Drought coincident with aeolian activity in a Great Lakes coastal dune setting during the Algoma Phase (3.1–2.4 ka), southwest Michigan

https://doi.org/10.1016/j.jglr.2021.04.017Get rights and content

Abstract

Aeolian studies of Lake Michigan’s coastal dunes have yet to elucidate what factors control their episodic activity over the past 5000 years. High lake levels exposing sand along with increased storminess is generally accepted for high perched dunes. This hypothesis, however, remains poorly tested for low perched dunes along the southeast Lake Michigan coastline. Here, small lakes in the lee of dune complexes contain aeolian sand and various biological proxies. Age and sedimentation rate models from Gilligan Lake cores guide analysis of aeolian sand, charcoal, pollen, and diatoms at high resolution (1 cm [10.4 yr/cm]) during the high-water Algoma Phase (3.1–2.4 ka) of the upper Great Lakes. The diatoms record a transition from a deep, more acidic lake to a shallower, more alkaline lake with fewer wetlands. This transition is accompanied by a stepped increase in the amount of aeolian sand. There is a weak correlation (R2 0.5, p < 0.01) between increasing abundances of charcoal chunks and sand. Peaks of sand follow peaks in charcoal threads and sheets, suggesting fire played a role in removing vegetation, presaging the landscape for increases in storminess. Arboreal pollen records a transition from a mesic forest Fagus-Acer-Quercus-Abies assemblage suggestive of moister conditions to one richer in mesic hardwoods tolerant of drier conditions. Together, the environmental proxy data record a shallowing lake concomitant with increasing aeolian sand, suggesting that drought-like conditions along the coastline conditioned the landscape for renewed aeolian activity. Once initiated, increased storminess and shoreline erosion maintained dune activity through increased sediment supply.

Introduction

Over the past 20 years considerable work along the upper Great Lakes (Lakes Superior, Michigan and Huron) coastlines has focused on when coastal sand dunes were active. During what might be called a discovery phase, the general consensus is that most coastal dune activity dates from the mid-Holocene to post-settlement time, or since the mid-Holocene Nipissing Phase highstand of the Great Lakes (e.g., Arbogast and Loope, 1999, Hansen et al., 2002, Hansen et al., 2010, Loope et al., 2004, Lepczyk and Arbogast, 2005, Arbogast et al., 2010, Blumer et al., 2012, Rawling and Hanson, 2014, Fulop et al., 2019). However, some of the back dunes, along with other dunes associated with higher elevation deglacial lake phases within the Lake Michigan basin, date back to the earliest Holocene or Late Glacial time (Hansen et al., 2010, Colgan et al., 2017). The ongoing question of what initiates aeolian sand dune activity along Great Lakes coastlines remains poorly answered. During the last 4000 years, water levels in Lake Michigan have shown two major highstands (from 3,600–2,400 years ago and 1,800–1,200 years ago) with a difference of ~1.5 m between water levels in the lowstands and highstands (Thompson and Baedke, 1997). Superimposed on these larger changes are two quasi-periodic water level cycles with periods of ~35 and ~160 years and water level variations on the order of 0.5–0.6 m and 0.5–1.5 m, respectively (Thompson and Baedke, 1997). One popular hypothesis is that dune activity is initiated by water level changes. Hence, most studies suggest new or renewed dune activity during periods of high-water levels which cause erosion, and expose sand to wind which can blow it further inland (Anderton and Loope, 1995, Arbogast and Loope, 1999, Loope and Arbogast, 2000, Loope et al., 2004). In this hypothesis, fresh supplies of sand moving into already established dune fields are responsible for their reactivation. Attempts to test this hypothesis have largely focused on searching for correlations between periods of dune activity and periods of high lake levels in lake level curves like those reconstructed by Baedke and Thompson (2000). Unfortunately, the relationship between lake level and dune activity is difficult to test when the dating errors for both data sets are considered. Moreover, errors are rarely considered for the disparate sediments and landforms used to reconstruct dune activity or dune stability.

