Late-Holocene atmospheric dust deposition in eastern Canada (St. Lawrence North Shore)

Abstract

Dust deposition in two ombrotrophic peatlands (Baie and Ile du Havre peatland (IDH) bogs) of the Estuary and Gulf of the St. Lawrence in eastern Canada was reconstructed using elemental geochemistry. The rare earth elements (REEs) and other lithogenic element concentrations were measured by ICP-oES and Q-ICP-MS along two peat cores spanning the last 4000 years. Principal component analyses on the geochemical profiles show that REEs display the same behavior as Al, Ti, Sc, and Zr, all conservative elements, which suggests that REEs are immobile in the studied peat bogs and can be used as tracers of dust deposition. Plant macrofossils were also used to infer past environmental and humidity changes. The dust fluxes were reconstructed using the sum of REEs (ΣREE). The range of dust deposition varies from 0.2 to 3.8 g m⁻² yr⁻¹ in the Baie bog, while the IDH bog shows lower fluxes ranging between 0.1 and 1.2 g m⁻² yr⁻¹. The highest dust fluxes in the Baie bog were recorded from 1750–1000 cal. BP to 600–100 cal. BP and occur at the same time as periods of high variability in the macrofossil record (i.e. successive layers dominated by Sphagnum or Ericaceae). The timing of these events in the dust and macrofossil records also corresponds to documented cold periods. These two periods have been identified as episodes of climatic instability, which could have been caused by changes in the wind regime.

Keywords

atmospheric dust climatic events eastern Canada geochemistry lead-210 peat radiocarbon

Introduction

Mineral dust plays an important role in the global climate system having effects on radiation budgets and the chemical composition of the atmosphere (Goudie and Middleton, 2001; Harrison et al., 2001). Atmospheric mineral dust can affect the biogeochemical cycles of several elements in marine (Meskhidze et al., 2003) and terrestrial environments (Goudie and Middleton, 2001) by supplying nutrients. Fluctuations in atmospheric dust can be indicative of changes in climatic conditions which are related to changes in various factors such as precipitation, temperature, wind regime or vegetation cover.

Paleoclimate records have shown greater dust deposition during glacial compared with interglacial periods. For example, dust deposition was 80 times greater during the Last Glacial Maximum (LGM) in comparison with the Holocene (Fischer et al., 2007). While variations in dust deposition during the Holocene are not of the same magnitude as those occurring on a glacial–interglacial scale, the complex spatial and temporal variability in climate during this period was important enough to affect dust production and transport (Albani et al., 2015; Wanner et al., 2008). Past variability in dust deposition and sources has been reconstructed in a range of archives, for example, ice cores (Albani et al., 2012; Delmonte et al., 2008; Zdanowicz et al., 2000), marine sediments (DeMenocal et al., 2000; Rea, 1994), lake sediments (Mischke et al., 2010), and loess deposits (Bettis et al., 2003; Muhs and Budahn, 2006). Despite the importance of atmospheric dust, the number of continental records of their changing rates and sources remains limited.

Recently, there has been an increased interest in using peatlands as archives of past variability in atmospheric dust deposition (De Vleeschouwer et al., 2014). Ombrotrophic peatlands (peat bogs) receive their nutrients and water exclusively from atmospheric deposition (Chambers and Charman, 2004). Peatlands present other advantages: they are broadly distributed, hence easily accessible, and show relatively high accumulation rates allowing high temporal resolution reconstructions. In comparison with ice cores, which record mainly long-range transport of atmospheric particles, peat bogs also record dust inputs from local and regional sources (Le Roux et al., 2012; Marx et al., 2010).

The geochemical analysis of peatland archives has helped to identify changes in dust deposition linked to the Younger Dryas in Switzerland (Le Roux et al., 2012; Shotyk et al., 2001), SE Asia (Weiss et al., 2002), and Tierra del Fuego (Vanneste et al., 2015); the Sahara aridification (Le Roux et al., 2012); and the ‘Little Ice Age’ (‘LIA’) in Europe (De Vleeschouwer et al., 2009). However, some questions remain as to under which conditions dust events are recorded. De Vleeschouwer et al. (2009) observed increases in dust deposition during ‘LIA’ dry episodes in northern Poland while other studies recorded some peaks during periods of climate instability (wet or dry shifts) (De Jong et al., 2007; De Vleeschouwer et al., 2012). The timing of some dust deposition episodes with known climatic events can also differ. For example, Le Roux et al. (2012) reported a Saharan dust event at 8.4 decoupled from the 8.2 event, which could have played a role in the onset of the latter. Hence, further research is needed to clarify the dust-climate linkage.

Over North America, continental dust records are scarce and mainly consist of loess deposits from the Great Plains, Mississippi Valley, and Alaska (Muhs, 2013), which usually cover glacial–interglacial or millennial time-scales but lack sufficiently high resolution for the Holocene. Furthermore, only a limited number of paleoclimatic studies have been conducted in the maritime peatlands of eastern Canada.

The climate of eastern Canada is currently governed by the relative strength of three major air masses (Bryson, 1966; Bryson and Hare, 1974). Mild, dry Pacific winds enter the region from the west; cold, dry Arctic air arrives from the north and northwest; warm, moist tropical maritime air comes from the south. North-Eastern Canada Holocene climate history has been characterized by a mid-Holocene thermal maximum (~4000 cal. BP; Viau et al., 2006). The proximity of the residual Laurentide Ice Sheet delayed this warm period relative to other regions of boreal North America. Long-term trends in paleohydrology were also linked with other periods of climate change such as the Neoglacial and ‘LIA’ coolings or the ‘Medieval Warm Period’ (‘MWP’) (Garneau et al., 2014; Magnan and Garneau, 2014). The climatic studies available in the region have focused on successional vegetation changes and/or surface humidity in relation to climate (Charman et al., 2015; Hughes et al., 2006; Magnan et al., 2014; Magnan and Garneau, 2014). In Newfoundland, Hughes et al. (2006) suggested that bog surface wetness (BSW) over the Holocene was controlled by a combination of oceanic and solar forcing. Magnan and Garneau (2014) reconstructed late-Holocene contrasted hydrological responses in peat bogs along the Estuary and the Gulf of the St. Lawrence. They suggested that a greater Sub-Arctic influence in the Gulf of the St. Lawrence explained the different responses recorded in each region. These results were based on paleohydrological or paleoecological proxies such as testate amoebae, plant macrofossils, and peat humification and interpreted as proxies of surface humidity, although they can also be influenced by temperature (Barber and Langdon, 2007; Lamentowicz et al., 2010) or differential decomposition in the case of humification (Hansson et al., 2013). Furthermore, they may be partially interdependent and can be influenced by site-specific factors other than climate (Caseldine et al., 2000; De Jong et al., 2010; Magnan and Garneau, 2014; Sullivan and Booth, 2011). In contrast, dust deposition is independent of bog ecohydrological dynamics and is a function of particle availability and atmospheric transport and deposition (Maher et al., 2010). Accordingly, paleodust records in eastern Canada may provide complementary information about past climate variations in the region.

This study is a first attempt to reconstruct dust deposition using two continuous peat bog records from the North Shore of the Gulf and the Estuary of the St. Lawrence, eastern Canada. This paper aims at (1) reconstructing changes in atmospheric dust deposition using lithogenic elements, namely, Al, Ti, Sc, and rare earth elements (REEs); (2) investigating the relationship between dust fluxes and hydroclimatic variability during the late-Holocene using plant macrofossil analyses; (3) reconstructing the paleoclimatic events occurring on the North Shore of the St. Lawrence and comparing these with other regional records.

Materials and methods

Study sites

The Baie peatland (49°04′N, 68°14′W; 1.5 km², 15 m a.s.l.) is a Sphagnum-dominated treeless raised bog located on the Manicouagan delta, 20 km southwest of Baie-Comeau, at the eastern end of the Lower St. Lawrence Estuary (Figure 1). The southern border of the peatland (700 m from the coring site) is open toward the coast of the St. Lawrence Estuary, where beaches occupy the coastline. More than 4.5 m of peat has accumulated over marine silty-clay sediments originating from the Laurentian transgression that ended around 4400–4200 cal. BP (Bernatchez, 2003; Magnan and Garneau, 2014). Deposits at a higher altitude (>14 m a.s.l.) consist of silty-sand deltaic terraces which emerged following the withdrawal of the Goldthwait Sea from 8 kyr (Bernatchez, 2003). A closed boreal forest dominated by Picea mariana, Abies balsamea, and Betula paperyfera covers most of the deltaic deposits (Sauvé, 2016).

Figure 1.

Map of North America showing the (a) main air masses over the region (from Bryson, 1966; Bryson and Hare, 1974). (b) Location of the Baie and IDH peatlands and main records discussed in the text including (1) Baie and Lebel bogs (Magnan and Garneau, 2014), (2) Mort and Plaine bogs (Magnan and Garneau, 2014), (3) COR0602-42 marine core (Lemay-Tougas, 2014), and (4) Nordan’s Pond bog (Hughes et al., 2006).

The bog surface vegetation is dominated by Sphagnum fuscum, Sphagnum capillifolium, ericaceous shrubs (Chamaedaphne calyculata, Kalmia angustifolia, and Rhododendron groenlandicum), and sparse dwarf P. mariana on the hummock microforms, while the lawn microforms are mainly composed of Sphagnum rubellum, Vaccinium oxycoccos, Andromeda polifolia, and Eriophorum spp.

