Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario,Canada

Abstract

Historical changes in Holocene climate in northeastern Ontario were quantified using analyses of sedimentary pollen, diatoms, and pigments in a small boreal lake. Modern analog reconstructions of average temperature from Holocene pollen assemblages of Charland Lake showed temperature was ~2°C warmer than present conditions ~7800–4500 cal. yr BP, a time period consistent with the Holocene thermal maximum (HTM). Pollen data suggest a two-phase HTM: warm and dry conditions based on the presence of primarily Pinus spp., followed by warm and wet conditions based on increases in cedar. Overall, algal production was low during the HTM, as reflected by low concentrations of pigments and diatoms. In the late HTM, increases in cedar pollen and planktonic diatoms suggest sustained increases in water levels for the remainder of the Holocene. During the Post-HTM Period (~4500–2000 cal. yr BP), a period that was warmer than today but cooler than the HTM, overall pigment production was significantly higher than all other periods. However, changes in diatom species composition suggest this period was not uniform, with variation occurring between diatoms indicative of higher and lower nutrient levels. The last ~2000 cal. yr BP was less productive than the Post-HTM Period but more productive than the HTM with higher production from diatoms and cyanobacteria. This study suggests that the relationship between climate and lake water production can be quite complex, and that changes in temperature, precipitation, light, lake levels, and mixing patterns are among factors that are related to changes in subfossil phototroph assemblages.

Keywords

diatoms landscape dynamics pollen sedimentary pigments thermal stratification

Introduction

The Holocene thermal maximum (HTM) was a period of enhanced warmth that occurred during the early and middle Holocene (Renssen et al., 2009, 2012; Viau et al., 2006). This warming has been well recorded and studied in the mid-to-high latitudes of the Northern Hemisphere showing that major climate alterations occurred following the last glacial maximum (LGM) and deglaciation (Carlson et al., 2008; Whitlock and Bartlein, 1997). The cause of the HTM has largely been attributed to an increase in solar insolation, because of orbital variation, which maximized at ~11 ka BP (1000 years before present; Bond et al., 2001; Ritchie et al., 1983). Despite the overall maximum solar insolation occurring at this time, the specific regions of North America may not have experienced maximum warmth at the same time or to the same degree (Renssen et al., 2009, 2012). In general, regions at higher latitudes experienced greater amounts of warming than those at lower latitudes (Briner et al., 2016; Renssen et al., 2009, 2012) and the western portion of the continent warmed earlier than the east, likely because of downwind cooling effects of the melting Laurentide ice sheet (LIS) in the northeast of the continent (Carlson et al., 2008; Viau and Gajewski, 2009). Alongside the variation in onset, duration, and magnitude of warmth experienced during the HTM in North America, reconstructions of Holocene precipitation levels suggest a climate dipole in which the west of the continent was dry, while the east was wet during the HTM (Shuman and Marsicek, 2016). To date, however, these studies have focused on the continental United States and the Arctic or subarctic regions (Briner et al., 2016; Shuman and Marsicek, 2016; Viau and Gajewski, 2009), and relatively little is known regarding the nature of the HTM in the boreal region of east-central Canada (e.g. northern Ontario).

The boreal forest of northern Ontario spans ~1500 km and is one of the largest continuous forests globally. One of its primary features is its abundant fresh water contained within its wetlands, rivers, and abundant lakes. Despite the vast size of the boreal region, it is susceptible to anthropogenically forced climate change (Price et al., 2013). Therefore, a characterization of landscape responses to past climates may be useful for predicting effects of atmospheric warming and changes in regional hydrology (Klemm et al., 2016; Navarro et al., 2018; Teller et al., 2018). Atmosphere–ocean–vegetation models have suggested that the HTM in northern Ontario was ~2–3°C warmer than the modern climate (Renssen et al., 2009, 2012). Recent work from northwest Ontario (~1200 km to the west of our study site) suggests that the HTM occurred from ~8500 to 4500 cal. yr BP (calendar years before present) and was ~2°C warmer than present conditions. This estimate was calculated using a modern analog technique (MAT) to reconstruct past temperatures, based on a regional set of modern pollen and climate data (Moos and Cumming, 2011). Arid conditions at this time resulted in lower lake levels, increased algal production, an eastward shift of the prairie–forest ecotone, and an increase in fire activity (Moos and Cumming, 2011, 2012; Moos et al., 2009; Karmakar et al., 2015a, 2015b). Despite this research, there has been little work examining the HTM of northeastern Ontario, Canada, which is a gap this paper seeks to address.

Only one Holocene-scale palynological study has documented Holocene climate variability in northeast Ontario (Liu, 1990). This study examined three sediment cores along a north–south transect of eastern Ontario. One of these sites, Lake Six is ~40 km from our study site, so comparisons with Liu (1990) refer to the pollen record from Lake Six. In Liu (1990), the author showed a delayed, but protracted HTM Period occurring from ~7 to 2.5 ka BP at Lake Six. This was defined by increases in Pinus strobus, a typical taxon of the warm Great Lakes–St Lawrence forest, and in Cupressaceae. Liu (1990) interpreted the presence of Cupressaceae pollen as belonging to Thuja based on its co-occurrence with Thuja macrofossils within the sediment core. These data suggest that the HTM of northeast Ontario was wet, agreeing with studies at similar longitudes to the south (Shuman and Marsicek, 2016; Viau et al., 2006). A similar interpretation has been made by Carcaillet et al. (2001) who used charcoal analysis to infer that low fire activity during this time in western Quebec was likely attributable to enhanced water availability. Unfortunately, Liu’s (1990) results from Lake Six are supported by only three radiocarbon dates. With these few data points, there is likely considerable error regarding the temporal boundaries of HTM warming. Improved geochronology, as well as replicate sites in the region, are important to better define the onset and duration of the HTM in northeast Ontario, as well as its effects on lake ecosystems.

This paper quantifies changes in climate systems in northeastern Ontario using robust analysis of fossil pollen to reconstruct past vegetation and climate, as well as fossil diatoms and pigments from aquatic phototrophs to quantify the effects of climate variability on lake production and community composition. Diatoms are ideal for these objectives, as certain species have well-documented optima for nutrients and other physical and chemical limnological characteristics (water depth, pH, salinity, stratification regime, etc.). By examining changes in diatom assemblages, changes in nutrient levels can be inferred (Cumming et al., 2015; Douglas and Smol, 1999; Lotter et al., 1999; Rühland and Smol, 2005). Similarly, fossil pigments (chlorophylls, carotenoids, derivatives) often preserve after the loss of morphological remains of non-siliceous phytoplankton and phytobenthos, and are used to estimate historical changes in the abundance of primary producers, as well as their gross community composition (Hall et al., 1999; Hodgson et al., 1998; Leavitt et al., 1994a, 1997). Taken together, analysis of these proxies can be used to answer three questions related to the HTM: (a) When and for how long did the HTM take place in northeast Ontario? (b) What was the degree of warming that occurred during the HTM of this region, and were the warmer conditions, associated with a wet or dry climate? and (c) What was the response in algal production to warmer conditions?

Methods

Study area

Charland Lake (48°34′50.8″N, 80°53′46.5″W; Figure 1) is located east of Timmins, Ontario, Canada, immediately west of Kettle Lake Provincial Park. Charland Lake, like other basins in the region, formed when Glacial Lake Ojibway regressed from the landscape ~8200 cal. yr BP (Carlson et al., 2008; Margold et al., 2018; Roy et al., 2015; Veillette, 1994). Charland Lake is small (14.2 ha), has a maximum depth of 17.7 m, and is situated at 274 m a.s.l. The climate of the region is humid continental (Koppen Dfb), which is characterized by wet summers and long, cold winters. Mean annual temperature is ~1.8°C (Environment Canada, 2019). Modern boreal forest in the area is composed mainly of white (Picea glauca) and black spruce (Picea mariana) in lowlands, and jack pine (Pinus banksiana) and white birch (Betula papyrifera ) on upland sites. Charland Lake occurs in the ‘Great Clay Belt’ of northern Ontario. The bedrock is composed of metamorphosed Precambrian granites and granodiorites which are overlain by Quaternary glacial deposits and extensive glaciolacustrine clays associated with Glacial Lake Ojibway (Roy et al., 2011, 2015; Veillette, 1994). Charland Lake is located at the edge of an esker complex at the interface between glaciolacustrine clays and sand and gravel deposits. The lake is imbedded in the groundwater system as evidenced by the presence of a lake outlet and many groundwater springs flowing on the eastern shore of the lake.

Figure 1.

(a) Bathymetric map of Charland Lake with 2 m contours. The black star represents the location of the piston core that was taken from a depth of ~16 m. (b) A reference map of Ontario with the Charland Lake coring site indicated by a black star.

Sample collection

Bathymetric data points were collected from Charland Lake using a Garmin GPS-Map Sounder 238 connected to a Toughbook computer with Bathymetric Automated Survey System (B.A.S.S.) software, v. 2.4 (Levec, 2001). These data points were converted to shape files and used to produce the bathymetric map in ArcMap 10.5 (Esri, 2016). A 473-cm long sediment core was collected from a depth of ~16 m on 19 June 2014 using a 1-m square rod Livingstone piston corer with an internal diameter of 5.1 cm (Glew et al., 2001; Wright, 1967; Wright et al., 1984). Approximately 10 cm of material from the top of the core were lost during horizontal extrusion in the field. To assure collection of an undisturbed sediment–water interface, and to be able to accurately date the top of the piston core, a gravity core (internal diameter 7.62 cm) was taken at the same location using a modified gravity corer (Glew, 1989), which was sectioned into 0.5-cm intervals in the field. The piston core exhibited distinct units, with organic gyttja from 0 to 313 cm, gray clays from 313 to 337 cm, organic-rich material from 337 to 417 cm, and basal clay deposition below that level. In this paper, we refer to these clay layers in order of deposition, making the clay deposit from 417 to 473 cm ‘the first clay layer’ and the deposit from 313 to 337 cm ‘the second clay layer’. The piston core was wrapped in 1-m sections on site and transported in a cooler to the Paleoecological Environmental Assessment and Research Laboratory (PEARL) at Queen’s University, where they were stored horizontally in a cold room at ~4°C. Half of the core was then sectioned into 1-cm intervals, while the other half was archived.