A variant of the lake-level hypothesis for dune mobility, based on their study of high-perched transgressive dunes on sandy glacial sediment further north on the northeastern shore of Lake Michigan, was put forward by Blumer et al. (2012). In their model, dune mobility is driven by fluctuations in lake level during periods of overall drought (Arbogast et al., 2011). The main evidence for this hypothesis was the initiation of a major period of dune activity ~1000 years ago during a period of dry conditions that coincided with the Medieval Warm Period. In this paper we focus on environmental factors associated with dune activity in low-perched dunes (see Hansen et al., 2010) with bases a few meters above lake level in the southern half of the western Michigan coastline. These dunes are distinguishable from high-perched, transgressive dunes resting on sandy glacial sediment further north along the Michigan coastline (e.g., Blumer et al., 2012). For the southern study area, clustering of radiocarbon (14C) and optically stimulated luminescence (OSL) ages suggests episodic dune activity (Fig. 1E) that loosely corresponds with higher water levels in the Lake Michigan basin since the Nipissing peak in lake level 4500 years ago (Hansen et al., 2010).

Transgressive dunes, like the ones in our study area (Fig. 1), are common throughout the southeastern shore of Lake Michigan (Hansen et al., 2020), and are also common on ocean coasts (Hesp and Thom, 1990). Of the three scenarios that Hesp (2013) proposed for the development of transgressive dune fields, scenario 2, initiation by foredune and or dune erosion, appears to be most applicable to Lake Michigan dunes. Along the lake’s southeastern shore, the erosion of the foredune complex appears to have led to the development of blowouts which migrated downwind, forming the transgressive complexes (Fig. 1B, C). The initial formation of these dunes appears to have coincided with the rise in lake water levels associated with the Algoma phase of Lake Michigan approximately 3200 years ago (Hansen et al., 2010). Since this time, the transgressive dune fields have undergone periods of stability punctuated by episodes of renewed dune mobility. Hesp (2013) notes that, after the initiation of transgressive dune activity, the growth of a new foredune complex can cut off the supply of new sand to the dune field and this appears to have been the case in many dune fields along the southeastern shore of Lake Michigan. After the formation of this barrier, renewal of dune migration requires either the cannibalization and remobilization of sand within the transgressive dune complex or a breach or bypass in the foredune system by which new sand can be supplied to the transgressive dunes. The established foredune ridges at the landward edge of the foredune complexes are typically greater than 10 m high with steep (>25°) slopes and lack the sand ramps through which sand can be transported to and beyond the crest in the mechanism proposed by Christiansen and Davidson-Arnott (2004). Although these established foredune ridges may be difficult to bypass, breaches in these ridges are relatively common today (see the maps in Hansen et al., 2010) and represent pathways through which sand from the beach can be funneled into the transgressive complexes.

While models for the reactivation of mobility in transgressive dunes along the Laurentian Great Lakes have tended to focus on changes in lake water levels, most models for ocean coasts have emphasized climatic factors. Particularly important in these models has been variations in the frequency and intensity of storms (storminess) with periods of dune reactivation in transgressive complexes attributed to high storminess (Mason et al., 1997, Wilson et al., 2004, Clarke and Rendell, 2009, Clemmensen et al., 2009, Björck et al., 2012, González-Villanueva et al., 2013, Costas et al., 2016). An increase in storminess increases the wind energy available to transport sand. However, it is widely recognized that in vegetated coastal dunes increased wind energy by itself will probably not cause dune reactivation unless combined with another trigger that removes or weakens dune vegetation. One obvious trigger, tied directly to an increase in storminess, is an increase in wave erosion which can remove protective vegetation, exposing sand to transportation by the wind. In a review paper on dune reactivation along European coasts during the Little Ice Age, Jackson et al. (2019a) noted that sea level oscillations, both rising and falling, have been suggested as contributing to dune mobility as has anthropogenic disturbance. They suggest that a decrease in air temperature can contribute to mobility by shortening growing seasons, inhibiting the stabilization of dunes by the growth of new vegetation. A decrease in rainfall could also place stress on vegetation, making it more difficult for dunes to stabilize once they become mobile. In recent years this has been illustrated by the converse process in which increases in rainfall contribute to an increase in the vigor of dune vegetation. In a paper which discusses the cause of a widespread increase in the stability of coastal dunes since 1984, Jackson et al. (2019b) attribute this change to an increase in dune vegetation due, in part, to both an increase in temperature and an increase in precipitation. In a similar fashion, Wolfe and Hugenholtz (2009) attribute the stabilization of noncoastal dunes in the northern North American Great Plains during the 19th century to an increase in both rainfall and temperature which allowed stabilization by growing vegetation to outpace destabilization by sand transport.