The Ile du Havre peatland (IDH) (50°13′N, 63°36′W, 0.5 km², 31 m a.s.l.) is a bog, raised at its center, located on a 12 km² island 2 km offshore in the eastern part of the Mingan archipelago in the Gulf of St. Lawrence (Figure 1). Peat accumulated on a depression over silty-clay sediments. At its deepest part, peat thickness reaches about 6.25 m, accumulated since ca. 7650 cal. BP. The underlying bedrock consists of Ordovician limestone intersected by layers of shale and sandstones from the Mingan formation (Grondin et al., 1980; Poole et al., 1970). The islands of the archipelago are characterized by cuesta landforms with cliffs to the north and gentle slopes toward the south (Grondin et al., 1980). The peatland being located at the northeast side of the island, the surrounding topography is characterized by cliffs and steep slopes. The coastline of the island (600 m away from the peatland) is dominated by pebbles and sandy beaches. A closed forest dominated by P. mariana occupies large parts of the island, while depressions are occupied by peatlands (Grondin et al., 1980).

The peatland is treeless and the dominant plant species are ericaceous shrubs (C. calyculata, K. angustifolia, and Empetrum nigrum) along with S. fuscum, Cladina spp., and isolated stands of dwarf P. mariana. The hummock microforms at the center of the peatland are dominated by a variable combination of S. fuscum, Cladina spp., E. nigrum, and Rubus chamaemorus, while hollows are dominated by S. rubellum and Carex limosa.

Coring and subsampling

A core was retrieved from the deepest part of each peatland, using a stainless steel box corer (10 × 10 cm width; Jeglum et al., 1992) to sample the upper meter and a Russian sampler (7.5 cm diameter; Jowsey, 1966) for the deeper sections. The cores were collected from two holes approximately 30 cm apart in an overlapping manner to ensure complete stratigraphic recovery. The cores were wrapped in plastic film, transferred into PVC tubes, and stored in a freezer (−20°C). In the laboratory, the cores were sliced frozen at 1-cm intervals (Givelet et al., 2004) using a stainless steel band saw at Ecolab (Toulouse, France). Each slice was rinsed with milliQ water and the edges of each subsample trimmed away to avoid any contamination from the saw. The thickness of each slice was then measured again to evaluate the loss from the slicing and correct the mid-point depth of each sample. The slices were then subsampled for each type of analyses.

Plant macrofossils

Plant macrofossil analyses were performed at systematic 4-cm intervals and refined (2-cm intervals), where visual changes in the stratigraphy were noted. Subsamples (4 cm³) were gently boiled in distilled water with addition of 5% KOH and carefully sieved (>125 µm). Macrofossils were scanned using a binocular microscope (×10–×40) and identified using reference collections (Garneau, 1995; Mauquoy and van Geel, 2007). Plant macrofossils were expressed as volume percentages (%) representative of cover estimates in a Petri dish (Sphagnum remains, ligneous fragments, etc.) except seeds, leaf fragments, and charcoal fragments, which were reported as numbers (n). Cenococcum spp. sclerotia and Myrica gale leaves/seeds were scaled from 1 to 5 (1 = rare, 5 = abundant). Sphagnum was determined to the highest taxonomic level possible based on stem and branch leaf characteristics using a photonic microscope (×40–×100). Terminology follows Marie-Victorin (1995) for vascular plants and Crum and Anderson (1979–1980) and Ireland (1982) for bryophytes. Zonation of the macrofossil diagrams was made using constrained cluster analysis by sum-of-square (CONISS) in psimpoll 4.25.

Chronological control

The chronological control of the cores is based on ¹⁴C AMS dates and ²¹⁰Pb measurements. For ¹⁴C dating, only aboveground plant macrofossils were selected based on a protocol developed by Mauquoy et al. (2004), in order to ensure a better accuracy of dates. Most of the macrofossils consisted of Sphagnum spp. stems (Table 1). When Sphagnum were not sufficient to provide enough material for dating, other types of macrofossils were selected (Larix laricina and P. mariana needles, Ericaceae leaves, etc.). A total of 24 samples (14 and 11 for Baie and IDH, respectively) were submitted to the Keck-CCAMS facility (Irvine, USA) for AMS radiocarbon dating.

Table 1.

Results of ¹⁴C AMS measurements, calibrations, and description of samples for the Baie and IDH cores.

Site and sample	Depth (cm)	Laboratory number	¹⁴C age (BP)	2σ range (cal. yr BP)	Material dated
Baie
45	51.6	UCIAMS-129510	425 ± 15	481–513	Charred Picea mariana needles
67	79.4	UCIAMS-151367	1015 ± 25	833–971	Sphagnum spp. stems
106	105.9	UCIAMS-129511	1120 ± 15	977–1059	Sphagnum spp. stems
128	130.0	UCIAMS-129512	1275 ± 15	1181–1271	Charred P. mariana needles
151	155.2	UCIAMS-135384	1905 ± 20	1818–1893	Sphagnum spp. stems
173	179.3	UCIAMS-129513	1740 ± 25	1570–1709	Sphagnum spp. stems, branches and leaves, charred P. mariana needles
196	206.7	UCIAMS-129514	2250 ± 15	2162–2336	Sphagnum spp. stems
217	233.0	UCIAMS-129515	2465 ± 15	2442–2704	Sphagnum spp. stems
232	250.2	UCIAMS-129516	2725 ± 15	2779–2853	Sphagnum spp. stems
256	283.2	UCIAMS-129517	2900 ± 15	2963–3136	Sphagnum spp. stems
286	321.8	UCIAMS-129518	3090 ± 15	3248–3362	Sphagnum spp. stems, Chamaedaphne calyculata leaf fragments
328	374.8	UCIAMS-129519	3590 ± 15	3842–3961	Sphagnum spp. stems, Ericaceae leaf fragments
376	435.3	UCIAMS-129520	3885 ± 15	4250–4409	Larix laricina needles
386	447.9	UCIAMS-135385	3995 ± 20	4421–4519	Bulk peat
IDH
54	69.7	UCIAMS-135378	140 ± 20	8–278	Sphagnum spp. stems
116	118.1	UCIAMS-135379	760 ± 20	670–726	Sphagnum spp. stems
137	143.6	UCIAMS-151363	905 ± 20	763–910	Sphagnum stems, branches and leaves
191	206.8	UCIAMS-151364	1900 ± 20	1815–1896	Sphagnum stems
220	240.7	UCIAMS-135380	2120 ± 20	2007–2150	Larix laricina needles, Myrica gale and C. calyculata leaf fragments, Sphagnum spp. stems
286	318.9	UCIAMS-135381	2785 ± 20	2804–2951	Sphagnum spp. stems
317	357.0	UCIAMS-151365	3320 ± 20	3479–3607	Sphagnum spp. stems
345	390.8	UCIAMS-151366	3480 ± 25	3651–3832	Sphagnum spp. stems
394	451.7	UCIAMS-135382	4380 ± 20	4857–5033	Sphagnum spp. stems
478	554.3	UCIAMS-135383	4960 ± 25	5612–5737	L. laricina needles and seeds, M. gale and C. calyculata leaf fragments
535	624.1	UCIAMS-123619	6950 ± 25	7699–7839	Carex spp. seeds

The ²¹⁰Pb activity was determined in the uppermost peat layers after extraction of its granddaughter product, ²¹⁰Po, from 0.5 g of powdered bulk peat spiked with a ²⁰⁹Po yield tracer following a sequential extraction HNO₃-HCl-HF-H₂O₂ digestion (Ali et al., 2008). Measurements were realized on an α-spectrometer (EGG Ortec 476A) at GEOTOP Research center (UQAM, Montreal). The Constant Rate of Supply (CRS) model was subsequently applied to build ²¹⁰Pb age-models (Appleby, 2002). Age–depth models were generated by combining both ¹⁴C and ²¹⁰Pb dates using the BACON software package (Blaauw and Christen, 2011). The radiocarbon dates were calibrated using the InterCal13 Northern Hemisphere terrestrial calibration curve (Reimer et al., 2013) integrated in the BACON package.

Chemical analyses

All geochemical sample preparations were performed under clean laboratory conditions (class 100) using acid-cleaned labware. Approximately 100 mg of powdered peat samples were acid-digested until complete dissolution, using a series of steps (HNO₃ + HF, then H₂O₂) in Savillex^® beakers on a hot plate (Le Roux and De Vleeschouwer, 2010). In a number of samples, an additional step involving aqua regia was necessary to fully digest the organic matter. The major elements (Al, Ti, Ca, Fe, K, and Na) analyses were made by inductively coupled plasma optical emission spectrometry (ICP-OES) at Ecolab laboratory (Toulouse, France) on a Thermo Electron IRIS Intrepid II. The trace element measurements (Sr, Rb, Zr, Sc, Pb, Th, U, and REEs) were completed on an Agilent Technologies 7500ce quadrupole inductively coupled plasma mass spectrometer (Q-ICP-MS) at Observatoire Midi-Pyrénée, in Toulouse. Prior to ICP-MS analyses, samples were diluted and an In-Re spike was added for internal normalization process.