Analyses

Chronology

The top 13 cm of the gravity core was dated using the constant rate of supply (CRS) model based on the unsupported ²¹⁰Pb gamma activity of the sediments. This was done to assess the date of the top of the piston core, as the piston cores were purposely turned horizontally in the field, which resulted in the loss of the watery uppermost sediments of the first core section. The ¹⁴C chronology for the piston core was determined by dating a concentrated sample of pollen grains from eight sediment samples with accelerator mass spectrometry (AMS). Pollen grains were isolated from lake sediments by LacCore using a procedure similar to Brown et al. (1989) and measured for ¹⁴C at the Lawrence Livermore National Laboratory. An age–depth relationship based on these radioisotopic data was constructed using Bayesian age modeling with the rBACON (v. 2.2) modeling package in R (R Core Team, 2015) with the IntCal13 ¹⁴C calibration curve (Blaauw and Christen, 2011; Reimer et al., 2013). Default settings outlined by Goring et al. (2015) were used to determine the gamma distribution of the accumulation rate, which were in agreement with the posterior distribution of activities (Blaauw and Christen, 2011). The prior memory was set to a mean of 0.3 and shape of 25 to accommodate small shifts in accumulation rates. The sensitivity of prior memory parameters was tested, and changes in mean and shape did not result in large changes to the model.

Dry mass and organic matter

Dry mass and percent organic matter was determined through standard loss-on-ignition procedures (Heiri et al., 2001) on 60 intervals throughout the sediment core.

Pollen

Pollen samples were prepared at every 8 cm (60 samples total) using a modification of the method of Bennett and Willis (2001). Sediment samples were spiked with two exotic Lycopodium tablets (batch no. 1031) and digested with 10% hydrochloric acid and 10% potassium hydroxide before being sieved through a 10-μm mesh. The remaining sediment was further digested in 40% hydrofluoric acid and then acetolyzed, deflocculated with 10% sodium metaphosphate, stained with Safranin, and mounted on microscope slides in silicon oil. Pollen grains were counted with a Leica light microscope with a 40× differential-interference-contrast objective. A minimum of 400 pollen grains were counted per sample except those in which pollen concentration was low and multiple coverslips needed to be counted. In those cases, a minimum of 300 grains were counted. Pollen grains were identified to the lowest possible taxonomic resolution based on published references (Bassett et al., 1978; Kapp et al., 2000; McAndrews et al., 1973).

Pigments

Subsamples of wet sediment were used for determination of photosynthetic pigment concentrations from 59 intervals throughout the core at the University of Regina’s Institute of Environmental Change and Society (IECS). Sedimentary pigment analysis was undertaken following procedures outlined in Leavitt and Hodgson (2001). Pigment concentrations are reported as nanomoles of the pigment per gram of the organic matter, a unit which is linearly proportional to the standing stock of phytoplankton in decadal-scale monitoring programs (Leavitt et al., 1994a). An Agilent model 1100 high-performance liquid chromatography (HPLC) system with photodiode array detector was calibrated using commercial pigment standards from DHI (Denmark). Analysis included the main a- and b-phorbins (chlorophyll, pheophytin), as well as chemically stable, taxonomically diagnostic pigments representing total algal abundance (β-carotene), cryptophytes (alloxanthin), total cyanobacterial (echinenone), colonial cyanobacteria (myxoxanthophyll), Nostocales cyanobacteria (canthaxanthin), chlorophytes (Chl b), chlorophytes + cyanobacteria (lutein–zeaxanthin), siliceous algae (fucoxanthin), mainly diatoms (diatoxanthin), and anaerobic purple sulfur bacteria (okenone) following Leavitt and Hodgson (2001). Historical changes in the lacustrine preservation environment were recorded as changes in the ratio of labile precursor (Chl a) to stable degradation products (pheophytin a), both ubiquitous pigments used to estimate total phototrophic abundance (Leavitt and Hodgson, 2001).

Diatoms

Diatom samples were prepared using ~0.2–0.3 g of wet sediment, taken every 8.0 cm, for a total of 59 subsamples. Samples were digested in a 1:1 molar solution of concentrated nitric and sulfuric acids before being brought to a slightly acidic pH through repeated rinses with double-deionized water. Diatom slurries were reduced to a volume of ~5 ml by aspiration following sedimentation and spiked with a solution of microspheres of known concentration equivalent to 4 ml of a 2.0 × 10⁷ spheres/ml solution for organic samples and 0.2 ml of a 2.0 × 10⁷ for clay-rich samples. Samples were plated on coverslips in a series of four dilutions and then mounted to slides using Naphrax^®. Diatoms valves were identified and enumerated using a Leica DMRB microscope under a 100× Fluotar objective (NA of objective = 1.3) and differential interference contrast optics at 1000× magnification. For most diatom samples, a minimum of 400 valves were counted, or, if the concentration of valves was exceptionally low, until five transects were completed. Diatoms were identified to species or lower taxonomic units using the standard references (Cumming et al., 1995; Krammer and Lange-Bertalot, 1986, 1988, 1991a, 1991b). Chrysophyte scales were enumerated alongside diatoms but were not taxonomically identified.

Numerical analyses

Fossil pollen, pigment, and diatom data were plotted using the computer program Tilia v. 2.0.2 (Grimm, 2004). Pollen taxa present at greater than 2% abundance in at least two samples and diatom species with greater than 5% abundance in at least three samples were included in the plots. Pollen concentration was calculated with the formula: (exotic Lycopodium spores added × fossil pollen counted)/exotic Lycopodium spores counted (Bennett and Willis, 2001) and standardized to dry mass. A depth-constrained cluster analysis (CONISS; Grimm, 1987) was performed on the pollen assemblage to identify major pollen zones. The statistical significance of zones delineated by CONISS was validated with a broken stick model by rioja package in R (Juggins, 2015; R Core Team, 2015).

MAT was used to quantitatively reconstruct mean annual temperature over the Holocene record of Charland Lake using the C2 software program (Juggins, 2003). Modern pollen taxa and climate data were collected from the North American pollen database (Whitmore et al., 2005; Williams et al., 2006) to form a regional calibration set. A total of 305 samples between 45°N−60°N and 75°W–90°W were selected to form the calibration set which captured the transition between the boreal and mixed-wood forests to the south of Charland Lake. The pollen abundances present in the core samples from Charland Lake are well represented in the modern pollen dataset, so it is likely that this group of sites should provide strong analogs over the Holocene. Cupressaceae pollen was removed from the reconstruction model as this pollen type can represent both Juniperus and Thuja which have very different climate preferences (Yu, 1997). Annual average temperature was reconstructed from an unweighted average of the five closest modern analogs compared with core pollen samples, using a squared chord measure of dissimilarity (Overpeck et al., 1985; Viau et al., 2006; Viau and Gajewski, 2009) and bootstrap cross-validation (bootstrapped r² = 0.95, root-mean-square-error-of-prediction (RMSEP) = 0.88). A distribution of the dissimilarity of the five lowest dissimilarities to each sample in the 305-site calibration set was calculated. Each of the five lowest dissimilarities (i.e. the closest analogs for the core samples) were within the top 20th percentile of the distribution with the majority of core samples occurring within the top 10th percentile of this distribution.

Pigments were restricted to common sedimentary compounds of known chemical stability (see above). Analysis of similarities (ANOSIM) using a Bray–Curtis dissimilarity coefficient and 999 permutations were used to test the null hypothesis that there was no difference in diatom and pigment composition between the pollen-inferred climate zones (Clarke and Warwick, 1994). Analyses were performed using both non-transformed and square root transformed species data. ANOSIM tests were performed on diatom relative abundance and concentration and pigment concentration data. ANOSIM tests were calculated using the PAST 3 software package (Hammer et al., 2001; Hammer and Harper, 2006). The sample from a depth of 440 cm was determined to be an outlier and was removed prior to any statistical analyses. Following the ANOSIM, SIMPER tests (also performed in PAST) were used to calculate the contribution of each species to the average dissimilarity between the two groups. Post hoc t-tests assuming unequal variance were performed to identify the significant difference between the pollen-inferred climate zones for individual sedimentary pigments using Microsoft Excel. The index of chrysophyte scales to diatom frustules (scale-to-diatom index) was calculated using the formula: scales/(diatoms + scales) × 100 (Moos et al., 2005).

Results

Age model

The activity of ²¹⁰Pb in the Charland Lake gravity core showed an exponential decay in total ²¹⁰Pb activity with increasing core depth, reaching supported levels by a depth of 12 cm (Supplemental Figure 1, available online). Core notes indicate that ~10 cm were lost from the first section of the piston core, which according to the age–depth CRS model, corresponds to ~1950, or a calibrated ¹⁴C age of ~0 cal. yr BP.

Concentrated pollen samples from eight sediment intervals were dated using AMS (Table 1). Bayesian modeling of the depth−¹⁴C age relationship and the estimate surface age from ²¹⁰Pb activities revealed a relatively constant rate of deposition over the period of study (Figure 2). Overall, the errors of the inferred ages for intervals in the sediment core were relatively low (± 30–40 years). The calibrated 14C dates increased in age with increasing depth and follows an approximately linear trend with cumulative core depth.

Table 1.

Summary of the ¹⁴C-dating results on pollen isolated from selected intervals from the sediment cores from Charland Lake. All analyses were performed based on pollen isolated at the LacCore Facility at the University of Minnesota and dated at Lawrence Livermore National Laboratory. The mean age of the distribution is presented in ‘Cal. yr BP’. The top date was generated through ²¹⁰Pb dating and a CRS model.