Chronologic data for most coastal dune studies consist of radiocarbon ages from charcoal, wood and soil underlying and within sand dunes, and optically stimulated luminescence (OSL) ages from dune sand. The general assumption is that organic matter within dune sand is from plants killed during or just slightly after sand burial and, hence, represents the onset of dune activity. On the other hand, OSL ages give the time at which the sand was buried and no longer exposed to sunlight and, hence, represent the onset of dune stability. Complicating matters is that charcoal and wood from dune sand can be re-transported and re-deposited whenever a dune reactivates. Thus, associating the death of the plant with dune activation is not always possible. In situ stumps within a soil are an exception. Unfortunately, many of the radiocarbon dates can only be considered maximum limiting ages for aeolian activity (Timmons et al., 2007). General practice is to sample the crest of a stabilized (vegetated) sand dune for determining last activity using OSL dating. Argyilan et al. (2014) OSL dated several geomorphic positions along two stabilized parabolic sand dunes at the southern end of Lake Michigan (e on Fig. 1A) and suggested that dates from the crest or lee slope record late stages of dune migration, possibly affected by short-term local disturbance instead of dune migration. Reported errors for radiocarbon dating are a few percent using the AMS method, while OSL age errors can be generally in the range of 8–15%. The same dating methods with the same errors are also inherent for the dated beach ridges used to develop the Lake Michigan relative lake level curve (LMRLLC, Baedke and Thompson, 2000; Fig. 1E). The LMRLLC is a Fourier smoothed line joining five strandplain sites within the Lake Michigan basin, with each site’s age assignment a best fit line between numerous minimum radiocarbon dates (Thompson and Baedke, 1997). The LMRLLC is a record of high lake levels only and illustrates the ~33-year (0.5–6 m) and ~160-year (0.5–1.5 m) fluctuations in high lake levels. Here we recognize that past linkages of dune activity with high lake level are hampered by uncertainty in how ages are assigned to specific dune activity or a specific beach ridge age, sharing dating errors (14C and OSL) that are similar if not greater than the ~33 yr or ~160 yr fluctuations in lake level. Additionally, for the time period of this study, water levels in the Superior, Michigan, and Huron basins formed a single lake (Farrand and Drexler, 1985) referred to here as the upper Great Lake.