The analytical performance was monitored through reference materials (NIST 1515 Apple leaves, NIST 1547 Peach leaves, NIST 1575a Pine needles, GBW07604 Bush branches and leaves, and NIMT/UOE/FM/001 peat). Measurements obtained by ICP-OES were generally within 10% of certified values with the exception of Al concentrations which vary between 6% and 17% of certified values. The elements measured by ICP-MS were all within 15% of certified values with the exception of Sc (22%), Lu (19%), and Ce (16%). The reproducibility of digestion procedure was evaluated by repeated analyses of NIST 1515 (n = 5), NIST 1547 (n = 5), GBW-07063 (n = 4), and 10 peat samples (each n = 2) and was generally better than 20%. The procedural blanks for the element analyzed were <15 ppt for trace elements and <1 ppm for Al, Ca, Ti, K, and Na.

Dust flux

The dust flux was calculated using the sum of REE concentrations (µg g⁻¹) in the bulk peat using the following equation (Shotyk et al., 2002):

\begin{array}{l} Dust flux (g m^{- 2} {yr}^{- 1}) = ({\sum [REE]}_{sample} / {\sum [REE]}_{UCC}) \\ * PAR * bulk density * 10,000 \end{array}

with Σ [REE]_sample the sum of REE concentrations (µg g⁻¹) in a sample, Σ [REE]_UCC the sum of REE concentrations in the upper continental crust (144.3 µg g⁻¹; Wedepohl, 1995), and PAR the peat accumulation rate (cm yr⁻¹).

Statistics

A principal component analysis (PCA) was performed on the elemental concentrations in each peat profile using SPSS 22.0.0 software. A PCA can help in the inference of the sources and processes affecting the distribution of the chemical elements. The PCA was performed in correlation mode on previously log transformed (log₁₀) and standardized (z-scores) data to avoid scaling effects, using a varimax rotation which is a fixed orthogonal rotation that maximizes the loadings of the variables on the components (Eriksson et al., 1999). The z-scores were calculated as (X_i − X_avg)/X_std, where X_i is the variable (i.e. concentration of the element), while X_avg and X_std are the series average and standard deviation of all samples for the variable X_i. Since we are working with chemical elements, the components will contain elements with similar records, likely controlled by the same environmental factors. Hence, using PCA, the chemical signals tend to be clearer and the underlying factors controlling them are more easily identified and interpreted.

Results

Chronologies

For both cores, the range of calibrated radiocarbon ages of the dated peat layers are presented in Table 1, while the supported and unsupported ²¹⁰Pb activities and CRS modeled ages are reported in Table 2. The age–depth models cover the period of 4500 cal. BP to present and 7650 cal. BP to present in the Baie and IDH cores, respectively (Figure 2). The Baie core reveals a rather constant peat accumulation rate, with a relatively high average value of 1.0 mm yr⁻¹ from the base of the core to 52 cm below the surface. An important diminution in accumulation rates is registered between 52 and 36 cm, yielding an average of 0.4 mm yr⁻¹. From 36 cm toward the surface, the accumulation rates are much higher and correspond to the acrotelm.

Table 2.

Results of ²¹⁰Pb measurements and CRS modeling on Baie and IDH cores.

Baie						IDH
Sample	Depth (cm)	²¹⁰Pb_tot activity (Bq kg⁻¹)	²¹⁰Pb_ex activity (Bq kg⁻¹)	CRS ²¹⁰Pb age (AD)	Uncertainty (yr)	Sample	Depth (cm)	²¹⁰Pb_tot activity (Bq kg⁻¹)	²¹⁰Pb_ex activity (Bq kg⁻¹)	CRS ²¹⁰Pb age (AD)	Uncertainty (years)
Baie 2	1.7	332	316	2010	1	IDH 1	0.8	449	436	2010	1
Baie 3	2.9	500	484	2008	1	IDH 2	2.4	480	468	2006	1
Baie 4	4.1	257	241	2006	1	IDH 3	3.6	436	423	2001	1
Baie 6	6.5	339	323	2003	1	IDH 5	6.0	388	375	1997	1
Baie 8	8.9	290	274	2002	1	IDH 7	8.5	373	360	1993	1
Baie 10	11	236	221	2001	1	IDH 9	10.9	420	408	1990	1
Baie 12	13.4	323	308	1997	1	IDH 11	13.3	227	215	1987	1
Baie 14	15.7	358	343	1991	1	IDH 13	15.8	334	321	1983	1
Baie 16	17.9	363	347	1987	1	IDH 15	18.4	262	249	1980	1
Baie 18	20.3	209	193	1984	2	IDH 17	21.1	261	249	1976	2
Baie 20	22.6	268	252	1979	2	IDH 19	23.8	193	180	1973	2
Baie 22	24.9	248	232	1974	2	IDH 21	26.4	215	202	1968	2
Baie 24	27.3	267	251	1966	2	IDH 23	29.1	267	255	1958	3
Baie 26	29.7	237	221	1953	3	IDH 25	32.0	175	162	1948	3
Baie 28	32.1	241	225	1921	4	IDH 27	34.7	116	104	1938	4
Baie 30	34.5	104	89	1869	8	IDH 29	37.5	62	50	1933	4
Baie 32	36.9	32	17			IDH 31	40.2	53	40	1928	5
Baie 34	39.1	16	0			IDH 33	42.9	53	40	1920	5
Baie 36	41.4	15	0			IDH 35	45.6	55	42	1909	6
						IDH 37	48.8	53	41	1891	6
						IDH 39	51.2	59	47	1865	7
						IDH 41	53.6	39	26
						IDH 43	58.6	14	0
						IDH 45	63.2	13	0

Figure 2.

Age–depth models for Baie (top) and IDH (bottom) cores constructed using BACON software (Blaauw and Christen, 2011) using both ²¹⁰Pb and ¹⁴C dates. The gray bands encompass all possible age–depth models, whereas the dotted lines represent the 95% confidence intervals. The red dotted line represents the weighted mean age of each modeled sample. Blue symbols indicate calendar age distribution of ¹⁴C dates and light green symbols show ²¹⁰Pb ages derived from the CRS model.

The IDH core displays a mean accumulation rate of 0.4 mm yr⁻¹ from peat initiation at 623 cm until 554 cm (Figure 2). Accumulation rates are higher between 554 and 452 cm with a mean of 1.1 mm yr⁻¹. Between 452 and 70 cm, the IDH core shows a relatively constant accumulation rate of 0.8 mm yr⁻¹. As in the Baie core, from 70 cm toward the surface, the accumulation rates are much higher and correspond to the acrotelm.

Plant macrofossils

The main changes in macrofossil composition are reported in Figure 3 and Tables 3 and 4. Following peat inception at ca. 4500 cal. BP, the Baie bog displays a high mineral content (ash content > 50%) and a dominance of Cyperaceae (Figure 3a; zone B-1). Zones B-2 (4470–4340 cal. BP) and B-3 (4340–3940 cal. BP) show high amounts of wet and mineral rich taxa such as L. laricina, M. gale (B-2), Cyperaceae, brown mosses, and Sphagnum section Cuspidata (B-3). Zone B-4 (3940–2270 cal. BP) is characterized by the replacement of S. section Cuspidata by Sphagnum section Sphagnum at the bottom followed by Sphagnum section Acutifolia, namely, S. rubellum/capillifolium to the top of the zone. Zone B-5 (2270–1240 cal. BP) is dominated by S. fuscum with intermittent layers of Ericaceae/woody debris and unidentified organic matter (UOM) with presence of charcoal layers (130 and 154.1 cm). In zone B-6 (1250–590 cal. BP), S. rubellum/capillifolium gradually replaces S. fuscum to dominate the vegetation assemblages. Similarly to zone B-5, zone B-7 (590 cal. BP-1950AD) is characterized by high amounts of S. fuscum interspersed with layers of Ericaceae debris and UOM. Finally, in zone B-8 (AD 1950 to present), a combination of S. fuscum and S. rubellum/capillifolium dominates again the vegetation composition.

Figure 3.

Macrofossil diagrams, ash content and bulk density for (a) Baie and (b) IDH cores. Peat type (%): Sphagnum (light gray), other mosses (white), ligneous (dark gray), Cyperaceae (horizontal lines), aquatics (gray/horizontal lines), unidentified organic matter (black).

Table 3.

Baie macrofossil zonations.

Macrofossil zones	Depth (cm)	Main characteristics
B-8	0–33	High presence of Sphagnum rubellum/capillifolium up to 60% with lower percentages of Sphagnum fuscum
B-7	33–65	Dominance of S. fuscum and diminution of S. rubellum/capillifolium presence (down to 20%), high presence of Cenococcum spp. sclerotia, presence of charcoal at the transition with zone B-8, recurrent peaks in UOM and Ericaceae rootlets
B-6	66–130	Co-dominance of S. fuscum and S. rubellum/capillifolium, presence of Picea mariana needles
B-5	130–209	Dominance of S. fuscum and disappearance of S. rubellum/capillifolium, occurrence of several peaks in UOM (up to 50%) and Ericaceae rootlets, presence of two charcoal layers at 130 and 154.1 cm, high abundance of Cenococcum spp. sclerotia
B-4	209–382	High presence, reaching up to 84% of S. rubellum/capillifolium; presence of Sphagnum section Sphagnum peaking between 295 and 302 cm
B-3	382–433	High abundance of Cyperaceae and Calliergon stramineum, numerous Larix laricina needles at the bottom of the zone, high percentage of Sphagnum section Cuspidata
B-2	448–433	High number of Myrica gale leaves and L. laricina needles, up to 50% of Ericaceae rootlets
B-1	448–453	Presence of Foraminifera tests and high abundance of Cyperaceae remains (more than 80%)

Table 4.