Core	Cumulative depth in piston core (cm)	δ¹³C	¹⁴C age ± SD	Cal. yr BP	2-σ cal. yr BP range(min–max)	CAMS #
C2S1	0	N/A	N/A	0	−17–17	N/A
C2S1	45.5	−28	1170 ± 30	1081	976–1187	171192
C2S2	143.5	−28	2505 ± 40	2623	2451–2749	171205
C2S2	178.5	−28	3035 ± 30	3260	3133–3360	171201
C2S3	204–206	−28	3480 ± 30	3762	3649–3863	176019
C2S3	262.5	−28	4400 ± 30	4925	4846–5037	171194
C2S4	305–306	−28	4560 ± 35	5390	5238–5570	176020
C2S4	401.5	−28	6235 ± 30	7159	6999–7272	171206
C2S5	437.5	−28	6995 ± 40	7815	7670–7947	171202

Figure 2.

A Bayesian age–depth model run on the Charland Lake core using the computer program rBACON (v. 2.2). Top left panel: Markov Chain Monte Carlo model iterations. Top middle panel: prior (heavy line) and posterior (solid) distribution of accumulation rate. Top right panel: prior (heavy line) and posterior (solid) distribution of model memory. Bottom panel: calibrated ¹⁴C dates and the age–depth model. The outer dotted lines indicate 95% confidence intervals. The central dotted line is the model based on the weighted mean age. Ages at depths > the last ¹⁴C age were estimated from extrapolation.

Pollen

Five statistically significant pollen zones were identified in the Charland Lake core through CONISS and validated with the broken stick model (Figure 3; Supplemental Figure 2, available online). CONISS analysis demonstrated that the highest level breaks between zones occur at cumulative depths of 445, 342, 253, and 85 cm. Reconstructed average temperature showed major changes in line with four zones labeled as the Pre-HTM, HTM, Post-HTM, and Modern climate zones (Figure 3). Standard error for the temperature reconstructions were generally low and on average did not exceed 0.5°C above or below the estimate (Figure 4). Error is larger in the Pre-HTM zone, but even considering the upward limit of the error, there is still a large and noticeable increase in temperature when entering the HTM zone (Figure 4).

Figure 3.

Relative abundance of abundant (>2% found in two sections of the core) pollen taxa in Charland Lake. Total pollen concentration is also shown. Darker zones in Sediment Type represent organic sedimentation in the piston core. Layered zones represent clay deposits. Climate zones are derived through CONISS (see text). Reconstructed average temperature was estimated through MAT calibration (bootstrapped r² = 0.95, RMSEP = 0.88).

Figure 4.

Reconstructed average temperature with error bars and percent dissimilarity of the analogs used in the MAT temperature reconstruction of the Charland Lake record. Percent dissimilarity refers to the mean dissimilarity of the five closest analogs used to reconstruct annual temperature for each fossil interval. Zones are the CONISS-derived climate zones from Figure 3.

The Pre-HTM zone occurred from 473 to 445 cm (~8200–7800 cal. yr BP) and showed mixed dominance of Picea and Pinus; Pinus banksiana and Pinus resinosa could not be distinguished. This zone also included a relatively low abundance of arboreal pollen compared with rest of the core, as well as more abundant spores and pollen from Sphagnum, fern, and Cyperaceae. Reconstructed average temperature was approximately −0.5 to 0°C in this zone. The HTM zone occurred from 445 to 253 cm (~7800–4500 cal. yr BP) and showed a major increase in reconstructed temperature (Figures 3 and 4). Average temperatures of ~2–3°C were consistent across this zone, but there are two distinct floral assemblages within the warm period. The first was marked by an increase in the Pinus banksiana/resinosa and the second with an increase in Pinus strobus and Cupressaceae. Organic sedimentation also began in this zone, concomitant with pollen concentration rapidly increasing before falling substantially at the onset of the second clay layer. The Post-HTM Period (253–85 cm, ~4500–1700 cal. yr BP) showed an increase in Picea, Pinus banksiana/resinosa, and Betula pollen, as Pinus strobus and Cupressaceae decline (Figure 3). This is reflected in the MAT reconstruction as average temperature fell by ~1°C. Pollen concentration was variable in this zone but generally increased during the earlier portion before declining after ~3300 cal. yr BP. The Modern zone (85–0 cm, ~1700–0 cal. yr BP) showed a general decrease in pollen concentration and further decreases of Pinus strobus. Picea became predominant in this zone with Pinus banksiana/resinosa and Betula experiencing minor decreases compared with the Post-HTM zone. Reconstructed average temperature slowly declined by ~0.5°C throughout this zone. The concentration of pollen types was also plotted and can be found in the supplemental materials (Supplemental Figure 3, available online).

Pigments

The pollen climate zones were used as a framework to assess the relationship between climate change and variation in pigment assemblages (Figure 5). Overall, preservation of pigments was poor in the Pre-HTM zone, with fossil concentrations below detection limit in most samples. A similar absence of fossil pigments in the second clay layer suggests that the near absence of sedimentary organic matter favored complete decomposition of carotenoids and chlorophylls, irrespective of their inherent chemical lability, similar to patterns seen in glacially fed alpine lakes (Bunting et al., 2010). In contrast, concentrations of ubiquitous pigments (β-carotene, Chl a, pheophytin a), and those from cryptophytes (alloxanthin), diatoms (diatoxanthin), total cyanobacteria (echinenone), Nostocales cyanobacteria (canthaxanthin), chlorophytes (pheophytin b) and chlorophytes + cyanobacteria (lutein–zeaxanthin) all increased markedly at the base of the HTM zone, whereas more labile compounds (fucoxanthin, Chl b) did not. These patterns reflect the increase in okenone from obligate anaerobic purple sulfur bacteria, an indicator of anoxia in deep waters or sediment. Transition from the HTM to Post-HTM zones was marked by significant increases in the concentration of most fossil pigments (Table 2, Figure 6). The Post-HTM zone was characterized by elevated concentrations of most pigments, often to a historical maximum relative to other zones (e.g. β-carotene, lutein–zeaxanthin, okenone, and canthaxanthin). In the Modern zone (0–85 cm; ~1700–0 cal. yr BP), pigment assemblages were marked by the first appearance of myxoxanthophyll from colonial cyanobacteria, increases in concentrations of compounds from Nostocales cyanobacteria (canthaxanthin) and secondarily total cyanobacteria (echinenone), and historical maxima of less chemically stable pigments from diatoms (fucoxanthin, diatoxanthin) and chlorophytes (Chl b). In contrast, fossil levels of ubiquitous β-carotene and the mixed chlorophyte–cyanobacterial indicator, lutein–zeaxanthin, declined slightly in the most recent zone. Relative to individual pigments, the ratio of labile chlorophyll a to stable pheophytin a (indicating the preservation environment) remained relatively constant through the core, with some variance within and between the clay bands of the Pre-HTM and HTM zones (Figure 5).

Figure 5.

Concentration of photosynthetic pigments (nmol/g organic matter) in Charland Lake sediment core over time (cal. yr BP). The ratio of chlorophyll a to pheophytin a, an indicator of preservation, and organic matter (%) are also shown. Darker zones in Sediment Type represent organic sedimentation in the piston core. Layered zones represent clay deposits. The dotted lines indicate the pollen-derived climate zones from Figure 3.

Table 2.

Summary table of the results from a series of two sample post hoc t-tests assuming unequal variance conducted to identify significant differences in pigment concentrations between the HTM, Post-HTM, and Modern zones for sedimentary pigments.

Zones	Canthaxanthin	Echinenone	Okenone	Alloxanthin	β-carotene
HTM, Modern	t (22) −8.7 = p = < 0.01	t (12) −2.5 = p = 0.03	t (21) = −6.3 = p = < 0.01	t (26) −5.1 = p = < 0.01	t (21) = −1.9 p = 0.07
HTM, Post-HTM	t (44) −5.0 = p = < 0.01	t (43) = −4.5 p = < 0.01	t (42) = −4.2 = p = < 0.01	t (31) −4.2 = p = < 0.01	t(43) = −4.1 p = < 0.01
Post-HTM, HTM	t (23) −4.2 = p = < 0.01	t (11) = −0.8 p = 0.44	t (25) = −2.2 = p = 0.04	t (32) −1.4 = p = 1.69	t (24) = 1.7 p = 0.1
	Lutein–zeaxanthin	Myxoxanthophyll	Fucoxanthin	Chlorophyll b	Diatoxanthin
HTM, Modern	t (32) −2.5 = p = 0.02	t(10) −3.3 = p = < 0.01	t (16) −19.4 = p = < 0.01	t (10) −8.4 = p = < 0.01	t (25) = −10.7 = p = < 0.01
HTM, Post-HTM	t (36) −4.4 = p = < 0.01	N/A	t (25) −6.3 = p = < 0.01	t (22) −3.1 = p = < 0.01	t (41) = −6.4 = p = < 0.01
Post-HTM, Modern	t (25) 2.3 = p = 0.03	t (10) −3.3 = p = < 0.01	t (31) −4.6 = p = < 0.01	t (26) 0.3 = p = 0.77	t (19) = −6.0 = p = < 0.01

HTM: Holocene thermal maximum.

Figure 6.

Boxplots of pigment concentrations (nmol/g OM) within the a priori defined pollen-derived climate zones (HTM (n = 24), Post-HTM (n = 21), Modern (n = 11)). ANOVA tests were run to test for significant differences between zones for (a) β-carotene (F(2, 53) = 8.2, p < 0.01), (b) lutein–zeaxanthin (F(2, 53) = 7.6, p < 0.01), (c) alloxanthin (F(2, 53) = 11.4, p < 0.01), (d) echinenone (F(2, 53) = 7.3, p < 0.01), (e) okenone (F(2, 53) = 10.1, p < 0.01), (f) canthaxanthin (F(2, 53) = 26.8, p < 0.01), (g) chlorophyll b (two outliers removed in the Post-HTM zone; (F(2, 53) = 6.5, p < 0.01)), (h) fucoxanthin (F(2, 53) = 52.1, p < 0.01), (i) diatoxanthin (F(2, 53) = 47.1, p < 0.01), and (j) myxoxanthophyll (F(2, 53) = 22.9, p < 0.01).