A different approach for improving age control of past dune activity, referred to here as sand signals, uses variations in the weight percent sand in well-dated lacustrine sequences immediately downwind of sand dunes as a proxy for dune activity. This methodology builds on early work by Keen and Shane (1990) in Wisconsin and in Michigan where Fisher and Loope (2005) proposed that peaks in sand from Silver Lake, MI (Fig. 1A) recorded the Nipissing highstand and subsequent ~160-year cyclic highstands in the LMRLLC. Timmons et al. (2007) developed sand signal time series for several cores in Gilligan Lake, the site chosen for this study. The lake is located downwind of several large parabolic dunes and its lacustrine sediment contains numerous fine laminations of aeolian sand. Using an age model assuming constant sediment rates, Timmons et al. (2007) observed that peaks in aeolian sand aligned with high levels in the LMRLLC only 50 percent of the time. Elsewhere, sand flux has been used in the North Atlantic region (e.g., Björck and Clemmensen, 2004, Clemmensen et al., 2009, Nielsen et al., 2016) and the coast of Chile (Björck et al., 2012) to reconstruct past storminess. The underlying assumption behind all of these methods is that the sand is a result of direct grainfall of aeolian transported sand onto open water, or snow and ice when active in wintertime. In Michigan, an aeolian origin for the sand signal was successfully evaluated at several sites in Allegan County against alternative sand delivery processes, including fluvial input, mass wasting and shoreline dynamics, such as sheet wash (DeVries-Zimmerman et al., 2014). They also observed long distance (~200 m) suspension transport of aeolian sand from dune crests to Gilligan Lake. In a follow-up study, Bodenbender et al. (2018) set out a series of traps designed to catch suspended aeolian sand in the lee of a large parabolic dune. After the period of winter storms, they found significant amounts of sand in their furthest trap, 215 m away from the base of the dune. A benefit of developing sand signals from aeolian sand in lacustrine sediment is the greater preservation potential of the aeolian record than might be expected from organic records within sand dunes. However, this method also suffers from dating errors inherent to calculating sedimentation rates.

Sand signals have been used elsewhere along the western coast of Michigan. Baca et al. (2014) found correlation of peaks in sand over the past 100 years (using 210Pb/137Cs/7Be dating), with peaks in water levels of Lake Michigan and the winter Palmer drought severity index, suggesting dune activity is linked with periods of wet conditions and storminess. Similarly, Fisher et al. (2007) found that highstands in Lake Michigan since the 1930s correspond with higher rates of sand dune migration along the western side of Silver Lake, MI (Fig. 1). At the Grand Mere Embayment, MI (d on Fig. 1A), Hanes et al. (2014) used spectral analysis to investigate sand signals from six cores taken from three adjacent lakes. After determining that sand deposition was not random, they found numerous periodicities within the sand signals corresponding to lake level cycles based on strandplain analysis (i.e., LMRLLC), solar cycles, and sand and climate cycles from other regional lakes (e.g., Dean, 1997, Fisher et al., 2012). Together, these results suggest that delivery of aeolian sand into lakes downwind of sand dunes along the Lake Michigan coastline is not random, but records periods of past storminess, presumably during periods of high lake level. Increased storminess over the Lake Michigan basin is expected to bring more geomorphologically effective wind. Thus, a positive correlation between high or rising lake levels and dune reactivation is expected (see Anderton and Loope, 1995, Loope and Arbogast, 2000).

An alternative model to high-lake levels and increased storminess leading to renewed dune activity is the drought model developed for the Great Plains of the USA. Here, the stabilizing vegetation is removed or significantly reduced by drought due to increasing moisture/plant stress, increasing susceptibility to pests/pathogens, and/or fire. During the Holocene, dunes were reactivated episodically during warmer and drier climate periods (Forman et al., 2001, Wolfe et al., 2017, Mason et al., 2020). The mesic forests along the Michigan coastline are impacted by both drought in the short-term and climate change in the long-term (e.g., Handler et al., 2014, Rollinson et al., 2020). Therefore, drought may be enough of a stressor, in combination with other factors, to result in renewed aeolian activity. Widespread dune activity across the eastern Upper Peninsula of Michigan from 10 to 8 ka is explained by drought (Loope et al., 2012). Researchers at inland sites in northeast Lower Peninsula of Michigan (Arbogast et al., 2010) and at inland sites along the Upper Peninsula of Michigan (Arbogast et al., 2002a, Arbogast et al., 2002b, Arbogast and Packman, 2004) suggested that drought was one of numerous contributing factors to dune activity between 7.0 and 5.5 ka. Arbogast et al., 2011, Blumer et al., 2012 suggest that drought was also a contributing factor to the remobilization of high-perched transgressive coastal dunes along the northeastern shore of Lake Michigan during the Medieval Warm Period roughly 1000 years ago.