IDH macrofossil zonations.

Macrofossil zones	Depth (cm)	Main characteristics
IDH-7	0–229	Abundant Sphagnum fuscum and S. rubellum/capillifolium together with numerous Ericaceae rootlets
IDH-6	229–267	Dominance of S. rubellum/capillifolium, peak in Larix laricina needles and presence of Myrica gale leaves at 230–250 cm
IDH-5	267–321	Dominance of S. fuscum and S. rubellum/capillifolium, high proportion of ericaceous rootlets, presence of Picea mariana needles
IDH-4	321–445	Sphagnum peat dominated by S. fuscum and S. rubellum/capillifolium
IDH-3	445–478	Peak in Sphagnum teres and Dicranum spp. at 450–486 cm, disappearance of brown mosses other than Dicranum spp., presence of Menyanthes trifoliata and M. gale, important diminution in Cyperaceae
IDH-2	478–607	Dominance of brown mosses Calliergon giganteum, Scorpidium scorpioides, and Warnstorfia spp.; abundant L. laricina needles and M. gale seeds and leaves; presence of M. trifoliata seeds
IDH-1	607–625	High abundance of Cyperaceae and Bryum spp. at the base of the section

IDH initiated at ca. 7650 cal. BP over silty-clay sediments. The first three zones, namely, IDH-1 (7650–7175 cal. BP), IDH-2 (7175–5150 cal. BP), and IDH-3 (5150–4750 cal. BP), are characterized by minerotrophic taxa dominated by Cyperaceae, Menyanthes trifoliata, M. gale; brown mosses Cinclidium stygium, Calliergon giganteum, Scorpidium scorpioides, and Warnstorfia spp.; and other bryophytes such as Sphagnum teres and S. section Cuspidata (Figure 3b and Table 4). Zone IDH-4 (4750 to 2910 cal. BP) is characterized by a dominance of S. fuscum and S. rubellum/capillifolium with high amount of Ericaceae rootlets and wood fragments. In zone IDH-5 (2910–2365 cal. BP), S. rubellum/capillifolium is the most abundant species in the assemblages reaching up to 75%. Zone IDH-6 (2365–2000 cal. BP) is still dominated by S. rubellum/capillifolium but a short period between 2200 and 2000 cal. BP is characterized by abundant L. laricina needles and M. gale leaves. In zone IDH-7 (last 2000 years), S. fuscum is the dominant species with relatively high amount of S. rubellum/capillifolium.

PCAs on elemental geochemistry

Three principal components (PCs) were extracted in the Baie record, accounting for 92% of the total variance (Figure 4a). The first PC explains 63% of the variance and is characterized by positive loadings of U, Th, REEs, Zr, Sc, Ti, and Al. The factor score profile displays several peaks, mainly between 200–100 cm (1760–1030 cal. BP) and 60–25 cm (650 cal. BP–AD 1955). PC2 accounts for 20% of the variance and has positive loadings for K, Rb, Na, and Pb. The factor scores of PC2 show relatively constant values up to 40 cm, where they increase toward the surface. Rb, Na, and K concentrations display low values except at 29.7 cm and near the surface (Figure 4). The PC3 explains 9% of the variance, and Ca and Sr show high positive loading factors. The factor scores of PC3 show a decreasing trend from the base of the profile until 200 cm, where they increase and remain relatively constant toward the surface.

Figure 4.

Communalities accounting for each element’s variance allocated to each principal component, log₁₀ of Ca, Na, K, La, and Pb concentrations as well as factor score profiles for each PC for (a) Baie and (b) IDH cores.

In the IDH profile, two PCs were extracted, accounting for 90% of the variance (Figure 4b). The first PC explains 70% of the variance and shows high loading factors for all REEs, U, Zr, Al, Sc, Th, and Ti. The factor score profile shows a ‘see-saw’ pattern with several peaks at 400–378 cm (3910 to 3700 cal. BP), 335 cm (3130 cal. BP), 206 cm (1810 cal. BP), 143 cm (875 cal. BP), and 40–30 cm (AD 1955 to AD 1925). The second PC explains 20% of the variability and includes Pb, K, Na, Rb, and high negative loadings with Sr and Ca. The factor scores for PC2 display a large peak between 70 and 20 cm.

Dust flux

The sum of REEs was selected to reconstruct dust fluxes as REEs show very high loading factors in both bog’s PC1. With the presence of an aluminum smelter in the region of the Baie bog and a Ti mine close to the IDH bog, these two elements were not used to reconstruct dust fluxes. Zirconium is known to be enriched in certain mineral phases, mostly zircons (Shotyk et al., 2002), and therefore was not used as well.

If we exclude the basal minerotrophic phase, the dust flux of the Baie profile can mainly be divided into four phases. A first phase from ca. 3800 to 1760 cal. BP (Figure 5a) shows a relative stability with fluxes ranging from 0.5 to 1.1 g m⁻² yr⁻¹ (mean: 0.81 ± 0.28 g m⁻² yr⁻¹). This period is followed by a period of increased dust fluxes ranging from 0.68 to 3.81 g m⁻² yr⁻¹, with greater variability (mean: 1.83 ± 0.84 g m⁻² yr⁻¹), between ca. 1760 and 1000 cal. BP. From ca. 1000 to 650 cal. BP, dust fluxes reach their lowest values in the core (mean: 0.24 ± 0.09 g m⁻² yr⁻¹) followed by a period of high variability with higher fluxes averaging 1.31 ± 0.31 g m⁻² yr⁻¹, which lasted until AD 1960.

Figure 5.

(a) Atmospheric dust fluxes (g m⁻² yr⁻¹) in Baie and IDH, (b) percentage of Sphagnum and Ericaceae in Baie (this study), (c) percentage of Sphagnum and Ericaceae in IDH (this study), (d) mean water table depth (WTD) for every 200 years from two peatlands in Baie-Comeau (Baie and Lebel; B-C) region (Magnan and Garneau, 2014), (e) mean WTD for every 200 years from two peatlands in Havre St-Pierre (Mort and Plaine; HSP) region (Magnan and Garneau, 2014), (f) bog surface wetness (BSW) from Newfoundland (Hughes et al., 2006), (g) sea-surface temperature (SST) from COR0602-42 (Lemay-Tougas, 2014), and (h) Δ¹⁴C curve (Reimer et al., 2013).

The dust flux of the IDH profile shows less variability and a lower rate of particle accumulation (Figure 5a). The profile shows an average dust flux of 0.4 g m⁻² yr⁻¹ with four peaks of interest. A peak reaching 2.1 g m⁻² yr⁻¹ is observed between ca. 4100 and 3700 cal. BP. A second peak reaching 0.9 g m⁻² yr⁻¹ is registered at ca. 3130 cal. BP. Then, the dust flux shows little variability until ca. 875 cal. BP where a third peak is observed (1.5 g m⁻² yr⁻¹). After this, the dust flux remains relatively low and stable until AD 1910 where it increases again up to 2.7 g m⁻² yr⁻¹ and higher fluxes (mean: 1.43 ± 0.85 g m⁻² yr⁻¹) persist until the present day.

Discussion

Trophic status of the peatlands

The ombrotrophic conditions of the Baie and IDH cores are confirmed by their ecological and geochemical characteristics. In the Baie bog, Cyperaceae, brown mosses, and S. section Cuspidata, that is, species indicative of minerotrophic conditions, dominate the assemblages from the base until 3800 cal. BP (Figure 3a; zones 1–3; Table 3). Plant macrofossil record (Figure 3a) suggests a rapid transition from fen to bog with a persistence of ombrotrophy from 3800 cal. BP to the present day as also reported by Magnan et al. (2014). The ombrotrophy is further confirmed by the low ash content (<5%) (Tolonen, 1984) in the core above 440 cm (Figure 3a).

In the IDH core, the first three zones in Figure 3b are composed of species typical of minerotrophic conditions, that is, dominance of Cyperaceae, brown mosses, and Sphagnum teres (Figure 3 and Table 4). Macrofossils and major elements suggest that ombrotrophy was reached at about 445 cm of depth (Figures 3b and 4b). The appearance of S. fuscum and S. rubellum/capillifolium at the end of zone IDH-3 (5000 cal. BP) indicates a transition from a fen to a bog environment (Figure 3b and Table 4). The low Sr (<30 ppm) and Ca values (<0.5%) (underlying limestone substratum) are also consistent with ombrotrophic conditions (Shotyk et al., 2001). However, the low concentrations of lithogenic elements further down core suggest that their supply was mainly atmospheric approximately 500 years earlier (Figure 4b), which is confirmed by the low ash content (Figure 3b).