One-way pair-wise ANOSIM tests confirmed that there were significant differences in pigment concentration among the three most recent zones. Specifically, the null hypothesis of no difference between pigment concentrations among zones was rejected in comparisons of the HTM and Post-HTM zones, and between the Post-HTM and Modern zones (Table 3). A SIMPER test identified lutein–zeaxanthin as the predominant pigment contributing to ~40% of the difference in composition between the HTM and Post-HTM zones (Table 4), and ~29% of the difference in composition seen between Post-HTM and Modern zones (Table 4).

Table 3.

Summary of the one-way ANOSIM pair-wise tests (Bray–Curtis dissimilarity) on Charland Lake diatom relative abundances, diatom concentrations, and HPLC data between the pollen-derived climate zones. Significance levels are indicated in brackets.

	Diatom relative abundance	Diatom relative abundance (sqrt)	Diatom Concentration	Diatom Concentration (sqrt)	HPLC concentration	HPLC concentration (sqrt)
Global R	0.25 (0.001)	0.28 (0.001)	0.23 (0.001)	0.24 (0.001)	0.23 (0.001)	0.31 (0.001)
Modern, Post-HTM	0.02 (0.294)	0.08 (0.12)	0.21 (0.008)	0.27 (0.003)	0.39 (0.001)	0.58 (0.001)
Modern, HTM	0.35 (0.002)	0.37 (0.001)	0.3 (0.002)	0.27 (0.002)	0.12 (0.1)	0.21 (0.017)
Post-HTM, HTM	0.34 (0.001)	0.33 (0.001)	0.22 (0.001)	0.24 (0.001)	0.21 (0.001)	0.27 (0.001)

Bold cells indicate significance (999 permutations). HPLC: high-performance liquid chromatography; HTM: Holocene thermal maximum.

Table 4.

Summary of the one-way SIMPER tests (Bray–Curtis dissimilarity) on Charland Lake diatom relative abundances, diatom concentrations, and HPLC data between pollen-derived climate zones.

	Diatom Relative Abundance	Diatom Relative Abundance (sqrt)	Diatom Concentration	Diatom Concentration (sqrt)	HPLC concentration	HPLC concentration (sqrt)
Modern, Post-HTM	S. minutulus (32.8%)	D. stelligera (18.8%)	S. minutulus (46.2%)	S. minutulus (26.7%)	Lutein (29.0%)	Myxoxanthophyll (17.4%)
Modern, HTM	S. minutulus (29.7%)	S. minutulus (17.1%)	S. minutulus (46.0%)	S. minutulus (26.3%)	Lutein (27.9%)	Diatoxanthin (17.4%)
Post-HTM, HTM	S. minutulus (22.8%)	S. minutulus (14.6%)	S. minutulus (27.7%)	D. stelligera (17.8%)	Lutein (40.2%)	Lutein (22.3%)

Percent contribution of driving taxa indicated in brackets. HPLC: high-performance liquid chromatography; HTM: Holocene thermal maximum.

Diatoms

Application of pollen climate zones to the diatom assemblages of Charland Lake showed that diatom concentrations were low in both the Pre-HTM zone and the second clay band, but increased in the organic layer between the clay bands during the HTM to ~23 × 10⁸ valves per gram dry weight (Figure 7). Diatom concentrations also increased after the second clay band and remained relatively stable between 284 and 145 cm (range ~9–21 × 10⁸ valves per gram dry weight). The scale-to-diatom index was low within both clay bands, with an increase in scaled chrysophyte abundance in the intervening organic layer. The abundance of scaled chrysophytes increased during the second clay band and continued to increase throughout the HTM and Post-HTM zones. After a Post-HTM zone peak, scaled chrysophyte abundance declined into the Modern zone and remained low to the top of the core (Figure 7).

Figure 7.

Relative abundance of dominant (>5% found in three sections of the core) diatom taxa in Charland Lake arranged by age (cal. yr BP). The diatom taxa are arranged by their weighted average optima based on cumulative depth in the core. The scale-to-diatom index and total diatom concentrations (valves/g dry weight × 10⁸) are also shown. Darker zones in Sediment Type represent organic sedimentation in the piston core. Layered zones represent clay deposits. The dotted lines indicate the pollen-derived climate zones from Figure 3.

The largest shift in diatom assemblages occurred after the second clay band, before the boundary of the HTM and Post-HTM zones. Prior to this point, the Pre-HTM and HTM zones were predominated by benthic taxa, while the Post-HTM and Modern zones were composed mainly of planktonic taxa. Only one sample from the Pre-HTM zone contained enough diatoms for enumeration, so this zone was removed from the analysis.

HTM taxa prior to the second clay band included Staurosira construens and Staurosirella pinnata. Lindavia intermedia appeared as an important taxon immediately before and after this clay band. After the resumption of organic sedimentation, Stephanodiscus minutulus was the predominant taxon for the remainder of the zone. The importance of Stephanodiscus minutulus increased in the Post-HTM zone and remained common until ~220 cm when relative abundance of Discostella stelligera increased. This pattern was reversed at ~105 cm. Diatom assemblages within the Modern zone were composed largely of Stephanodiscus minutulus and Stephanodiscus parvus; however, Discostella stelligera abundance increased again at ~70 cm until it became a subdominant species at the top of the core (Figure 7).

The difference in diatom assemblages between the pollen-derived climate zones were assessed using ANOSIM tests. The null hypothesis of no difference between diatom assemblages in the HTM, Post-HTM, and Modern zones was rejected in two of the pair-wise tests comparing the HTM zone with the Post-HTM zone and the HTM zone with the Modern zone (Table 3). SIMPER tests identified Stephanodiscus minutulus as the predominant species contributing to ~23% of the difference in species composition between the HTM and Post-HTM zones (Table 4). Similarly, there was a significant shift in species composition between the HTM and Modern zones with Stephanodiscus minutulus driving species change and contributing to ~30% of the difference in species composition (Table 4).

Discussion

The HTM in northeast Ontario took place from ~7800 to 4500 cal. yr BP based on large changes in pollen data that suggest elevated temperatures. Average temperature reconstructions from the pollen assemblages in Charland Lake showed that the climate was ~1.5–2.0°C warmer than the modern day. In this context, the ‘modern day’ refers to the top of our piston core which corresponds to ~1950. All discussion referring to ‘the modern day’ or ‘present day’ relate to this time frame. Algal response to climate change over the Holocene of this region was variable and somewhat unexpected as the warm HTM zone did not experience the greatest amount of algal abundance. Instead, the cooler Post-HTM and Modern zones showed increased algal production which may be related to various indirect climate and landscape effects. These effects, along with more thorough interpretations of pollen and algal data, are discussed in the forthcoming sections.

Pollen as a climate proxy in northeastern Ontario

Pre-HTM zone (~8200–7800 cal. yr BP)

The climate during this interval was cooler than present as is shown in the temperature reconstruction (Figure 3). This inference of cooler temperatures arises because of the predominance of Picea pollen, combined with a relatively low amount of Pinus spp. pollen. Thermophilous taxa, such as deciduous hardwood trees, were rare in the Pre-HTM zone, further contributing to the inference of cool conditions. In addition, the landscape was likely poorly developed with low vegetation cover and immature, highly inorganic soils, as inferred by the relatively high abundances of ferns and other spore-producing plants. It is likely that the boreal forest had not completely colonized the watershed of Charland Lake so soon after the draining of Glacial Lake Ojibway ~8200 cal. yr BP (Liu, 1990; Prentice et al., 1991).

HTM zone (~7800–4500 cal. yr BP)

Reconstructed temperature shows a sharp increase by ~2°C at the onset of this zone ~7800 cal. yr BP. This climate shift is most driven in the decline of Picea and the increase in Pinus banksiana/resinosa, followed by increases in Pinus strobus and Cupressaceae at ~6000 cal. yr BP. This is similar to patterns recorded in the mid-Holocene of Lake Six as described by Liu (1990). This two-phase HTM likely represents an increase in precipitation levels after ~6000 cal. yr BP as inferred by the increase of Cupressaceae.

Interpretation of the Cupressaceae peak is difficult, as this pollen morphotype cannot be easily distinguished between Juniperus and Thuja based on morphological characteristics alone (McAndrews et al., 1973; Yu, 1997). As Juniperus and Thuja are indistinguishable as pollen types, and the two genera have differing climate optima, we have elected to remove Cupressaceae from our temperature reconstructions. It is for this reason as well, that we did not attempt to reconstruct annual precipitation over the Holocene based on the pollen in the Charland Lake core. Juniperus is known to prefer dry habitats, while Thuja prefers wetter conditions (Johnston, 1990; Yu, 1997; Yu et al., 1996). Nonetheless, we infer that Cupressaceae were composed mainly of Thuja at Charland Lake because of the presence of Thuja stomata observed on pollen slides. This interpretation agrees with that of Liu (1990) who inferred that Thuja represented most Cupressaceae pollen at Lake Six based on the occurrence of fossil Thuja seeds. Carcaillet et al. (2001) also inferred the presence of Thuja rather than Juniperus because charcoal analysis shows that forest fire frequency did not increase in the area during the HTM which would be consistent with a wetter, Thuja-rich environment.

Pinus strobus, which increased in relative abundance alongside Cupressaceae likely expanded northward from the mixed Great Lakes–St Lawrence forest as average temperature increased (Bartlein et al., 1984; Hall et al., 1994; Liu, 1990; Richard, 1980; Terasmae and Anderson, 1970). Concomitant changes in Pinus strobus and Cupressaceae during the HTM also argue for the presence of Thuja, as competition for the drier upland areas would not have allowed the coexistence of Juniperus and Pinus strobus (Liu, 1990). Thuja would have had minimal habitat competition with Pinus strobus, allowing the species to coexist (Fowells, 1965; Liu, 1990). It is therefore likely that Thuja proliferated in the widespread wetlands of the Clay Belt lowlands, while Pinus strobus occupied dry upland sites (Liu, 1990).

