61.

Slemmons

KEH

Saros

(2012) Implications of nitrogen-rich glacial meltwater for phytoplankton diversity and productivity in alpine lakes. Limnology and Oceanography 57(6): 1651–1663.

62.

Slemmons

KEH

Saros

Simon

(2013) The influence of glacial meltwater on alpine aquatic ecosystems: A review. Environmental Sciences: Processes and Impacts 15: 1794–1806.

63.

Slemmons

KEH

Saros

Stone

. (2015) Effects of glacier meltwater on the algal sedimentary record of an alpine lake in the central US Rocky Mountains throughout the late- Holocene. Journal of Paleolimnology 53: 385–399.

64.

Smol

Wolfe

Birks

HJB

. (2005) Climate-driven regime shifts in the biological communities of arctic lakes. Proceedings of the National Academy of Sciences of the United States of America 102: 4397–4402.

65.

Stone

Fritz

(2004) Three-dimensional modeling of lacustrine diatom habitat areas: Improving paleolimnological interpretation of planktic: Benthic ratios. Limnology and Oceanography 49: 1540–1548.

66.

Straile

Jöhnk

Rossknecht

(2003) Complex effects of winter warming on the physicochemical characteristics of a deep lake. Limnology and Oceanography 48: 1432–1438.

67.

Stuiver

Reimer

(2005) CALIB 5.0 (WWW program and documentation). Available at: calib.org/calib/.

68.

Golden Software (1999) SURFER (version 7.0). Golden, CO: Golden Software.

69.

Uehlinger

Robinson

Hieber

. (2010) The physico-chemical habitat template for periphyton in alpine glacial streams under a changing climate. Hydrobiologia 657: 107–121.

70.

Vincent

Callaghan

Dahl-Jensen

. (2011) Ecological implications of changes in the Arctic cryosphere. AMBIO 40: 87–99.

71.

Vinther

Clausen

Fisher

. (2008) Synchronizing ice cores from the Renland and Agassiz Ice Caps to the Greenland Ice Core Chronology. Journal of Geophysical Research 113: D08115.

72.

Wigdahl

Saros

Fritz

. (2014) The influence of basin morphometry on the regional coherence of patterns of diatom-inferred salinity in lakes of the northern Great Plains (USA). The Holocene 24: 603–613.

73.

Willemse

Koster

Hoogakker

. (2003) A continuous record of Holocene eolian activity in West Greenland. Quaternary Research 59: 322–334.

74.

Williams

Bautel

Nurse

. (2016) Phytoplankton responses to nitrogen enrichment in Pacific Northwest, USA Mountain Lakes. Hydrobiologia 776: 261–276.

75.

Williams

Knauf

Cory

. (2007) Nitrate content and potential microbial signature of rock glacier outflow, Colorado Front Range. Earth Surface Processes Landforms 32: 1032–1447.

76.

Wolfe

Baron

Cornett

(2001) Anthropogenic nitrogen deposition induces rapid ecological changes in alpine lakes of the Colorado Front Range (USA). Journal of Paleolimnology 25: 1–7.

77.

Wynn

Hodson

Heaton

THE

. (2007) Nitrate production beneath a high arctic glacier, Svalbard. Chemical Geology 244: 88–102.