Home


       

Habitat Change in a Perched Dune System Along Lake Superior

Pitcher's Thistle and Lake Level Change

 

Pitcher's thistle (Fig. 1), listed as a threatened species by the U.S. Fish and Wildlife Service, inhabits the sandy shores of the upper Great Lakes. Its life cycle requires a dynamic habitat that periodically provides both highly disturbed, open patches (affording sites for seed germination on mineral soil in high light under limited competition) and more stable patches for flowering, seed set, and seedling establishment before burial or other site destruction (McEachern 1992). Since water levels of the lakes greatly influence the effects of waves upon the shore, the fine-scale history of lake-level change is relevant to understanding past habitat change, the persistence of Pitcher's thistle, and habitat requirements for restoration and protection of the species.

Fig. 1. Pitcher's thistle.
Courtesy K. McEachern, USGS

Recent studies of lake-edge dunes and beach ridges (Fraser et al. 1990; Lichter 1995; Thompson and Baedke 1995) suggest that because of climatic variation, past levels of Lake Michigan have differed from those of the present by as little as several tenths of a meter to as much as a meter or more. These changes have taken place periodically over decades to centuries during the past 5,000 years. The growth of open habitat patches through dune building in lake-edge dunefields is greatest as low lake levels bare broad sand flats, expanding the source of sand for transport by the wind (Fraser et al. 1990). Perched dunefields, which form high atop lake-facing bluffs, also respond to lake- level change (Marsh and Marsh 1987; Anderton and Loope 1995) but in a mirror-image fashion to those on the lake edge (Olson 1958). In perched dunefields, growth of open habitats (dune building) occurs as rising lake levels destabilize lake-facing bluffs, creating active colluvial slopes and increasing sand supply to the bluff tops.

   

For both lake-edge and perched dunes, the mix of habitat change may depend upon smaller-scale episodes of lake-level change that are nested within larger, longer-term trends (Fraser et al. 1990; Thompson and Baedke 1995; Fig. 2). Lake-level changes, then, have probably mediated expansion and contraction of habitat patches suitable for Pitcher's thistle on several scales (Snyder 1985; Businski 1992; McEachern et al. 1994) since the appearance of the Great Lakes about 10,000 years ago.

Fig. 2. Hypothetical late Holocene hydrograph proposed for Lake Michigan (Fraser et. al. 1990) in relation to historical lake levels.
© Geological Society of America

Episodes of habitat change, driven by changes in levels of the Great Lakes, must be considered when assessing human effects upon coastal vegetation and rare species (Schultz 1988; Businski 1992). Paleoecological studies, baseline inventories, and long-term monitoring programs within the Grand Sable Dunes, a perched-dune system along Lake Superior, provide a window on vegetation change at different spatial and temporal scales and also provide an illustrative case study.

   

Vegetation Change

 

The modern Grand Sable Dunes are characterized by a shifting mosaic of plant communities and physical dune forms periodically disturbed as sand builds, stabilizes, and erodes away from the dune system (Fig. 3). Exhumed forests and buried soils in this landscape (Fig. 4) attest that vegetation cover has varied significantly over hundreds to thousands of years and give us a crude picture of coarse-scale changes (Anderton and Loope 1995). Analysis of present-day vegetation and plant population dynamics helps reveal how plant communities may change within tens to hundreds of years and allows us to predict habitat suitability for disturbance-adapted species like Pitcher's thistle for the next few decades.

a.
Fig. 3. Grand Sable Dunes vegetation mosaic: a) pine forest, b) grassland, and c) dunes.

Courtesy W. L. Loope, USGS

 

b.
Courtesy W. L. Loope, USGS

 

c.
Courtesy W. L. Loope, USGS

Paleoecological Change

 

The presence of buried soils within Grand Sable Dunes implies sharply contrasting rates of sand supply to the dunefield. Radiocarbon dating of buried soils suggests that at least 5 and perhaps as many as 11 episodes of soil burial have occurred there over the past 5,500 years (Anderton and Loope 1995). Soil burial events were probably related to a much greater supply of sand than that of the present (Fig. 5a), which occurred as lake-facing bluffs were destabilized by the rising waters of Lake Superior (Fig. 5b). Periods of bluff stability during low water on Lake Superior (Fig. 5c) allowed vegetation to invade sand-starved dunes. Whether each buried soil represents complete forestation of Grand Sable Dunes is questionable, but the presence of charcoal in several soil profiles supports the possibility that vegetation occasionally became continuous enough to carry a fire. During periods of afforestation, Pitcher's thistle would have been restricted to small, isolated, and disturbed areas along the bluff edge or in rare inland blowouts. During periods of high sand supply and dune building, Pitcher's thistle would have been afforded a broader spectrum of open habitat. Rapid dune building may also have limited the availability of intermediate sites, which were stable long enough to permit completion of the flowering cycle of Pitcher's thistle but were open enough to permit germination of new seedlings.

