Changes in the Upland Forests of the Douglas Lake Region Over the Past 75 Years - Claudia L. Jolls, 1983
Changes in the Upland Forest of the Douglas Lake region over the past 75 years
Claudia L. Jolls
A Lecture given at the Station's Diamond (75th) Jubilee - 1983
When Dave Gates suggested I prepare a talk describing changes in the terrestrial ecosystems in the Douglas Lake region, my excitement was surpassed only by the cold grip of fear which seized me. As a plant ecologist, I was exhilarated at the thought of reviewing and synthesizing the literature. After a cursory review of the list of 1,700 Station publications, my exhilaration changed to trepidation, and after a more extensive review of 58 publications, dealing just with changes in terrestrial vascular flowering plants, my excitement and trepidation turned to exhaustion.
The most overwhelming aspects I encountered during my preparation for this talk were (1) the dramatic changes that have occurred in our vicinity during the past 75 years and (2) the vast amount of work, both published and unpublished, documenting these changes. Douglas Lake is unique in having a rich, extensive, and well documented history, gained while fulfilling our mission of teaching and research in field biology.
Thus, I was left with 1,700 publications, 114 courses, 150 faculty, several hundred researchers, and 7,000 students, most of whom did independent projects for their courses. The compilation of all this in formation has been tackled in earnest through the efforts centered around the Station's list of publications, our archives, copies of student papers, and, most recently, the Gazetteer and our efforts with data management. The accumulation of this information has taken 75 years and I was beginning to feel that its synthesis would take an additional 75.
I had a title: "Changes in the Terrestrial Ecosystems Over 75 Years"; the words in the title connote several concepts and ideas. The word "changes" implies the process of succession. The phrase "terrestrial ecosystems" suggests those extensive plant communities that control to some degree the other organisms inhabiting them. The most conspicuous of the plant communities is the forest and our re search history for this community, particularly the upland forest, is richest. I have deferred to my colleague, Win Fairchild, for a discussion of the wetlands and have opted to ignore the dunes and shoreline vegetation. My thirty years of life give me a very myopic perspective given that much of our forest is at least 70 years old; even one of our youngest communities, the 1980 burn, has a tenure at UMBS longer than mine. My caveat is I have overlooked some very important contributions. Although my views have undergone considerable evolution as the result of discussion with my colleagues, they are nonetheless my views and interpretation of others' work. I will attempt to illustrate some of the changes in this area during the past 75 years using selected studies and generalizations from many efforts, only some of which I can acknowledge in this discussion.
Succession is defined as a progressive, directional, and predictable or nonrandom process of change in species composition in a given area, generally accompanied by changes in diversity, productivity, and biomass. Some schools of thought conclude that this process yields a steady, self-perpetuating state termed climax, however, others feel climax is an academic state that is never realized.
Views of succession have changed. The earliest considerations, championed by F.E. Clements (1916, 1936), viewed succession as a community-level process. The community was considered an organ ism and succession was its life cycle. Transitions in a given area were believed to be replacement of one community by another rather than a continuous process of replacement of species. Changes mediated by the early successional species were to their own detriment, paving the way for later successional species. Clements viewed succession as a progression toward a homogeneous species composition or monoclimax, independent of initial site and vegetation composition.
Ecologists have argued whether a climax community is unique and predictable for a given regional climate or whether each region is characterized by a mosaic of climax types (the polyclimax of Whittaker 1953, 1973). This is a very important view, given the diversity of habitats in the Douglas Lake region. The most recent embellishments on traditional views of succession have trended away from the Clementsian view of a community level process to greater emphasis on population-level interactions. These views emphasize the life histories of individual species, the role of disturbance in shaping communities and the strong influence of random events (West et al. 1981). Both schools of thought are represented by Station research.
There are several methods of documenting succession, all rep resented by studies at UMBS. These methods include: (1) single measures of one plot through time, (2) observations on nearby plots of different successional ages, (3) analysis of historical and pre historical records, (4) experimental manipulations, and (5) mathematical modeling.
The present day vegetation patterns and the faunal communities dependent on the primary producers are a reflection of the vegetation here prior to 1909 and the forces which shaped it. These initial vegetation patterns were shaped by glaciation, climate, and soils, as well as by internally and externally caused disturbance following initial establishment. I will begin by describing the forest the settlers saw, the impact of human settlement (or dissettlement as you will see) and the resultant changes that have been monitored at UMBS during the past 75 years.