We infer that the HTM manifested in two phases in the Charland Lake region. The first phase, from ~7800 to 6000 cal. yr BP, saw the increase of Pinus banksiana/resinosa and increased average temperature ~2°C compared with the previous zone. The second phase occurred from ~6000 to 4500 cal. yr BP and experienced increases in Pinus strobus, a large increase in Thuja-inferred Cupressaceae, and a further increase in temperature by ~0.5–1.0°C to a maximum of ~3°C. This reconstruction makes the maximum HTM temperature ~1.5–2.0°C warmer than current conditions, a value which is in line with temperature estimates provided by climate modeling for this region (Renssen et al., 2009, 2012). We also postulate that the second phase of the HTM was wetter than present in this region, as Thuja would have inhabited widespread wetlands which developed during this time. This interpretation is consistent with Prentice et al. (1991) whose precipitation reconstructions of the Holocene of eastern North America shows an increase in annual precipitation in northeast Ontario after ~6000 cal. yr BP.

The Post-HTM zone (~4500–1700 cal. yr BP)

The Post-HTM zone was a transitional time between the warmer HTM and cooler modern climate regimes. The reconstructed temperature shows a decline in temperature from the maximum Holocene value of ~3.0°C at ~5000 cal. yr BP to ~1.5°C ~4500 cal. yr BP. Average temperature remained around 1.5°C for the majority of this zone. Floristically, this zone saw clear decreases in Pinus strobus and Cupressaceae and increases in Picea and Pinus banksiana/resinosa. In addition, the observed increase in Betula may suggest increased fire activity and drier conditions which in turn would not allow for the continued high abundance of Thuja after ~3700 cal. yr BP (Carcaillet et al., 2001; Supplemental Figure 3, available online).

Modern zone (~1700–0 cal. yr BP)

During the Modern interval, Picea increased, while Pinus strobus declined further, contributing to the inferred decrease in average temperature to ~1.0°C. Betula and other common and uncommon arboreal tree types remained largely unchanged throughout this period. It is quite likely that this climate zone experienced an increase in precipitation as represented by the increased abundance in Picea and decrease in Pinus banksiana/resinosa (Liu, 1990; Prentice et al., 1991).

Lake and algal responses to Holocene climate change

Definition of climate zones based on historical changes in terrestrial vegetation allows us to evaluate both how regional climate variation may have affected the linkage between land and water, and the responses of past changes in the production and community composition of aquatic primary producers. Here we evaluate changes in community composition of phototrophs (using fossil pigments and diatom assemblages) to better understand changes in climate on aquatic environments.

Pre-HTM zone (~8200–7800 cal. yr BP)

Northeastern Ontario was inundated by Glacial Lake Ojibway during the early Holocene. This glacial lake deposited thick clay layers which form the Great Clay Belt in the Cochrane District of northern Ontario before draining into the Tyrrell Sea ~8200 cal. yr BP (Roy et al., 2011, 2015; Veillette, 1994) resulting in the formation of Charland Lake. Algal production (both pigments and diatoms) was too low to be detected or analyzed statistically in the Pre-HTM zone. This may be because of the deposition of the first clay layer during this time. Clastic or mineral-rich sedimentation is common in the early ontogeny of post-glacial boreal lakes (Liu, 1990; Teller et al., 2018) and is known to interrupt fossil preservation and signals of algal abundance when sedimentary organic matter is extremely low (Bunting et al., 2010; Leavitt and Hodgson, 2001).

HTM zone (~7800–4500 cal. yr BP)

The climate rapidly warmed at the onset of this zone as shown in the temperature reconstruction based on pollen (Figure 3), but pollen data suggest a two-phase warming event; warm and dry conditions followed by warm and wet conditions. Charland Lake continued to experience major clastic input at the beginning of this zone, although microfossils from algae and cyanobacteria begin to appear at this time. High abundance of Staurosira construens and Staurosirella pinnata during the early HTM suggests low-light or turbid conditions (Bradshaw et al., 2000; Fritz et al., 2004; Haworth, 1976) consistent with high clastic influx. These benthic diatoms are known to be tolerant of low-light environments (Kingsbury et al., 2012; Punning and Puusepp, 2007) which, along with the low concentrations of both diatoms and pigments, suggests an unproductive and light-limited environment. Influx of clastic material in Charland Lake may reflect high rates of terrestrial erosion from the undeveloped landscape. The young forest surrounding Charland Lake was likely open during the early HTM, favoring high rates of weathering and transport of clay particles to the lake (Almquist-Jacobson et al., 1992; Dearing, 1983; Dearing and Foster, 1986).

Deposition of organic matter begins in the HTM, although algal fossils remain rare until ~7000 cal. yr BP (Figure 5). The appearance of fossil pigments from many but not all groups was likely a response to HTM warming, soil development, and the influx of organic matter that would have increased oxygen consumption potentially resulting in anoxia; a better environment for pigment preservation (Leavitt, 1993; Leavitt et al., 2003; Leavitt and Hodgson, 2001). Consistent with this pattern, the ratio of chlorophyll a to pheophytin a suggests a high degree of pigment preservation (Leavitt and Hodgson, 2001) following the period of high clastic influx. Lower light penetration during the final period of fine inorganic matter influx may have reduced photo-oxidation of pigments and favored elevated fossil concentrations (Furlong and Carpenter, 1988; Hurley and Armstrong, 1990, 1991; Leavitt, 1993). Finally, inferred low water levels (based on the high abundances of benthic diatoms) may have favored development of benthic mats of chlorophytes and cyanobacteria (as canthaxanthin). Typically, labile pigments (such as Chl a) are preserved better if they are produced in benthic mats (Leavitt et al., 2003, 1994b).

At ~6000 cal. yr BP, Thuja-inferred Cupressaceae increases rapidly, signifying the onset of wet conditions (Figure 3). Several major phycological and sedimentological changes coincided with this wetter climate. First, elevated organic matter sedimentation persisted until ~6100 cal. yr BP and was characterized by increased pigment concentrations, elevated relative abundance of planktonic diatoms, and an increase in chrysophyte scales indicative of deeper lake conditions (Zeeb and Smol, 2001; Figures 5 and 7). Contemporaneous with increased temperatures and lake levels, landscape stabilization was occurring, as indicated by a hiatus in clastic sedimentation which suggests forest closure and reduced erosion (Liu, 1990). The transition to gyttja-based sediments also suggests the presence of soluble nutrients leaching from fresh, organic soils (Liu, 1990).

Organic sedimentation was interrupted by the deposition of the second clay layer ~6000–5500 cal. yr BP, at which time concentrations of phototrophic fossils declines and diatom assemblages reverted to benthic taxa. Paradoxically, high Cupressaceae pollen abundance suggests a wet climate and high lake levels. We propose that water levels increased sufficiently to let Charland Lake join with surrounding aquatic ecosystems, including nearby Fredrick House Lake, a site which is only ~2 m below the study basin. As Fredrick House Lake is turbid, conjoined waters may have introduced clastic material from the Fredrick House Lake catchment into Charland Lake. Aquatic conditions in this second clay band were similar to those seen in the early HTM Period, where diatom assemblages were composed mainly of benthic species and pigment and frustule concentrations were low.

After ~5500 cal. yr BP, organic sedimentation resumed with no noticeable change in pollen assemblages. In contrast, diatom assemblages rapidly shifted to planktonic taxa, similar to that observed within the organic layer between clay bands, but with elevated abundance of the eutrophic taxon Stephanodiscus minutulus. Stephanodiscus taxa are common in productive waters (Cumming et al., 2015) and compete for silica better than other planktonic species (Mechling and Kilham, 1982). Past research has associated higher water levels with lower Si:P ratios and an associated dominance of Stephanodiscus (Kilham and Kilham, 1990). The eutrophication may reflect increased internal loading of nutrients from anoxic bottom water, as indicated by the presence of okenone, a pigment from purple sulfur bacteria. These prokaryotic taxa are obligate anaerobes and are present only when light penetrates to anoxic environments (see Leavitt et al., 1989). Regardless of the mechanism, the increase in nutrients must have been relatively minor, as neither frustule concentration nor that of most pigments suggests a strong increase in primary production during this period.

Post-HTM zone (~4500–1700 cal. yr BP)

Reconstructed average temperature shows that this zone experienced a cooler climate than the HTM but was still warmer than modern conditions. Algal groups responded to these climate changes; both diatom frustules and pigment concentrations increased moderately, possibly signaling an overall increase in primary production. The eutrophic Stephanodiscus minutulus continued to dominate the diatom assemblage until ~3900 cal. yr BP when it was replaced by the more oligotrophic Discostella stelligera (Figure 7). Presently, we are uncertain whether observed changes in fossil patterns reflect direct effects of climate, or indirect variation in lake structure, such as a change in seasonality and mixing regimes (Bradbury et al., 2002; Dean et al., 1994; Saros et al., 2012; Wiltse et al., 2016). The interpretation of longer periods of stratification is suggested by changes in the chrysophyte scale-to-diatom index, as chrysophytes have a competitive advantage over diatoms when the water column is strongly stratified (Eimers et al., 2009). An increase in the abundance of chrysophytes has also been related to less eutrophic conditions, which agrees with an increase in Discostella stelligera (Bradbury et al., 2002; Reynolds, 1988).