Fig. 4. Buried soil in a perched dunefield.
Courtesy W. L. Loope, USGS

Contemporary Vegetation Change

 

The most striking evidence of contemporary vegetation change within Grand Sable Dunes is the increase of jack pine forest over the last several hundred years. An aerial photo time series shows a fivefold increase in forest cover over the past 50 years. Snyder (1985) and Businski (1992) report similar results at Sleeping Bear Dunes, a perched dune system along Lake Michigan. Stand-age structure within forest patches at Grand Sable Dunes suggests that afforestation began at the landward edge of the dunes at least 125 years ago. The plant species richness in the forest increases with stand age and is strikingly higher than in the fire-influenced pine stands to the south and east of Grand Sable Dunes. Although charcoal from ancient soil profiles (Anderton and Loope 1995) suggests fire-prone vegetation occurred there in the past, burned snags and species usually associated with fire, such as blueberries, wintergreen, and bracken, are absent from modern pine patches.

Fig. 5. Drivers of change a) within the modern Grand Sable Dunes, b) under a dune-building future, and c) under a stabilization future.


Within the large-scale trend toward increasing forest cover, open patches are still being created on a smaller scale just landward of the lake bluff and in inland blowouts. Species composition and cover within open patches are determined by proximity to the lake-facing bluff edge and by fine-scale patterns of stability along the bluff face. Dominance of American beachgrass and wormwood along the bluff edge, both of which tolerate burial, reflects a high sand supply and fast patch turnover because of periodic sloughing of the bluff face. Shrubs such as wooly beachheater and bearberry and the bunch-grass, little bluestem, dominate more stable, open patches farther landward of the lake bluff, where there is less blowing sand.

   

Linking Nested Episodes of Change

 

Coastal geomorphology is important in understanding and predicting species persistence (Pavlovic et al. 1991). Periods when open habitats become rare or inaccessible to disturbance-adapted species (the bottlenecks of Loveless and Hamrick 1988) must have occurred at Grand Sable Dunes during nested landscape changes beginning about 5,500 years ago and continuing into the present.

   

Repeated episodes of both afforestation and soil burial at Grand Sable Dunes imply varying habitat quality for Pitcher's thistle over late Holocene time. Depending on the magnitude and duration of lake-level changes, open habitats may have been restricted to the bluff edge during low water. During high water, advancing dunes may have limited the extent of intermediate habitat required for the 8- to 10-year life cycle of Pitcher's thistle. The temporal and spatial details of landscape history can be linked with species' life histories, which allows for more realistic and spatially explicit population models.

   

Present successional trends at Grand Sable Dunes seem to be toward increasing forest cover. The landward half of Grand Sable Dunes preserves a record of successional change since the last major destabilization about 500 years before the present; the lakeward half presents a composite picture of changes over the last several hundred years in response to localized changes along the bluff. The same characteristics of perched dunes that make them valuable for studying buried soils and vegetation change over the last 5,000 years also make them valuable for studying recent vegetation trends. The perched dunes are remote from direct wave action and apparently respond only to more sustained changes of lake levels.

   

Grand Sable Dunes has experienced relative stability for the last 150 years, allowing jack pine to invade portions of the central and eastern dunes. An episode of dune building about 500 years ago buried the red pine and other early successional species that had begun to invade the dunes during a previous stable period. The duration and timing of such episodes have constrained the distribution of dunes-adapted plants during the late Holocene, alternately favoring species adapted to open sites with high sand supply and then favoring those adapted to shaded sites in early stages of succession. These changes in the dunefield habitat mosaic appear controlled by changes in the water levels of Lake Superior.

   

Regional reconstructions of paleolandscape dynamics have implications for studies of evolutionary ecology of narrowly distributed plants along the shores of the Great Lakes. Current research suggests that the hypothetical multiple successional pathways of Bach (1978) have indeed been a part of the recent history of Grand Sable Dunes. The turnover rate for small patches is presently quite rapid along the lakeward edge of the dunefield and decreases inland. The present turnover rate depends on a relatively low volume of sand along the lake-facing bluff. The sizes, distribution, and turnover rates of patches seem to have changed significantly throughout the late Holocene.

   
  Authors
Walter L. Loope
U.S. Geological Survey
Biological Resources Division
P. O. Box 40
Munising, Michigan 49862
A. Kathryn McEachern
U.S. Geological Survey
Biological Resources Division
1901 Spinnaker Drive
Ventura, California 93001

References


Home