To make progress on understanding the mechanisms responsible for coastal dune reactivation, the age control for the various data sets must be the same. The approach used in this paper is to develop all data sets from the same sediment core, sampled at the same resolution of 1-cm for each environmental proxy. This approach permits direct comparison of all datasets using the same relative time scale. Thus, the purpose of this paper is to develop multi-proxy paleoenvironmental records that include the aeolian sand signal (upwind dune activity) to determine processes that may be triggering coastal dune activity. The specific datasets include fire history (charcoal), vegetation (pollen) and water levels (diatoms) in the cored lake.

The study site at Gilligan Lake is 0.75 km inland from the Lake Michigan shoreline (Fig. 1B) and 10 km south of the town of Holland, MI (c on Fig. 1A). The site is well described in Timmons et al., 2007, DeVries-Zimmerman et al., 2014. Briefly, the 8-ha crescent-shaped lake formed during the Nipissing transgression when rising water levels in Lake Michigan flooded a stream valley and a barrier bar that is since masked by dunes formed along the shoreline (see Dorr and Eschman, 1970). The lake drains northward into Kelly Lake and then through Lake Macatawa into Lake Michigan (Fig. 1C). Backdunes close to Gilligan Lake record dune migration following the Nipissing highstand. After the post-Nipissing interlude and during the Algoma highstand, many of the high-relief parabolic dunes developed and migrated over the back dunes (Fig. 1B, E; Hansen et al., 2010). From ~2 ka to ~0.5 ka, the well-developed Holland paleosol formed in dunes along 200 km of the southeast coast of Lake Michigan, recording regional landscape stability (Arbogast et al., 2004). Approximately 500 years ago, many of the dunes were reactivated. Other than a few active dunes with some reaching 65 m height, the landscape is stabilized west of Gilligan Lake with a forest that was predominantly a hemlock-beech-maple (Tsuga-Fagus-Acer) forest before European settlement (Kenoyer, 1934, Brewer et al., 1984, Comer and Albert, 1997). Coastal dunes record westerly wind directions.

An unusually large number of radiocarbon dates from dune paleosols (Arbogast et al., 2002a, Arbogast et al., 2002b), and OSL ages from dune sand (Hansen et al., 2002, Hansen et al., 2010) exists for the dunes to the west of Gilligan Lake, allowing us to reconstruct an especially detailed chronology of dune growth and evolution. There is good overlap between chronology of dune mobility derived from these studies in the dunes and the chronology of sand peaks in cores from Gilligan Lake (Timmons et al., 2007, DeVries-Zimmerman et al., 2014). This indicates that, in this lake at least, peaks in sand abundance in sediment cores can be a reliable indicator of sand mobility in the adjacent dunes. The simplest explanation is that the sand in the lake sediments is the result of direct grainfall of aeolian transported sand as discussed in the previous section on Sand Signals. Even if the sand was delivered to the lake by some other mechanism, for example sheet wash, this activity appears to have been correlated with periods of mobility in the neighboring dunes, possibly because they both were a response to the same underlying factors. Thus, correlations between the sand signal and other environmental proxies in the Gilligan Lake core can potentially give useful insights into the factors responsible for mobility in the dunes.

Section snippets

Coring and core sampling

A canoe-pontoon raft was used as a coring platform (Fig. 1D) in 6.4 m of water at the center of the larger of the two sub-basins within Gilligan Lake (Fig. 1). Once visible sand laminae were identified from a Livingstone core having a stratigraphy similar to that observed in core GL104-1 from Timmons et al. (2007), one vibracore, GL13V2, (Fisher, 2004) and one Livingstone core, GL13L, were collected. The Livingstone core was extruded directly into a black plastic tube and sent to the Optical

Lithology, chronology & sedimentation rates

The chronology and lithostratigraphy of the studied cores, GL13V2 and GL13L, are shown in Fig. 2. The two cores studied here contain the same type of sediment as core GL104-1 collected previously from this lake (Timmons et al., 2007). The sediment consists of black sapropel with subtle color banding and laminations of sand (Fig. 2C, D) overlying sapropelic sand and silt, all of which is described in greater detail by Timmons et al. (2007). The single radiocarbon age of 2080 cal yr BP