Despite a trend toward oligotrophication as suggested by diatoms, interpretation of fossil pigment data suggests an increase in algal production during this period. Several mechanisms could underlie this pattern. First, increased thermal stratification can lead to the formation of metalimnetic blooms which tend to be over-represented in lake sediments (Leavitt et al., 1989; Leavitt and Hodgson, 2001). Alternately, changes in the seasonality of lake stratification could favor production of phototrophic groups other than diatoms. For example, both lutein–zeaxanthin, a stable indicator of green algae and cyanobacteria, and alloxanthin, a stable indicator of cryptophytes, had significantly higher concentrations (Figure 6) during this period compared with the HTM zone. Lower concentrations of other pigments from cyanobacteria (e.g. echinenone, canthaxanthin, myxoxanthophyll) suggest limited contributions of those taxa to the lutein–zeaxanthin signal (Brock et al., 2006). Both cryptophytes and chlorophytes (together ‘flagellates’) tend to replace negatively buoyant diatoms during periods of thermal stratification, because of their motility (Hickman, 1974; Reynolds, 1984). As such, higher concentrations of these algal groups are consistent with an interpretation of a longer period of thermal stratification during the Post-HTM Period and, perhaps, reduced diatom recruitment, which favored growth of other phytoplankton groups, such as cryptophytes and chlorophytes.

Modern zone (~1700–0 cal. yr BP)

At Charland Lake, the Modern Period of northeastern Ontario was characterized by a cold and wet environment based on the decrease of Pinus strobus and the dominance of Picea. However, algal proxies from Charland Lake suggest the onset of more eutrophic conditions as concentrations of diatoms, and pigments characteristic of diatoms (fucoxanthin, diatoxanthin), were highest during this period. Statistical analyses of individual pigments revealed that canthaxanthin and okenone increased significantly (Figure 6), indicating an abundance of colonial cyanobacteria and purple sulfur bacteria, respectively (Leavitt and Hodgson, 2001). Given that the Chl a:pheophytin ratio declined slightly prior to the Modern Period (Figure 5), but okenone increased, we infer that these changes may reflect in part a minor change in the preservation environment at Charland Lake. However, because this interval is also marked by the first occurrence of chemically robust myxoxanthophyll, a ubiquitous compound in colonial cyanobacteria, we infer that the Modern zone has also been productive.

The movement toward colder conditions during this period may have also promoted a change in lake seasonality, with potentially more lake mixing and less-stable thermal stratification, allowing some algal species, such as Stephanodiscus minutulus, to proliferate (Bradbury et al., 2002; Kilham et al., 1986; Makulla and Sommer, 1993; Reynolds and Reynolds, 1985). Growth requirements of Stephanodiscus species may have been largely satisfied by the regeneration of phosphorus from nutrient-dense water to the photic zone during spring circulation, resulting in large blooms of Stephanodiscus minutulus and Stephanodiscus parvus (Bradbury et al., 2002; Kilham and Kilham, 1978). The inference of extended periods or vigorous lake mixing is also supported by the decrease in the chrysophyte scale-to-diatom index, further suggesting less stratification (Reynolds, 1984; Rott, 1984).

Interpretation of the influence of climate on canthaxanthin (Nostocales cyanobacteria) and okenone (purple sulfur bacteria) is not straightforward, as both of these algal groups would be expected to decrease under cool and wet conditions (Lami et al., 2009). Okenone is found in anaerobic purple sulfur bacteria which prefer the oxic–anoxic interface in lakes (Massé et al., 2002), hence higher concentrations of okenone can be related to seasonal anoxia in lakes (Maheaux et al., 2016). However, low concentrations of purple sulfur bacteria have been related to cool conditions, likely coupled with increased mixing of the water column (Schmidt et al., 2002). Overall, concentrations of okenone are much lower than those seen in meromictic or strongly stratified ecosystems (Leavitt et al., 1989) and suggests the presence of only seasonal (summer) anoxia. Furthermore, canthaxanthin, an indicator of Nostocales cyanobacteria, is often associated with high nutrients in the water column (Kleppel et al., 1988; Lami et al., 2009), and can also be a marker of N₂ fixation (Hayes et al., in press). One possibility is that the Modern Period experienced in an influx of dissolved organic matter (DOM) which may have stimulated heterotrophic growth of the cyanobacterial groups (Stevenson et al., 2016). Regardless, it seems likely that factors other than climate must be considered to understand how limnological conditions became ideal for these groups in the Modern Period.

Conclusion

Pollen data from Charland Lake suggest that the HTM in northeast Ontario took place from ~7800 to 4500 cal. yr BP. The onset of warmth is ~1000 years later than recorded in northwest Ontario (Moos and Cumming, 2011, 2012; Moos et al., 2009; Karmakar et al., 2015a, 2015b). Differences between our findings and Liu’s (1990) results may be because of the more rigorous carbon dating model used in our study as our pollen analysis agrees with Liu’s study quite well. Analysis of fossil pollen showed that the HTM in northeastern Ontario was warm and dry from ~7800 to 6000 cal. yr BP and then became wetter, but still warm from ~6000 to 4500 cal. yr BP as indicated by the high abundance of Thuja-inferred Cupressaceae pollen. Changes in algal abundance was complex over the Holocene, as inferred from both fossil pigment and diatoms.

This work represents one of the first investigations in northeast Ontario to assess the relationship between climate and in-lake production during the Holocene epoch. Overall, the interaction between climate and limnological conditions were complex within Charland Lake, providing a striking contrast with lakes in northwest Ontario at millennial time scales. In general, we demonstrate that limnological conditions were more optimal for algae over the last ~4500 years than in the preceding millennia. This pattern is in stark contrast to those seen in lakes from northwest Ontario, where paleolimnological studies have linked the warmer HTM to enhanced lake water production (Karmakar et al., 2015b; Moos et al., 2009). In Charland Lake, such changes did not indicate the warmer temperatures of the HTM directly influenced lake water production, but warmer and wetter conditions in the Post-HTM Period provides some evidence of enhanced production. Similarly, production during the Modern Period is higher than the HTM. These findings suggest that the link between climate and production can be very complex. If this were broadly true, it suggests that lakes in northeast Ontario may respond differently to future climate change compared with lakes elsewhere in northern Ontario, and that the response and susceptibility of lakes to climate change can vary across an ecozone.

Supplemental Material

SFigure_1_Gamma_dating_curves – Supplemental material for Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario, Canada

Supplemental material, SFigure_1_Gamma_dating_curves for Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario, Canada by Brett G Elmslie, Cale AC Gushulak, Maxime P Boreux, Scott F Lamoureux, Peter R Leavitt and Brian F Cumming in The Holocene

Supplemental Material

SFigure_2_Bstick_model – Supplemental material for Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario, Canada

Supplemental material, SFigure_2_Bstick_model for Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario, Canada by Brett G Elmslie, Cale AC Gushulak, Maxime P Boreux, Scott F Lamoureux, Peter R Leavitt and Brian F Cumming in The Holocene

Supplemental Material

SFigure_3_Pollen_Concentration – Supplemental material for Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario, Canada

Supplemental material, SFigure_3_Pollen_Concentration for Complex responses of phototrophic communities to climate warming during the Holocene of northeastern Ontario, Canada by Brett G Elmslie, Cale AC Gushulak, Maxime P Boreux, Scott F Lamoureux, Peter R Leavitt and Brian F Cumming in The Holocene

Footnotes

Acknowledgements

The authors thank the LacCore team at the University of Minnesota and Tom Brown at the Lawrence Livermore National Laboratory for sample preparation and carbon dating. They also thank Matthew Peros for his helpful advice on pollen sample preparation and Kathleen Laird for her assistance with diatom processing and taxonomy. They thank Deirdre Bateson at U. Regina for analysis of fossil pigments. They extend their special thanks to Graham Mushet, Cécilia Barouillet, and Gladys Kong for their help in the field. The authors also thank the two anonymous reviewers whose comments greatly improved this manuscript.

Funding

The author(s) received the following financial support for the research, authorship and/or publication of this article: Funding for this project was provided by an NSERC Discovery Grant to BFC, SFL, and PRL; an NSERC PGS-D scholarship to CACG; and funding from the Canada Research Chair and Canada Foundation for Innovations programs.

ORCID iDs

Cale AC Gushulak

Maxime P Boreux

Supplemental material

Supplemental material for this article is available online.

References

Almquist-Jacobson

Almendinger

Hobbie

(1992) Influence of terrestrial vegetation on sediment-forming processes in kettle lakes of west-central Minnesota. Quaternary Research 38: 103–116.

Bartlein

Webb

III Fleri

(1984) Holocene climatic change in the northern Midwest: Pollen-derived estimates. Quaternary Research 22: 361–374.

Bassett

Crompton

Parmelee

(1978) An Atlas of Airborne Pollen Grains and Common Fungus Spores of Canada. Hull, QC, Canada: Canada Department of Agriculture, Ministry of Supply and Services Canada.

Bennett

Willis

(2001) Pollen. In: Smol

Birks

HJB

Last

(eds) Tracking Environmental Change Using Lake Sediments: Terrestrial, Algal, and Siliceous Indicators, vol. 3. Dordrecht: Kluwer Academic Publishers, pp. 5–32.

Blaauw

Christen

(2011) Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Analysis 6: 457–474.

Bond

Kromer

Beer

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

Bradbury

Cumming

Laird

(2002) A 1500-year record of climatic and environmental change in Elk Lake, Minnesota III: Measures of past primary productivity. Journal of Paleolimnology 27: 321–340.

Bradshaw

Jones

Birks

HJB

, et al. (2000) Diatom responses to late-glacial and early-Holocene environmental changes at Kråkenes, western Norway. Journal of Paleolimnology 23: 21–34.

Briner

McKay

Axford

, et al. (2016) Holocene climate change in Arctic Canada and Greenland. Quaternary Science Reviews 147: 340–364.

10.

Brock

Leavitt

Schindler

, et al. (2006) Spatial variability of stable isotopes and fossil pigments in surface sediments of Alaskan coastal lakes: Constraints on quantitative estimates of past salmon abundance. Limnology and Oceanography 51: 1637–1647.

11.

Brown

Nelson

Mathewes

, et al. (1989) Radiocarbon dating of pollen by accelerator mass spectrometry. Quaternary Research 32: 1205–1212.

12.

Bunting

Leavitt

Weidman

, et al. (2010) Regulation of the nitrogen biogeochemistry of mountain lakes by subsidies of terrestrial dissolved organic matter and the implications for climate studies. Limnology and Oceanography 55: 333–345.