The present-day landscape was released from glaciers 10,000-12,000 years (Gold, Gannon, and Paddock 1979) and sometime after 11,400 years ago plants reentered Cheboygan County (Kilburn 1957). Even after the final retreat of the ice, fluctuating lake levels shaped the landscape around Douglas Lake. Pleistocene Lake Algonquin connected Burt, Douglas, and Mullett Lakes at an elevation of 225 m above sea level 11,000 to 8,000 year ago. Much of UMBS was under Lake Algonquin, except for the highlands. Lake Nipissing existed 4,000 to 3,500 years ago at a level of 189 m above sea level and combined Lakes Superior, Huron, and Michigan. Despite its large size, it affected only small areas in the region around Douglas Lake (Spurr and Zumberge 1956). Thus, near UMBS, various areas have been exposed to vegetation for different lengths of time. Areas above Lake Algonquin, that is areas now above 225 m, have been exposed since ice retreat, approximately 11,000 years. Areas between 189 m and 225 m, below Lake Algonquin and above Lake Nipissing, have been exposed for 8,000 years, while areas below 189 m have only been free of ice and water for 3,500 - 4,000 years.
Topography and soils are two factors which determine plant distributions. Glaciation has produced a diversity of topographic regions and soil types in the UMBS region. These include erosion resistant ridges on the lakeshore such as Roberts (Bentley), Ingleside, and Grapevine Points. The effects of glaciation also include the deposition of moderate to heavy textured till, coarse-textured outwash sands and gravels, and lakebed deposits.
After the Pleistocene, temperate mesic plant species began their migration northward from their southern glacial refuges. Davis (1981) documented the arrival of various tree species in the Great Lakes region based on pollen stratigraphy of cores from north central and eastern United States lakes. Jack pine arrived 10,500-11,000 years ago, along with white pine and oak. Hemlock arrived 500 to 1,000 years later. Beech, the most recent arrival, is first seen in the pollen record between 5,000 and 6,000 years ago.
Pollen profiles and plant macrofossils also help us document the post-glacial vegetation of this region, most notably work by Sears (1942), Wilson and Potzger (1943), Deevey (1949), Benninghoff and his students (1960), Farrand, Zahner and Benninghoff (1969), Miller and Benninghoff (1969), Andresen (1976), and Futyma (1982). These profiles record a transition from boreal conifers to deciduous tree genera. Initially, spruce was dominant, with small amounts of pine and fir. This spruce dominance was followed by an increase in pine pollen and the first appearance of birch, oak, and tamarack, then by hemlock, beech, and maple. After this pine dominance, conifers decreased to 50% with a concomitant rise in hemlock, birch, oaks, and beech.
THE FOREST THE SETTLERS SAW
Land survey records also are a powerful tool used to reconstruct the presettlement forest. Between 1840 and 1855, before the earliest settlers had arrived, the General Land Office Survey took place in Cheboygan County. In 1957, Paul Kilburn interpreted these land survey records. Using site reconnaissance of the existing vegetation and soil types, Kilburn constructed a Resettlement vegetation map, documented successional trends in the post settlement vegetation and proposed future trends. In order for the surveyors to locate and map section and township corners, the distances and directions to two or four trees were recorded, as well as tree species and diameter at breast height (DBH). These so called witness or bearing trees give us a glimpse of part of the presettlement forest.
Little disturbance characterized the presettlement forest except for that created by native peoples. No lumbering occurred prior to 1840 and little cutting had taken place by 1855. Fire probably occurred only once per century in the Resettlement forest. Based on the presence of fire scars, Kilburn concluded that only five major fires had occurred in Cheboygan County during the lifetime of the trees present in 1957. These fires occurred primarily in dry woodlands dominated by pine.
Kilburn (1957, 1060, 1961) recognized several types of pre settlement forest. On organic soils, the swamp forest was composed of cedar, and to a lesser degree tamarack, or spruce and fir. The northern hardwood forest occurred on the better soils, well-drained loams, silt loams, clay loams, and some mesic sandy loams. On such sites, sugar maple dominated with some basswood, yellow birch, and elm. On drier sandy loams, beech abundance rose to 50% or shared dominance on loamy sands with hemlock. Lastly, Kilburn recognized the pine wood land, dominated by jack, red, and white pine, with lesser representation of hemlock, bigtooth aspen, white oak, red oak, paper birch, and beech.