Paleoenvironmental/paleoclimatological changes from macrofossils

In the Baie macrofossil record, the period between 3800 and 2100 cal. BP shows stability with assemblages typical of a lawn microform (S. rubellum/capillifolium and S. section Sphagnum with Ericaceae rootlets). Between 2100–1250 cal. BP (zone B-5) and 500 cal. BP to AD 1950 (zone B-7), a lower water table probably linked to drier climatic conditions as well as a greater variability in the hydroclimatic conditions are suggested by the changes in vegetation assemblages (S. fuscum, several peaks of UOM and Ericaceae and two charcoal layers). The dominance of S. section Acutifolia almost throughout the core suggests relatively dry conditions over the studied period with substitutions by Ericaceae and UOM indicating even drier periods.

In the IDH ombrotrophic section, zone IDH-4 (4750–2900 cal. BP) is characterized by the dominance of S. fuscum and S. rubellum/capillifolium and the presence of Ericaceae rootlets, which indicates drier conditions than in the previous sections. In zones 5 and 6 (2910–2000 cal. BP), S. rubellum/capillifolium dominates the assemblages, which suggests relatively more humid conditions similar to a lawn microform. A short period between 2200 and 2000 cal. BP is characterized by abundant L. laricina needles and M. gale leaves, which can be explained by more humid conditions or a return to minerotrophic conditions (IDH-6). Over the last 2000 years (IDH-7), little variability is observed in the assemblages in which S. fuscum is the dominant species with relatively high amount of S. rubellum/capillifolium. The high presence of S. section Acutifolia and Ericaceae remains in the ombrotrophic section suggests relatively dry conditions over the entire period.

Drivers of the geochemical signals

PCAs were realized in order to identify variables (i.e. chemical elements) with similar behavior and likely to be controlled by the same processes. The PCA analyses focused on the last 4300 years in the Baie core and 5500 years in the IDH cores. PC1 displays high positive loading factors with all lithogenic elements (U, Th, REEs, Zr, Sc, Ti, and Al) (Figure 4). The fact that REEs load on the same PC as other lithogenic elements, known for their conservative nature, such as Ti and Sc, confirms that REEs are immobile in the peat column as suggested in Krachler et al. (2003) and Aubert et al. (2006). Since all lithogenic elements are associated with PC1 in both cores, it reflects the mineral content of the peat from dust deposition, probably derived from soil erosion. In both cores, PC2 shows high loading factors for Na, Rb, K, Pb (Figure 4) and represents a combination of different controlling factors of the geochemical profiles. Lead has a history of anthropogenic use and is known to be emitted as by-product of human activities. It shows an increase from the beginning of the 19th century and a peak in the 1950s–1970s (Gallon et al., 2005; Pratte et al., 2013). Elements such as Rb, K, and Na are all essential elements known to be bioaccumulated by bog vegetation, which would explain the increase in the factor score at the surface (Damman, 1978; Steinmann and Shotyk, 1997). The presence of a sand layer at 29.7 cm depth in the Baie core likely explains the enrichment in Rb, K, Na but also in other lithogenic elements. Furthermore, the low values, except in the above-mentioned layers, probably explain the fact that these elements are on the same component as Pb although they normally do not display similar geochemical behavior in the peat column. Calcium and strontium show high loading factors on the PC3 in the Baie core, while they show high negative loadings with PC2 in the IDH core (Figure 4). Diagenetic processes, such as changes in redox conditions and upward leaching from basal sediments following mineral dissolution, are known to affect the mobility of Ca and Sr in peat (Steinmann and Shotyk, 1997). Both cores show a gradual decrease in Ca and Sr concentrations (less negative values in the case of the IDH core) over time up to the fen to bog transition, which suggest upward leaching from the basal sediments (Shotyk et al., 2001). Hence, PC3 in Baie bog and the Ca and Sr part of PC2 in the IDH core will not be discussed further as they do not provide information about dust deposition or origin.

Paleoclimatic events recorded by the dust records

Period 4100–3700 cal. BP

Between 4100 and 3700 cal. BP, the IDH core displays increased dust fluxes (mean: 0.95 ± 0.76), greater factor scores for PC1 (Figure 4b), and a greater proportion of Ericaceae remains, suggesting slightly drier conditions (Figure 5a and c). Although the Baie record covers this time period, it will not be discussed because the dust signal was not solely atmospheric in origin which renders its interpretation difficult in terms of past climate. Several archives recorded a dry episode at low and mid-latitudes around 4200 cal. BP (Booth et al., 2005 and references therein), while the higher latitudes recorded more humid conditions (Yu et al., 2003). Mayewski et al. (2004) report a strengthening of the westerlies over North America from 4200 to 3800 cal. BP, which is in agreement with studies conducted in eastern North America (Almquist et al., 2001; Jessen et al., 2011). Hence, the combination of drier conditions to the south and increased westerly strength could explain the increased dust deposition in the IDH bog over the period.

Period 3700–1750 cal. BP

The Baie and IDH records show relative stability in the dust flux at ca. 3800–2000 cal. BP (mean: 0.81 ± 0.28 g m⁻² yr⁻¹) and ca. 3700–2000 cal. BP (mean: 0.45 ± 0.17 g m⁻² yr⁻¹), respectively, although some slight increases are observed in both records (Figure 5a). The macrofossils composition also shows stability over this period with the dominance of S. rubellum/capillifolium (Figures 3 and 5b and c). Other regional studies have shown rather stable water table depths (Figure 5d), carbon accumulation rates, and summer sea-surface temperature (Figure 5; SST) during this period (Lemay-Tougas, 2014; Magnan et al., 2014; Magnan and Garneau, 2014). At IDH, a trend toward lower dust deposition is observed from ca. 3600 to 3000 cal. BP and, for the same period, higher WTD was reported in a peat bog along the shore of the Gulf of St. Lawrence (Magnan and Garneau, 2014). In the same record, a gradual drying of the peat surface is documented after 3000 cal. BP reaching its maximum around 1900 cal. BP. Over the same period (3000–1900 cal. BP), both dust records do not display significant changes. The macrofossil analyses as well as the low dust fluxes recorded in both Baie and IDH bogs during this period suggest a rather stable wind and climatic regime, which is supported by other records (Figure 5d, f, and g), although a number of regional records (Figure 5e) suggest a dryer climate in the region of the IDH bog, not recorded by our proxies.

The apparent discrepancies between our dust records and the biological proxies could be explained by their different response time to changes. Furthermore, these proxies can be controlled by internal processes during certain periods (Sullivan and Booth, 2011). In contrast to the biological proxies, dust deposition is not affected by internal processes, hence regardless of the geographical scale, it is likely more strongly linked to climate.

Period 1750–1000 cal. BP

The highest dust fluxes were recorded during the period from 1750 to 1000 cal. BP, with values ranging between 1.2 and 3.8 g m⁻² yr⁻¹ (mean: 1.83 ± 0.84 g m⁻² yr⁻¹) (Figure 5a), while macrofossils show several changes in species composition (Figures 3a and 5b), with several peaks in Ericaceae remains and charcoal horizons. The IDH bog recorded a lower dust flux (mean: 0.39 ± 0.16 g m⁻² yr⁻¹) over the same time period. Although a relatively warmer climate is suggested from Δ¹⁴C production (Figure 5h), the different peaks recorded during this period correspond with periods of higher ¹⁴C production, that is, lower insolation (Bond et al., 2001; Chambers et al., 1999). The highest dust flux values in the Baie bog are recorded when summer SST was at its warmest at ca. 1500 cal. BP as indicated by a marine core in the St. Lawrence Estuary (Figure 5g; Lemay-Tougas, 2014). Afterward, a decrease in summer SST and sea-surface salinity (greater precipitation) centered at ca. 1200 cal. BP is reported (Lemay-Tougas, 2014).

All the records in Figure 5 agree that a trend toward more humid conditions occurred around ca. 1750 cal. BP. Furthermore, both the Newfoundland (BSW) and the Δ¹⁴C records show greater variability from 1750 cal. BP (Figure 5f and h). The same variability is observed for the dust flux and the macrofossil assemblages in the Baie record. Testate amoebae assemblages from records in the Baie-Comeau region also display species associated with periods of high decadal to centennial hydroclimate variations (Magnan and Garneau, 2014). This period is therefore characterized by a certain degree of hydroclimatic instability expressed by high variability of vegetation and testate amoebae assemblages and higher dust loads as in Europe. For example, De Jong et al. (2007) noted peaks in eolian activity during hydrological shifts regardless of their direction (wet or dry) in South Sweden. In Belgium, De Vleeschouwer et al. (2012) also suggested that higher dust loads were favored by hydroclimatic instabilities (i.e. shift toward wet or dry periods). In our records, this hypothesis is supported, especially in the Baie bog, which displays higher dust fluxes during a period where other biological proxies show high variability. Such instability could be explained by a change in atmospheric circulation. Around 1650 ± 200 cal. BP, Viau et al. (2002) identified a major transition in pollen records of North America (including our study regions) associated with a rearrangement of atmospheric circulation. Furthermore, cooling events during the late-Holocene in eastern Canada were associated with the incursion of Arctic air masses (Carcaillet and Richard, 2000; Girardin et al., 2004), which could partially explain the occurrence of dust peaks with cold events and increased hydroclimatic instability.