13.

Carcaillet

Bergeron

Richard

, et al. (2001) Change of fire frequency in the eastern Canadian boreal forests during the Holocene: Does vegetation composition or climate trigger the fire regime? Journal of Ecology 89: 930–946.

14.

Carlson

LeGrande

Oppo

, et al. (2008) Rapid early Holocene deglaciation of the Laurentide ice sheet. Nature Geoscience 1: 620–624.

15.

Clarke

Warwick

(1994) Similarity-based testing for community pattern: The two-way layout with no replication. Marine Biology 118: 167–176.

16.

Cumming

Laird

Gregory-Eaves

, et al. (2015) Tracking past changes in lake-water phosphorus with a 251-lake calibration dataset in British Columbia: Tool development and application in a multiproxy assessment of eutrophication and recovery in Osoyoos Lake, a transboundary lake in Western North America. Frontiers in Ecology and Evolution 3: 1–18.

17.

Cumming

Wilson

Hall

, et al. (1995) Diatoms From British Columbia (Canada) Lakes and Their Relationship to Salinity, Nutrients, and Other Limnological Variables (Bibliotheca Diatomologica, Band 31). Stuttgart: Schweizerbart Science.

18.

Dean

Bradbury

Anderson

, et al. (1994) A high-resolution record of climatic change in Elk Lake, Minnesota for the last 1500 years (Report no. 94-578). Reston, VA: U.S. Geological Survey.

19.

Dearing

(1983) Changing patterns of sediment accumulation in a small lake in Scania, southern Sweden. Hydrobiologia 103: 59–64.

20.

Dearing

Foster

IDL

(1986) Lake sediments and paleohydrological studies. In: Berglund

(ed.) Handbook of Holocene Palaeoecology and Palaeohydrology. Chichester: John Wiley & Sons, pp. 67–90.

21.

Douglas

MSV

Smol

(1999) Freshwater diatoms as indicators of environmental change in the High Arctic. In: Stoermer

Smol

(eds) The Diatoms: Applications for the Environment and Earth Sciences. Cambridge: Cambridge University Press, pp. 227–244.

22.

Eimers

Watmough

Paterson

, et al. (2009) Long-term declines in phosphorus export from forested catchments in south central Ontario. Canadian Journal of Fisheries and Aquatic Sciences 66: 1682–1692.

23.

Environment Canada (2019) Canadian Climate Normals 1981-2010 Station Data. Available at: http://climate.weather.gc.ca/climate_normals/results_1981_2010_e.html?searchType=stnName&txtStationName=Timmins&searchMethod=contains&txtCentralLatMin=0&txtCentralLatSec=0&txtCentralLongMin=0&txtCentralLongSec=0&stnID=4180&dispBack=1 (accessed 26 June 2019).

24.

Esri (2015) Arcgis Desktop: Release 10.5. Redlands, CA: Environmental Systems Research.

25.

Fowells

(1965) Silvics of Forest Trees of the United States (Number 271). Washington, DC: United States Department of Agriculture.

26.

Fritz

Engstrom

Juggins

(2004) Patterns of early lake evolution in boreal landscapes: A comparison of stratigraphic inferences with a modern chronosequence in Glacier Bay, Alaska. The Holocene 14: 828–840.

27.

Furlong

Carpenter

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

28.

Glew

(1989) A new trigger mechanism for sediment samplers. Journal of Paleolimnology 2: 241–243.

29.

Glew

Smol

Last

(2001) Sediment core collection and extrusion. In: Last

Smol

(eds) Tracking Environmental Change Using Lake Sediments, Basin Analysis Coring and Chronological Techniques, vol. I. Dordrecht: Kluwer Academic Publishers, pp. 73–105.

30.

Goring

Dawson

Simpson

, et al. (2015) Neotoma: A programmatic interface to the neotoma paleoecological database. Open Quaternary 1: 1–17.

31.

Grimm

(1987) CONISS: A FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Computers and Geosciences 13: 13–35.

32.

Grimm

(2004) TILIA and TILIA GRAPH Computer Programs (Version 2.0.2). Springfield, IL: Illinois State Museum Research and Collections Center.

33.

Hall

Duff

Quinby

(1994) A 10,000-Year Vegetation History of the Temagami Region of Ontario With Special Emphasis on White Pine. Temagami, ON, Canada: Temagami Wilderness Society.

34.

Hall

Leavitt

Quinlan

, et al. (1999) Effects of agriculture, urbanization and climate on water quality in the northern Great Plains. Limnology and Oceanography 43: 739–756.

35.

Hammer

Harper

(2006) Paleontological Data Analysis. Malden, MA: Blackwell Publishing.

36.

Hammer

Harper

Ryan

(2001) PAST: Palaeontological statistics software package for education and data analysis. Palaeontologica Electronica 4: 1–9.

37.

Haworth

(1976) Two late-glacial (Late Devensian) diatom assemblage profiles from northern Scotland. New Phytology 77: 227–256.

38.

Hayes

Patoine

Haig

, et al. (in press) Spatial and temporal variation in nitrogen fixation and its importance to phytoplankton growth in phosphorus-rich lakes. Freshwater Biology 64: 269–283.

39.

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.

40.

Hickman

(1974) The seasonal succession and vertical distribution of the phytoplankton in Abbot’s Pond, North Somerset, U.K. Hydrobiologia 44: 127–147.

41.

Hodgson

Wright

Tyler

, et al. (1998) Analysis of fossil pigments from algae and bacteria in meromictic Lake Fidler, Tasmania, and its application to lake management. Journal of Paleolimnology 19: 1–22.

42.

Hurley

Armstrong

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

43.

Hurley

Armstrong

(1991) Pigment preservation in lake sediments: A comparison of sedimentary environments in Trout Lake, Wisconsin. Canadian Journal of Fisheries and Aquatic Sciences 48: 472–486.

44.

Johnston

(1990) Thuja occidentalis L. Northern white-cedar. Silvics of North America 1: 580–589.

45.

Juggins

(2003) C2 Software for Ecological and Palaeoecological Data Analysis and Visualization User Guide 1.3. Newcastle: University of Newcastle.

46.

Juggins

(2015) Rioja: Analysis of quaternary science data (R package version (0.9-6)). Available at: https://cran.r-project.org/web/packages/rioja/index.html.

47.

Kapp

Davis

King

(2000) Pollen and Spores. 2nd edition. College Station, TX: The American Association of Stratigraphic Palynologists.

48.

Karmakar

Laird

Cumming

(2015a) Diatom-based evidence of regional aridity during the mid-Holocene period in boreal lakes from northwest Ontario (Canada). The Holocene 25: 166–177.

49.

Karmakar

Leavitt

Cumming

(2015b) Enhanced algal abundance in northwest Ontario (Canada) lakes during the warmer early-to mid-Holocene period. Quaternary Science Reviews 123: 168–179.

50.

Kilham

(1990) Endless summer: Internal loading processes dominate nutrient cycling in tropical lakes. Freshwater Biology 23: 379–389.

51.

Kilham

Hecky

(1986) Hypothesized resource relationships among African planktonic diatoms. Limnology and Oceanography 31: 1169–1181.

52.

Kilham

(1978) Natural community bioassays: Predictions of results based on nutrient physiology and competition. Verhandlungen des Internationalen Verein Limnologie 20: 68–74.

53.

Kingsbury

Laird

Cumming

(2012) Consistent patterns in diatom assemblages and diversity measures across water-depth gradients from eight Boreal lakes from north-western Ontario (Canada). Freshwater Biology 57: 1151–1165.

54.

Klemm

Herzschuh

Pestryakova

(2016) Vegetation, climate and lake changes over the last 7000 years at the boreal treeline in north-central Siberia. Quaternary Science Reviews 147: 422–434.

55.

Kleppel

Frazel

Pieper

, et al. (1988) Natural diets of zooplankton off southern California. Marine Ecology Progress Series 49: 231–241.

56.

Krammer

Lange-Bertalot

(1986) Bacillariophyceae. 1: Teil: Naviculaceae (Band 2/1, eds Ettl

Gärtner

Gerloff

, et al.). Stuttgart: Süßwasserflora von Mitteleuropa.

57.

Krammer

Lange-Bertalot

(1988) Bacillariophyceae. 2: Teil: Bacillariaceae, Epithmiaceae, Surirellaceae (Band 2/2, eds Ettl

Gärtner

Gerloff

, et al.). Stuttgart: Süßwasserflora von Mitteleuropa.

58.

Krammer

Lange-Bertalot

(1991a) Bacillariophyceae. 3: Teil: Centrales, Fragilariaceae, Eunotiaceae (Band 2/3, Ettl

Gärtner

Gerloff

, et al.). Stuttgart: Süßwasserflora von Mitteleuropa.

59.

Krammer

Lange-Bertalot

(1991b) Bacillariophyceae. 4: Teil: Achnanthaceae (Band 2/4, eds Ettl

Gärtner

Gerloff

, et al.). Stuttgart: Süßwasserflora von Mitteleuropa.

60.

Lami

Musazzi

Marchetto

, et al. (2009) Sedimentary pigments in 308 alpine lakes and their relation to environmental gradients. Advances in Limnology 62: 247–268.

61.

Leavitt

(1993) A review of factors that regulate carotenoid and chlorophyll deposition and fossil pigment abundance. Journal of Paleolimnology 9: 109–127.

62.

Leavitt

Hodgson

(2001) Sedimentary pigments. In: Smol

Birks

HJB

Last

(eds) Tracking Environmental Change Using Lake Sediments: Terrestrial, Algal, and Siliceous Indicators, vol. 3. Dordrecht: Kluwer Academic Publishers, pp. 295–325.

63.

Leavitt

Carpenter

Kitchell

(1989) Whole-lake experiments: The annual record of fossil pigments and zooplankton. Limnology and Oceanography 34: 700–717.

64.

Leavitt

Cumming

Smol

, et al. (2003) Climatic control of UV radiation effects on lakes. Limnology and Oceanography 48: 2062–2069.