Discussion

The high-resolution datasets sharing the same age control make this a novel paleoenvironmental investigation of aeolian sediments within a lake in the lee of sand dunes. Summary proxy environmental data are plotted beside each other in Fig. 6 with the sand signal showing increasing abundance of sand concomitant with a transition to more drought like conditions. The sand signal data of stepped increases in sand abundance (Fig. 2F, 6) up core span ~600 years between ~3100–2500 years ago. OSL ages

Conclusions

Four high resolution (~10 yr) paleoenvironmental datasets sharing the same age model allow for detailed environment reconstructions of a coastal dune setting along the southeast coast of Lake Michigan during the Algoma Phase. Up-core changes in charcoal, diatom, and pollen data spanning ~600 years record initial moist conditions and high-water levels in Gilligan Lake within a hemlock-beech-maple forest that transition to drier conditions represented by low water levels, increased ratios of

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by the National Geographic Society grant #9649-15 to TGF, and a Course-based Research Experience (CRE) award from Hope College's HHMI grant to SDVZ. Mitchel Dziekan and Amy Towell assisted with collecting the cores. Laboratory assistance provided by Randi Ulmer and Ben Johnson from Hope College and Kayleigh Alme from North Dakota State University. We would like to acknowledge the support of The Ohio State University Nuclear Reactor Laboratory and the assistance of

References (116)

  • S. Costas et al.

    Windiness spells in SW Europe since the last glacial maximum

    Earth Planet. Sci. Let.

    (2016)
  • S.L. Forman et al.

    Temporal and spatial patterns of Holocene dune activity on the Great Plains of North America: megadroughts and climate links

    Glob. Planet. Change

    (2001)
  • E.C.F. Fulop et al.

    A geochronology-supported soil chronosequence for establishing the timing of shoreline parabolic dune stabilization

    Catena

    (2019)
  • R. González-Villanueva et al.

    Impact of atmospheric circulation patterns on coastal dune dynamics, NW Spain

    Geomorphology

    (2013)
  • E.C. Hansen et al.

    Geomorphic history of low-perched, transgressive dune complexes along the southeastern shore of Lake Michigan

    Aeolian Res.

    (2010)
  • D.W.T. Jackson et al.

    Large-scale transgressive coastal dune behavior in Europe during the Little Ice Age

    Global Planet. Chg.

    (2019)
  • Z. Kilibarda et al.

    Morphology and origin of the Fair Oak Dunes in NW Indiana, USA

    Geomorphology

    (2011)
  • J.E. Kutzbach et al.

    Climate and biome simulations for the past 21,000 years

    Quat. Sci. Revs.

    (1998)
  • W.L. Loope et al.

    Dominance of an ~150-year cycle of sand-supply change in Late Holocene dune-building along the eastern shore of Lake Michigan

    Quat. Res.

    (2000)
  • W.L. Loope et al.

    A Holocene history of dune-mediated landscape change along the southeastern shore of Lake Superior

    Geomorphology

    (2004)
  • J.A. Mason et al.

    Late Pleistocene dune activity in the central Great Plains, USA

    Quat. Sci. Revs.

    (2011)
  • A.S. Murray et al.

    Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol

    Radiat. Meas.

    (2000)
  • J.R. Prescott et al.

    Cosmic ray and gamma ray dosimetry for TL and ESR

    Nucl. Tracks Radiat. Measurement

    (1988)
  • J.R. Prescott et al.

    Cosmic ray contributions to dose rates for luminescence and ESR dating: large depths and long-term time variations

    Radiat. Measurements

    (1994)
  • J.E. Rawling et al.

    Late Pleistocene dune construction in the Central Sand Plain of Wisconsin, USA

    Geomorphology

    (2008)
  • R.J. Shedlock et al.

    Interactions between ground water and wetlands, southern shore of Lake Michigan, USA

    J. Hydro.

    (1993)
  • S. Shumack et al.