‘MWP’

In both study regions, the ‘MWP’ (roughly 1000–650 cal. BP) is characterized by higher WTD around 700 cal. BP in peat bogs (Figure 5d and f). Further east, Hughes et al. (2006) also reported a wet shift between 900 and 750 cal. BP. Arseneault and Payette (1997) reported greater tree growth associated to warmer conditions (1000–800 cal. BP) in northern Quebec, while Lemay-Tougas (2014) observed slightly cooler and wetter conditions from a record of summer SST (1200–800 cal. BP Figure 5g) and sea-surface salinity. Following this period, the summer SST record is characterized by a transition toward a warmer/drier climate. The discrepancies in the age of the different events in the marine record could be ascribed to the lower dating control compared with our record or to slower response of the oceanic system relative to changes. In both Baie and IDH records, a diminution of Ericaceae remains and increase in S. rubellum/capillifolium suggest a slightly wetter climate. This period is also recorded in the dust fluxes and PC1 factor scores (Figure 4) from Baie (mean: 0.24 ± 0.09 g m⁻² yr⁻¹) and IDH (mean: 0.60 ± 0.52 g m⁻² yr⁻¹), which display lower values, usually typical for a wet period. This has already been observed by Filion (1984), who reported lower eolian activities in parallel with warmer and/or moister conditions during the same period in Northern Quebec. The 1000–650 cal. BP period is therefore characterized by warmer but also wetter conditions.

‘LIA’

Following the ‘MWP’, the ‘LIA’ is characterized by a period of increased dust fluxes in the Baie bog between 650 and 100 cal. BP (mean: 1.31 ± 0.31 g m⁻² yr⁻¹), where values reach 2.3 and 1.7 g m⁻² yr⁻¹ at 540 and 150 cal. BP, respectively, while IDH bog shows small peaks around 600–400 cal. BP and 200 cal. BP (mean: 0.36 ± 0.16 g m⁻² yr⁻¹). Factor scores of PC1 also display higher values during this period in the Baie core but do not show any change in the IDH core (Figure 4). Macrofossil assemblages in both cores suggest dry conditions and display high variability in the Baie core (Figure 3), similar to the previous period of increased dust deposition (1750–1000 cal. BP). Previous research in northern Quebec showed a dry shift between 530 and 350 cal. BP (Loisel and Garneau, 2010), while cooling associated with wet shifts were reconstructed in a peatland in Newfoundland around 600 and 200 cal. BP (Figure 5; Hughes et al., 2006). Moreover, Magnan and Garneau (2014) observed differences between the Baie-Comeau and Havre St-Pierre regions, the latter being drier during the ‘LIA’ (Figure 5) influenced by longer frost duration periods linked to a greater Sub-Arctic influence. The timing of the dust peaks during the ‘LIA’ in the Baie record also corresponds with minima in solar activity (Δ¹⁴C). The first peak (650–400 cal. BP) encompasses both Spörer and Maunder minima, while the second peak (150–100 cal. BP) corresponds with the Dalton minimum. The ‘LIA’ is therefore characterized by colder conditions, but the wetness probably varied from one region to another, with some places being drier than others in relation to longer frost duration under the influence of arctic air masses.

A more meridional atmospheric circulation or a stronger north polar vortex has been proposed to explain greater dust deposition and colder temperatures in Greenland (GISP2 core) during the ‘LIA’ (O’Brien et al., 1995). Likewise to the period between 1750 and 1000 cal. BP, such a change in atmospheric zonal and meridional circulation may as well have affected mid- and low-latitude circulation, which could explain the increase in dust deposition in our bogs during the ‘LIA’. A more meridional circulation could have allowed more frequent intrusion of cold and dry Arctic air masses. This is further confirmed by a transition in North American pollen records during the ‘LIA’ which would have been provoked by a reorganization of atmospheric circulation potentially linked to a solar forcing (Viau et al., 2002). The relative agreement between increases in dust deposition in the Baie record and solar minima supports this hypothesis.

Modern period (AD 1850 to present)

Both peat bogs recorded peaks in dust flux (mean: 0.98 ± 0.42 g m⁻² yr⁻¹ and 1.43 ± 0.85 g m⁻² yr⁻¹ for Baie and IDH cores respectively) following AD 1850 (100 cal. BP), which are likely anthropogenic in origin. More precisely, a large peak is found in both cores following the 1930s. The North shore was increasingly developed during this period with the development of several towns, among them Baie-Comeau, located 20 km northeast from Baie bog, and Havre St-Pierre, 2 km north of IDH bog on the mainland shore. The land clearance probably increased the amount of dust available for transport and can explain the high fluxes over this period. The construction of hydropower dams (1960–1985AD), and an aluminum smelter, in the region of Baie bog and the development of mining activities, more precisely a Ti mine (1950AD), near IDH bog have most likely contributed to the dust signal recorded in the two regions. The so-called ‘dust bowl,’ a period of severe drought conditions during the 1930s in the Central United States, was proposed as a potential contributor to dust deposited in Greenland ice over this period of time (Donarummo et al., 2003). However, isotopic analyses (Nd and Pb isotopes) of the mineral fraction in the peat would be needed to verify the presence of dust from Central United States in our peat bogs over this period. The factor scores of PC2 for both cores, which includes Pb, show a sharp increase from 100 cal. BP until the 1970s (Figure 4). The timing of this increase in Pb concentrations is in agreement with the historical use of leaded gasoline (Gallon et al., 2005; Pratte et al., 2013) and further confirms that both study regions were impacted by anthropogenic activities over the last century.

Insights from the comparison of the two dust records

The two studied peat bogs show differences in the magnitude of the dust signal. A potential explanation would be that both archives record a more regional scale signal or are affected by different air masses since the two sites are found in a zone of mixing air masses (Figure 1). The two study regions are located more than 500 km apart and likely developed in different climatic contexts. Distinct regional signals were already reported by Magnan and Garneau (2014) using testate amoebae, with drier conditions and greater variability on peat bogs near IDH, which are located in the forest tundra ecozone, more exposed and influenced by the Labrador Current (Figure 5).

The discrepancies between the two records could also be, at least partially, ascribed to within-bog spatial heterogeneity. The use of a single core per site could limit the representativeness of the records for the whole bogs as vegetation and micro-topography affect the capture and retention efficiency of atmospheric particles (Bindler et al., 2004). The difference between hummocks and hollows is well established, where hummocks are known to have higher interception of atmospheric particles (Norton et al., 1998; Oldfield et al., 1995). However, both records were collected onto low hummock microforms with similar plant composition (i.e. S. fuscum and S. rubellum/capillifolium dominated), which reduces the potential for the dust flux to be controlled by spatial heterogeneity. Furthermore, both records display similar ²¹⁰Pb inventories of 0.32 and 0.35 kBq m⁻² for Baie and IDH, respectively, which is similar to the inventory of 0.36 kBq m⁻² in another record from the Baie bog (Pratte et al., 2013). These similar ²¹⁰Pb inventories are as expected (Le Roux and Marshall, 2011) given the similar average annual rainfall in both the areas (1000 mm; Environment Canada, 2010). This suggests that particle deposition has not been overly affected by the micro-topography or vegetation in this study.

Although previous studies suggest that the Havre St-Pierre region was exposed to a drier and colder climate, the IDH bog did not record any significant increase in dust deposition during these periods, which could be explained by the availability of dust particles. While the peatlands along the coast of the Gulf of St. Lawrence are largely exposed to wind and long freezing seasons (Magnan and Garneau, 2014), the IDH peatland may have been protected by the tree fringe that surrounds its open center, which may have acted as a barrier for dust accumulation. Furthermore, the fact that the peatland is located on an island about 2 km offshore probably contributed to the lower dust fluxes. For changes in climatic/wind regime to be recorded in peat bogs, specific conditions need to be met. For example, changes in wind regime and climatic conditions were recorded as a result of the presence of sand dunes and/or seashore providing erodible material to certain peat bogs of northern Europe (De Jong et al., 2007; De Vleeschouwer et al., 2009). The same conditions can be applied to Baie bog, which is located near the seashore. Although IDH bog meets similar requirements, it does not record the variations in dust deposition. This discrepancy could be ascribed to the afore-mentioned ‘isolation’ of the peatlands, which would prevent the deposition of a part of the dust.

To summarize, results from the two dust records and comparison with other regional records highlight the importance of local and regional factors on late-Holocene variations in climate and dust deposition in eastern Canada.

Conclusion

Using peat cores recovered from two peat bogs in the Estuary (Baie) and Gulf (IDH) of the St. Lawrence in eastern Canada, 4000 years of paleodust deposition was reconstructed using REE concentrations. Plant macrofossils and other records of past surface wetness (testate amoeba) from the literature were used to reconstruct hydroclimatic variations over the same period. PCAs showed that REEs display the same behavior as other conservative lithogenic elements in the peat column (Ti, Al, Zr), hence they can be used to reconstruct mineral dust deposition in peat. Both cores show periods of increased dust deposition. The Baie bog shows the greatest variability with peaks between 1750–1000 cal. BP and 600–100 cal. BP. During both periods, comparison with macrofossils assemblages and other surface humidity records suggests periods of climatic instability (wet or dry shift) which affected dust deposition. A reorganization of atmospheric circulation during these periods is proposed to explain this instability. The agreement between the dust deposition record in the Baie bog and past solar minima suggests that dust deposition increased during cold periods. Over the last 100 years, dust deposition in both peat cores has been affected by anthropogenic activities which increased the atmospheric dust loads through different processes (land clearing, mining, smelting) in both regions.