65.

Leavitt

Hann

Smol

, et al. (1994a) Paleolimnological analysis of whole-lake experiments: An overview of results from Experimental Lakes Area Lake 227. Canadian Journal of Fisheries and Aquatic Science 51: 2322–2332.

66.

Leavitt

Schindler

Paul

, et al. (1994b) Fossil pigment records of phytoplankton in trout-stocked alpine lakes. Canadian Journal of Fisheries and Aquatic Science 51: 2411–2423.

67.

Leavitt

Vinebrooke

Donald

, et al. (1997) Past ultraviolet radiation environments in lakes derived from fossil pigments. Nature 388: 457–459.

68.

Levec

(2001) B.A.S.S.2.0 User Manual: Inventroy, Monitoring, and Assessment. Peterborough, ON: Ontario Ministry of Natural Resources and Forestry.

69.

Liu

(1990) Holocene paleoecology of the boreal forest and Great Lakes-St. Lawrence forest in northern Ontario. Ecological Monographs 60: 179–212.

70.

Lotter

Pienitz

Schmidt

(1999) Diatoms as indicators of environmental change near arctic and alpine treeline. In: Stoermer

Smol

(eds) The Diatoms: Applications for the Environmental and Earth Sciences. Cambridge: Cambridge University Press, pp. 205–226.

71.

McAndrews

Berti

Norris

(1973) Key to the Quaternary Pollen and Spores of the Great Lakes Region. Toronto, ON, Canada: Life Science Miscellaneous Publications.

72.

Maheaux

Leavitt

Jackson

(2016) Asynchronous onset of eutrophication among shallow prairie lakes of the northern Great Plains, Alberta, Canada. Global Change Biology 22: 271–283.

73.

Makulla

Sommer

(1993) Relationships between resource ratios and phytoplankton species composition during spring in five north German lakes. Limnology and Oceanography 38: 846–856.

74.

Margold

Stokes

Clark

(2018) Reconciling records of ice streaming and ice margin retreat to produce a palaeogeographic reconstruction of the deglaciation of the Laurentide Ice Sheet. Quaternary Science Reviews 189: 1–30.

75.

Massé

Pringault

De Wit

(2002) Experimental study of interactions between purple and green sulfur bacteria in sandy sediments exposed to illumination deprived of near-infrared wavelengths. Applied and Environmental Microbiology 68: 2972–2981.

76.

Mechling

Kilham

(1982) Temperature effects on silicon limited growth of the Lake Michigan diatom Stephanodiscus minutus (Bacillariophyceae). Journal of Phycology 18: 199–205.

77.

Moos

Cumming

(2011) Changes in the parkland-boreal forest boundary in northwestern Ontario over the Holocene. Quaternary Science Reviews 30: 1232–1242.

78.

Moos

Cumming

(2012) Climate–fire interactions during the Holocene: A test of the utility of charcoal morphotypes in a sediment core from the boreal region of north-western Ontario (Canada). International Journal of Wildland Fire 2: 1640–1652.

79.

Moos

Laird

Cumming

(2005) Diatom assemblages and water depth in Lake 239 (Experimental Lakes Area, Ontario): Implications for paleoclimatic studies. Journal of Paleolimnology 34: 217–227.

80.

Moos

Laird

Cumming

(2009) Climate-related eutrophication of a small boreal lake in northwestern Ontario: A palaeolimnological perspective. The Holocene 19: 359–367.

81.

Navarro

Harvey

AÉ

Ali

, et al. (2018) A Holocene landscape dynamic multiproxy reconstruction: How do interactions between fire and insect outbreaks shape an ecosystem over long time scales? PLoS ONE 13: e0204316.

82.

Overpeck

Webb

Prentice

(1985) Quantitative interpretation of fossil pollen spectra: Dissimilarity coefficients and the method of modern analogs. Quaternary Research 23: 87–108.

83.

Prentice

Bartlein

Webb

(1991) Vegetation and climate change in eastern North America since the last glacial maximum. Ecology 72: 2038–2056.

84.

Price

Alfaro

Brown

, et al. (2013) Anticipating the consequences of climate change for Canada’s boreal forest ecosystems. Environmental Reviews 21: 322–365.

85.

Punning

Puusepp

(2007) Diatom assemblages in sediments of Lake Juusa, Southern Estonia with an assessment of their habitat. Hydrobiologia 586: 27–41.

86.

R Core Team (2015) R: A language and environment for statistical computing. Vienna: The R Foundation. Available at: https://www.R-project.org/.

87.

Reimer

Bard

Bayliss

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

88.

Renssen

Seppä

Crosta

, et al. (2012) Global characterization of the Holocene thermal maximum. Quaternary Science Reviews 48: 7–19.

89.

Renssen

Seppä

Heiri

, et al. (2009) The spatial and temporal complexity of the Holocene thermal maximum. Nature Geoscience 2: 411–414.

90.

Reynolds

(1988) Functional morphology and the adaptive strategies of freshwater phytoplankton. In: Sandgren

(ed.) Growth and Reproductive Strategies of Freshwater Phytoplankton. Cambridge: Cambridge University Press, pp. 388–433.

91.

Reynolds

(1984) The Ecology of Freshwater Phytoplankton. Cambridge: Cambridge University Press.

92.

Reynolds

(1985) The atypical seasonality of phytoplankton in Crose Mere, 1972: An independent test of the hypothesis that variability in the physical environment regulates community dynamics and structure. British Phycological Journal 20: 227–242.

93.

Richard

(1980) Histoire postglaciaire de la végétation au sud du lac Abitibi, Ontario et Québec. Geographie Physique et Quaternaire 34: 77–94.

94.

Ritchie

Cwynar

Spear

(1983) Evidence from north-west Canada for an early Holocene Milankovitch thermal maximum. Nature 305: 126–128.

95.

Rott

(1984) Phytoplankton as biological parameter for the trophic characterization of lakes. Verhandlungen des Internationalen Verein Limnologie 22: 1078–1085.

96.

Roy

Dell’Oste

Veillette

, et al. (2011) Insights on the events surrounding the final drainage of Lake Ojibway based on James Bay stratigraphic sequences. Quaternary Science Reviews 30: 682–692.

97.

Roy

Veillette

Daubois

, et al. (2015) Late-stage phases of glacial Lake Ojibway in the central Abitibi region, eastern Canada. Geomorphology 248: 14–23.

98.

Rühland

Smol

(2005) Diatom shifts as evidence for recent Subarctic warming in a remote tundra lake, NWT, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 226: 1–16.

99.

Saros

Stone

Pederson

, et al. (2012) Climate-induced changes in lake ecosystem structure inferred from coupled neo-and paleoecological approaches. Ecology 93: 2155–2164.

100.

Schmidt

Psenner

Müller

, et al. (2002) Impact of late glacial climate variations on stratification and trophic state of the meromictic lake Längsee (Austria): Validation of a conceptual model by multi proxy studies. Journal of Limnology 61: 49–60.

101.

Shuman

Marsicek

(2016) The structure of Holocene climate change in mid-latitude North America. Quaternary Science Reviews 141: 38–51.

102.

Stevenson

McGowan

Anderson

, et al. (2016) Impacts of forestry planting on primary production in upland lakes from north-west Ireland. Global Change Biology 22: 1490–1504.

103.

Teller

Rühland

Smol

, et al. (2018) Holocene history of Lake of the Woods: Ontario, Manitoba, and Minnesota. GSA Bulletin 130: 3–23.

104.

Terasmae

Anderson

(1970) Hypsithermal range extension of white pine (Pinus strobus L.) in Quebec, Canada. Canadian Journal of Earth Sciences 7: 406–413.

105.

Veillette

(1994) Evolution and paleohydrology of glacial lakes Barlow and Ojibway. Quaternary Science Reviews 13: 945–971.

106.

Viau

Gajewski

(2009) Reconstructing millennial-scale, regional paleoclimates of boreal Canada during the Holocene. Journal of Climate 22: 316–330.

107.

Viau

Gajewski

Sawada

, et al. (2006) Millennial-scale temperature variations in North America during the Holocene. Journal of Geophysical Research: Atmospheres 111: D09102.

108.

Whitlock

Bartlein

(1997) Vegetation and climate change in northwest America during the past 125 kyr. Nature 388: 57–61.

109.

Whitmore

Gajewski

Sawada

, et al. (2005) Modern pollen data from North America and Greenland for multi-scale paleoenvironmental applications. Quaternary Science Reviews 24: 1828–1848.

110.

Williams

Shuman

Bartlein

, et al. (2006) An Atlas of Pollen-Vegetation-Climate Relationships for the United States and Canada. Dallas, TX: American Association of Stratigraphic Palynologists Foundation.

111.

Wiltse

Paterson

Findlay

, et al. (2016) Seasonal and decadal patterns in Discostella (Bacillariophyceae) species from bi-weekly records of two boreal lakes (Experimental Lakes Area, Ontario, Canada). Journal of Phycology 52: 817–826.

112.

Wright

(1967) A square-rod piston sampler for lake sediments. Journal of Sedimentary Research 37: 975–976.

113.

Wright

Mann

Glaser

(1984) Piston corers for peat and lake sediments. Ecology 65: 657–659.

114.

(1997) Late Quaternary paleoecology of Thuja and Juniperus (Cupressaceae) at Crawford Lake, Ontario, Canada: Pollen, stomata and macrofossils. Review of Palaeobotany and Palynology 96: 241–254.

115.

McAndrews

Siddiqi

(1996) Influences of Holocene climate and water levels on vegetation dynamics of a lakeside wetland. Canadian Journal of Botany 74: 1602–1615.

116.

Zeeb

Smol

(2001) Chrysophyte scales and cysts. In: Smol

Birks

HJB

Last

(eds) Tracking Environmental Change Using Lake Sediments: Terrestrial, Algal, and Siliceous Indicators, vol. 3. Dordrecht: Kluwer Academic Publishers, pp. 203–223.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.23 MB

0.01 MB

0.06 MB