    The impact of fire on sand dune stability: Surface coverage and biomass recovery after fires on Western Australian coastal dune systems from 1988 to 2016

    Geomorphology

    (2017)
  • M.J. Aitken

    An Introduction to Optical Dating: The Dating of Quaternary Sediments by the Use of Photon-Stimulated Luminescence

    (1998)
  • A.F. Arbogast et al.

    Middle-Holocene mobilization of aeolian sand in western upper Michigan and the potential relationship with climate and fire

    Holocene

    (2004)
  • A.F. Arbogast et al.

    Widespread middle Holocene dune formation in the eastern Upper Peninsula of Michigan and the relationship to climate and outlet-controlled lake level

    Geology

    (2002)
  • A.F. Arbogast et al.

    The Holland Paleosol: an informal pedostratigraphic unit in the coastal dunes of southeastern Lake Michigan

    Can. J. Earth Sci.

    (2004)
  • Arbogast, A.F., Monaghan, G.W., Black, G.A., Lovis, W.A., 2011. Exploring the potential linkage between coastal sand...
  • Argyilan, E.P., Lepper, K., Thompson, T.A., 2014. Late Holocene coastal development along the southern shore of Lake...
  • Baca, K.J., Fisher, T.G., Gottgens, J.F., 2014. Temporally constrained eolian sand signals and their relationship to...
  • S.J. Baedke et al.

    A 4,700-year record of lake level and isostasy for Lake Michigan

    J. Great Lakes Res.

    (2000)
  • B.V. Barnes et al.

    Michigan Trees: A Guide to the Trees of the Great Lakes Region

    (2004)
  • R.W. Battarbee

    A new method for the estimation of absolute microfossil numbers, with reference especially to diatoms

    Limnol. Oceanogr.

    (1973)
  • Belford, A., Kenbeek, S.D., VanHorn, J., van Dijk, D., 2014. Using remote sensing and geospatial analysis to understand...
  • S. Björck et al.

    Aeolian sediment in raised bog deposits, Halland, SW Sweden: a new proxy record of Holocene winter storminess variation in southern Scandinavia?

    Holocene

    (2004)
  • Bodenbender, B.E., Hansen, E.C., Yurk, B.P., DeVries-Zimmerman, S.J., Stid, J.T., 2018. Sand deposition by grainfall...
  • R.K. Booth et al.

    A high-resolution record of late-Holocene moisture variability from a Michigan raised bog, USA

    Holocene

    (2003)
  • L.G. Brewer et al.

    Presettlement vegetation of southwestern Michigan

    Michigan Botanist

    (1984)
  • K.E. Camburn et al.

    The genus Melosira from soft-water lakes with special reference to northern Michigan, Wisconsin and Minnesota

  • Changnon, S.A. Jr., 1987. Detecting drought conditions in Illinois. Illinois State Water Survey Circular 169....
  • M.B. Christiansen et al.

    Rates of landward sand transport over the foredune at Skallingen, Denmark and the Role of Dune Ramps

    Geografisk Tidsskrift Danish J. Geogr.

    (2004)
  • M.L. Clarke et al.

    The impact of North Atlantic storminess on western European coasts: a review

    Quat. Int.

    (2009)
  • P.M. Colgan et al.

    Inland dunes on the abandoned bed of Glacial Lake Chicago indicate eolian activity during the Pleistocene-Holocene transition, southwestern Michigan, USA

    Quat. Res.

    (2017)
  • Comer, P.J., Albert, D.A., 1997. Vegetation circa 1800 of Allegan County, Michigan: An interpretation of the General...
  • W.E. Dean

    Rates, timing, and cyclicity of Holocene eolian activity in north-central United States: Evidence from varved lake sediments

    Geology

    (1997)
  • DeVries-Zimmerman, S., Fisher, T.G., Hansen, E.C., Dean, S., Bjorck, S., 2014. Sand in lakes and bogs in Allegan...
  • Cited by (0)

    View full text