Footnotes

Acknowledgements

We are grateful to Gaël Le Roux (Ecolab, Toulouse), David Baqué (Ecolab, Toulouse), Aurélie Lanzanova (Geoscience Environnement Toulouse), and Bassam Ghaleb (GEOTOP, Montreal, Canada) for their help with major and trace elements analyses and ²¹⁰Pb dating. Thanks to Gabriel Magnan, Nicole Sanderson, Antoine Thibault, and Hans Asnong for help during fieldwork as well as Julien Gogot, Julien Baudet-Lancup, and Marie-Josée Tavella for lab assistance. Thanks to Les Tourbeux for fruitful discussions.

Funding

Financial support was provided by Natural Sciences and Engineering Research Council of Canada (NSERC; #250287) discovery grant to MG. Scholarships to SP were provided by the Fonds de Recherche Québec – Nature et Technologie (FRQNT; #176250 and #180723). An additional mobility grant was provided by Institut National Polytechnique de Toulouse (‘Soutien à la mobilité’) grant to FDV.

References

Albani

Delmonte

Maggi

. (2012) Interpreting last glacial to Holocene dust changes at Talos Dome (East Antarctica): Implications for atmospheric variations from regional to hemispheric scales. Climate of the past 8: 741–750.

Albani

Mahowald

Winckler

. (2015) Twelve thousand years of dust: The Holocene global dust cycle constrained by natural archives. Climate of the past 11: 869–903.

Ali

Ghaleb

Garneau

. (2008) Recent peat accumulation rates in minerotrophic peatlands of the Bay James region, Eastern Canada, inferred by 210Pb and 137Cs radiometric techniques. Applied Radiation and Isotopes 66: 1350–1358.

Almquist

Dieffenbacher-Krall

Flanagan-Brown

. (2001) The Holocene record of lake levels of Mansell Pond, central Maine, USA. The Holocene 11: 189–201.

Appleby

(2002) Chronostratigraphic techniques in recent sediments. In: Last

Smol

(eds) Tracking Environmental Change Using Lake Sediments. Dordrecht: Springer, pp. 171–203.

Arseneault

Payette

(1997) Reconstruction of millennial forest dynamics from tree remains in a subarctic tree line peatland. Ecology 78: 1873–1883.

Aubert

Le Roux

Krachler

. (2006) Origin and fluxes of atmospheric REE entering an ombrotrophic peat bog in Black Forest (SW Germany): Evidence from snow, lichens and mosses. Geochimica et Cosmochimica Acta 70: 2815–2826.

Barber

Langdon

(2007) What drives the peat-based palaeoclimate record? A critical test using multi-proxy climate records from northern Britain. Quaternary Science Reviews 26: 3318–3327.

Bernatchez

(2003) Évolution littorale holocène et actuelle des complexes deltaïque de Betsiamites et Manicouagan-Outardes: synthèse, processus, causes et perspectives. Québec: Department of Geography, Université Laval.

10.

Bettis

EAI

Muhs

Roberts

. (2003) Last Glacial loess in the conterminous USA. Quaternary Science Reviews 22: 1907–1946.

11.

Bindler

Klarqvist

Klaminder

. (2004) Does within-bog spatial variability of mercury and lead constrain reconstructions of absolute deposition rates from single peat cores? The example of Store Moss, Sweden. Global Biogeochemical Cycles 18: GB3020.

12.

Blaauw

Christen

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

13.

Bond

Kromer

Beer

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

14.

Booth

Jackson

Forman

. (2005) A severe centennial-scale drought in midcontinental North America 4200 years ago and apparent global linkage. The Holocene 15: 321–328.

15.

Bryson

(1966) Air masses, streamlines and the boreal forest. Geographical Bulletin 8: 228–266.

16.

Bryson

Hare

(1974) Climates of North America. In: Bryson

Hare

(eds) Climates of North America: World Survey of Climatology, Volume 11. New York: Elsevier, pp. 1–47.

17.

Carcaillet

Richard

PJH

(2000) Holocene changes in seasonal precipitation highlighted by fire incidence in eastern Canada. Climate Dynamics 16: 549–559.

18.

Caseldine

Baker

Charman

. (2000) A comparative study of optical properties of NaOH peat extracts: Implications for humification studies. The Holocene 10: 649–658.

19.

Chambers

Charman

(2004) Holocene environmental change: Contributions from the peatland archive. The Holocene 14: 1–6.

20.

Chambers

Ogle

Blackford

(1999) Palaeoenvironmental evidence for solar forcing of Holocene climate: Linkages to solar science. Progress in Physical Geography 23: 181–204.

21.

Charman

Amesbury

Hinchcliffe

. (2015) Drivers of Holocene peatland carbon accumulation across a climate gradient in northeastern North America. Quaternary Science Reviews 121: 110–119.

22.

Crum

Anderson

(1979–1980) Mosses of Eastern North America. New York: Columbia University Press.

23.

Damman

AWH

(1978) Distribution and movement of elements in ombrotrophic peat bogs. Oikos 30: 480–495.

24.

De Jong

Schoning

Björck

(2007) Increased aeolian activity during humidity shifts as recorded in a raised bog in south-west Sweden during the past 1700 years. Climate of the past 3: 411–422.

25.

De Jong

Blaauw

Chambers

. (2010) Climate and peatlands. In: Dodson

(ed.) Changing Climates, Earth Systems and Society. Dordrecht: Springer, pp. 85–121.

26.

De Vleeschouwer

Ferrat

McGowan

. (2014) Extracting paleodust information from peat geochemistry. Past Global Changes Magazine 22: 88–89.

27.

De Vleeschouwer

Pazdur

Luthers

. (2012) A millennial record of environmental change in peat deposits from the Misten bog (East Belgium). Quaternary International 268: 44–57.

28.

De Vleeschouwer

Piotrowska

Sikorski

. (2009) Multiproxy evidence of ‘Little Ice Age’ palaeoenvironmental changes in a peat bog from northern Poland. The Holocene 19: 625–637.

29.

Delmonte

Andersson

Hansson

. (2008) Aeolian dust in East Antarctica (EPICA-Dome C and Vostok): Provenance during glacial ages over the last 800 kyr. Geophysical Research Letters 35: L07703.

30.

DeMenocal

Ortiz

Guilderson

. (2000) Abrupt onset and termination of the African Humid Period: Rapid climate responses to gradual insolation forcing. Quaternary Science Reviews 19: 347–361.

31.

Donarummo

Ram

Stoermer

(2003) Possible deposit of soil dust from the 1930’s U.S. dust bowl identified in Greenland ice. Geophysical Research Letters 30: 1269.

32.

Environment Canada (2010) Canadian Climate Normals 1981–2010. Available at: http://climate.weather.gc.ca/climate_normals/ (accessed 18 January 2016).

33.

Eriksson

Johansson

Kettaneh-Wodl

. (1999) Introduction to Multi and Mega-Variate Data Analysis Using Projection Methods (PCA & PLS). Umeå: Umetrics AB.

34.

Filion

(1984) A relationship between dunes, fire and climate recorded in the Holocene deposits of Quebec. Nature 309: 543–546.

35.

Fischer

Siggaard-Andersen

M-L

Ruth

. (2007) Glacial/interglacial changes in mineral dust and sea-salt records in polar ice cores: Sources, transport, and deposition. Reviews of Geophysics 45: RG1002.

36.

Gallon

Tessier

Gobeil

. (2005) Sources and chronology of atmospheric lead deposition to a Canadian Shield Lake: Inferences from Pb isotopes and PAH profiles. Geochimica et Cosmochimica Acta 69(13): 3199–3210.

37.

Garneau

(1995) Collection de référence de graines et autres macrofossiles végétaux de taxons provenant du Québec méridional et boréal et de l’arctique canadien. Sainte-Foy: Geological Survey of Canada, Division de la science des terrains, 47 pp.

38.

Garneau

van Bellen

Magnan

. (2014) Holocene carbon dynamics of boreal and subarctic peatlands from Québec, Canada. The Holocene 24: 1043–1053.

39.

Girardin

Tardif

Flannigan

. (2004) Multicentury reconstruction of the Canadian Drought Code from eastern Canada and its relationship with paleoclimatic indices of atmospheric circulation. Climate Dynamics 23: 99–115.

40.

Givelet

Le Roux

Cheburkin

. (2004) Suggested protocol for collecting, handling and preparing peat cores and peat samples for physical, chemical, mineralogical and isotopic analyses. Journal of Environmental Monitoring 6: 481–492.

41.

Goudie

Middleton

(2001) Saharan dust storms: Nature and consequences. Earth-Science Reviews 56: 179–204.

42.

Grondin

Blouin

J-L

Bouchard

(1980) Étude phyto-écologique de I’archipel de Mingan, vols 1–3 (Report for Groupe Dryade). Québec: Office de planification et de développement du Québec.

43.

Hansson

Rydberg

Kylander

. (2013) Evaluating paleoproxies for peat decomposition and their relationship to peat geochemistry. The Holocene 23: 1666–1671.

44.

Harrison

Kohfeld

Roelandt

. (2001) The role of dust in climate changes today, at the last glacial maximum and in the future. Earth-Science Reviews 54: 43–80.

45.

Hughes

PDM

Blundell

Charman

. (2006) An 8500 cal. year multi-proxy climate record from a bog in eastern Newfoundland: Contributions of meltwater discharge and solar forcing. Quaternary Science Reviews 25: 1208–1227.

46.

Ireland

(1982) Moss Flora of the Maritime Provinces. Ottawa: National Museums of Canada.

47.

Jeglum

Rothwell

Berry

. (1992) A peat sampler for rapid survey (Technical note). Canadian Forestry Service 13: 921–932.

48.

Jessen

Solignac

Nørgaard-Pedersen

. (2011) Exotic pollen as an indicator of variable atmospheric circulation over the Labrador Sea region during the mid to late-Holocene. Journal of Quaternary Science 26: 286–296.

49.

Jowsey

(1966) An improved peat sampler. New Phytologist 65: 245–248.

50.

Krachler

Mohl

Emons

. (2003) Two thousand years of atmospheric rare earth element (REE) deposition as revealed by an ombrotrophic peat bog profile, Jura Mountains, Switzerland. Journal of Environmental Monitoring 5: 111–121.

51.

Lamentowicz

van der Knaap

van Leeuwen

JFN

. (2010) A multi-proxy high-resolution approach to reconstructing past environmental change from an Alpine peat archive. PAGES News 18: 13–15.

52.

Lemay-Tougas

(2014) Changements climatiques le long de la côte nord de l’estuaire du Saint-Laurent Durant l’Holocène: relation entre les conditions hydrographiques et le développement des tourbières ombrotrophes côtières. MSc Thesis, University of Quebec in Montreal (UQAM).

53.

Le Roux

De Vleeschouwer

(2010) Preparation of peat samples for inorganic geochemistry used as palaeoenvironmental proxies. Mires & Peat 7: Article 4.

54.

Le Roux

Marshall

(2011) Constructing recent peat accumulation chronologies using atmospheric fall-out radionuclides. Mires & Peat 7: Article 8.

55.

Le Roux

Fagel

De Vleeschouwer

. (2012) Volcano- and climate-driven changes in atmospheric dust sources and fluxes since the Late Glacial in Central Europe. Geology 40: 335–338.

56.

Loisel

Garneau

(2010) Late-Holocene paleoecohydrology and carbon accumulation estimates from two boreal peat bogs in eastern Canada: Potential and limits of multi-proxy archives. Palaeogeography, Palaeoclimatology, Palaeoecology 291: 493–533.

57.

Magnan

Garneau

(2014) Evaluating long-term regional climate variability in the maritime region of the St. Lawrence North Shore (eastern Canada) using a multi-site comparison of peat-based paleohydrological records. Journal of Quaternary Science 29: 209–220.

58.

Magnan

Garneau

Payette

(2014) Holocene development of maritime ombrotrophic peatlands of the St. Lawrence North Shore in eastern Canada. Quaternary Research 82: 96–106.

59.

Maher

Prospero

Mackie

. (2010) Global connections between aeolian dust, climate and ocean biogeochemistry at the present day and at the last glacial maximum. Earth-Science Reviews 99: 61–97.

60.

Marie-Victorin

(1995) Flore Laurentienne. 3rd Edition. Montreal: Les Presses de l’Université de Montréal.

61.

Marx

Kamber

McGowan

. (2010) Atmospheric pollutants in alpine peat bogs record a detailed chronology of industrial and agricultural development on the Australian continent. Environmental Pollution 158: 1615–1628.

62.

Mauquoy

van Geel

(2007) Plant macrofossil methods and studies: Mire and peat macros. In: Elias

(ed.) Encyclopedia of Quaternary Science. Oxford: Elsevier, pp. 2315–2336.

63.

Mauquoy

van Geel

Blaauw

. (2004) Changes in solar activity and Holocene climatic shifts derived from ¹⁴C wiggle-match dated peat deposits. The Holocene 14: 45–52.

64.

Mayewski

Rohling

Stager

. (2004) Holocene climate variability. Quaternary Research 62: 243–255.

65.

Meskhidze

Chameides

Nenes

. (2003) Iron mobilization in mineral dust: Can anthropogenic SO₂ emissions affect ocean productivity? Geophysical Research Letters 30: 2085.

66.

Mischke

Zhang

Börner

. (2010) Lateglacial and Holocene variation in aeolian sediment flux over the northeastern Tibetan Plateau recorded by laminated sediments of a saline meromictic lake. Journal of Quaternary Science 25: 162–177.

67.

Muhs

(2013) The geologic records of dust in the Quaternary. Aeolian Research 9: 3–48.

68.

Muhs

Budahn

(2006) Geochemical evidence for the origin of late Quaternary loess in central Alaska. Canadian Journal of Earth Sciences 43: 323–337.

69.

Norton

Evans

Kahl

(1998) Comparison of Hg and Pb fluxes to hummocks and hollows of ombrotrophic Big Heath bog and to nearby Sargent Mt. Pond, Maine, USA. Water, Air, and Soil Pollution 292: 271–286.

70.

O’Brien

Mayewski

Meeker

. (1995) Complexity of Holocene climate as reconstructed from a Greenland ice core. Science 270(5244): 1962–1964.

71.

Oldfield

Richardson

Appleby

(1995) Radiometric dating (²¹⁰Pb, ¹³⁷Cs, ²⁴¹Am) of recent ombrotrophic peat accumulation and evidence for changes in mass balance. The Holocene 5: 141–148.

72.

Poole

Sanford

Williams

. (1970) Geology of southeastern Canada. In: Douglas

RJW

(ed.) Geology and Economic Minerals of Canada. Ottawa: Geological Survey of Canada, pp. 228–304.

73.

Pratte

Mucci

Garneau

(2013) Historical records of atmospheric metal deposition along the St. Lawrence Valley (eastern Canada) based on peat bog cores. Atmospheric Environment 79: 831–840.

74.

Rea

(1994) The paleoclimatic record provided by eolian deposition in the deep sea: The geologic history of wind. Reviews of Geophysics 32: 159–195.

75.

Reimer

Bard

Bayliss

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

76.

Sauvé (2016) Reconstitution holocène de la végétation et du climat pour les régions de Baie-Comeau et de Havre-Saint-Pierre, Québec. MSc Thesis. Montreal: University of Quebec in Montreal (UQAM).

77.

Shotyk

Krachler

Martinez-Cortizas

. (2002) A peat bog record of natural, pre-anthropogenic enrichments of trace elements in atmospheric aerosols since 12 370 C-14 yr BP, and their variation with Holocene climate change. Earth and Planetary Science Letters 199: 21–37.

78.

Shotyk

Weiss

Kramers

. (2001) Geochemistry of the peat bog at Etang de la Gruère, Jura Mountains, Switzerland, and its record of atmospheric Pb and lithogenic trace metals (Sc, Ti, Y, Zr, and REE) since 12,370 ¹⁴C yr BP. Geochimica et Cosmochimica Acta 65: 2337–2360.

79.

Steinmann

Shotyk

(1997) Geochemistry, mineralogy, and geochemical mass balance on major elements in two peat bog profiles (Jura Mountains, Switzerland). Chemical Geology 138: 25–53.

80.

Sullivan

Booth

(2011) The potential influence of short-term environmental variability on the composition of testate amoeba communities. Environmental Microbiology 62: 80–93.

81.

Tolonen

(1984) Interpretation of changes in the ash content of ombrotrophic peat layers. Bulletin of the Geological Society of Finland 56: 207–219.

82.

Vanneste

De Vleeschouwer

Martínez-Cortizas

. (2015) Late-glacial elevated dust deposition linked to westerly wind shifts in southern South America. Scientific Reports 5: 11670.

83.

Viau

Gajewski

Fines

. (2002) Widespread evidence of 1500 yr climate variability in North America during the past 14 000 yr. Geology 30: 455–458.

84.

Viau

Gajewski

Sawada

. (2006) Millennial-scale temperature variations in North America during the Holocene. Journal of Geophysical Research: Atmospheres 111(D9). DOI: 10.1029/2005JD006031.

85.

Wanner

Beer

Bütikofer

. (2008) Mid- to Late-Holocene climate change: An overview. Quaternary Science Reviews 27: 1791–1828.

86.

Wedepohl

(1995) The composition of the continental crust. Geochimica et Cosmochimica Acta 59: 1217–1232.

87.

Weiss

Shotyk

Rieley

. (2002) The geochemistry of major and selected trace elements in a forested peat bog, Kalimantan, SE Asia, and its implications for past atmospheric dust deposition. Geochimica et Cosmochimica Acta 66: 2307–2323.

88.

Campbell

. (2003) Carbon sequestration in western Canadian peat highly sensitive to Holocene wet-dry climate cycles at millennial timescales. The Holocene 13: 801–808.

89.

Zdanowicz

Zielinski

Wake

. (2000) A Holocene record of atmospheric dust deposition on the Penny Ice Cap, Baffin Island, Canada. Quaternary Research 53: 62–69.