From 1841 to 1871, the federal government offered land for $1.25 per acre to encourage the extension of the railroad into the forests (O'Neil 1977 cites Inglis 1898). Although labeled "swamp," this was valuable pineland in our area. Due to high yield, much land was cruised, purchased, and set to the axe.
In 1845, Cheboygan County saw its first settler (Voss 1956). By 1870, these numbers exceeded 2,000 and in 1910, the county population peaked at 17,872. By 1900, more than 20% of the county existed as farmlands, mostly situated on loamy soils (Kilburn 1957). Thus, the northern hardwoods were more affected by human settlement.
Human settlement also had a drastic impact on the terrestrial fauna of this area. According to O'Neil (1977), in 1875, 160 acres of land were awarded to Civil War veterans. After a rugged winter in this area, the homesteaders were met by the arrival of millions of passenger pigeons, returning to nest. Colonies were reported to be 30 miles long and three to four miles wide, particularly near Cross Village and Petoskey. The story of the uncontrolled slaughter is well documented and by 1898, Michigan's last passenger pigeon was gone (Schorger 1955).
The years between 1840 and 1900 were the heyday of Michigan lumbering (O'Neil 1977). Between 1847 and 1872, lumbering in creased rapidly; six mills were operating in Cheboygan by 1872 (Voss, 1956). "Pine went first," as Voss (1956) recounts. In general, trees between 15 and 30 inches DBH, from 70 to 160 years old were cut, but smaller trees were lumbered as well (O'Neil 1977). By 1880, all the better pine was gone from Douglas Lake and Pellston. Between 1880 and 1900, most lumbering operations involved only small scale salvaging.
In addition to white and red pine for lumber, hemlock was taken for its bark for the extraction of tannins. By 1900, the conifers were gone and attention turned to the high-grade hardwoods: ash, elm, maple, birch, cedar, spruce, and smaller pines. In 1902, the Tyndall Jackson Mill at Pellston began to lumber the hardwoods west of town. As J. G. Inglis reported in his 1898 Handbook for Travelers: Northern Michigan (cf. O'Neil 1977):
For miles and miles this desolate wilderness of stumps stretches on either side with gaunt bare pine 'stubs' sprinkled among them and decaying logs scattered in wild confusion everywhere. The stubby undergrowth of oak and poplar adds to rather than relieves the desolateness.
The period between 1880 and 1920 was a time of constant fires (Kilburn 1957). Due either to farmers clearing the land or to transients (depending on whether you ask a farmer or a visitor), major fires occurred in 1892, 1901, 1911, and 1923, according to Frank C. Gates. From his examination of fire scars, Kilburn (1957) reported nine fires in 43 years, or a fire at least once every four to five years. These frequent fires kept aspen and other timber down and stimulated the huckleberries and blueberries.
Human impact, trapping, logging, and agriculture also shaped the faunal patterns of distribution and abundance. The fur trade dominated Indian white relations during the seventeenth and eighteenth centuries. Even at this early date, O'Neil (1977) reported that the in coming settlers were pushing out the Indians and destroying the for est. Gates, Clarke, and Harris (1983) summarized changes in wildlife of the Great Lakes states; these changes probably parallel abundance patterns of animals which occurred here in northern Michigan. Gates, Clarke, and Harris concluded that early trapping did not exterminate any wildlife species, however, it did deplete populations of wolverine, fisher, marten, otter, and beaver, the mainstay of the trapping industry. Beavers created frequent forest clearings, elevated water tables and made ponds as catchments for meltwater and eroding soils. The drastic decline in beaver populations must have produced dramatic environmental changes.
The landscape opened with farming and the subsequent growth of aspen. Logging, clearing, and subsequent fires radically changed the microclimates as well as the landscape. Soil and air temperatures increased, and water content of the soils decreased, producing drastic shifts in bird and mammal populations.
By 1900, large mammals (moose, bison, elk) had disappeared due to exploitation by humans for food and fur in addition to habitat changes. Systematic hunting exterminated puma and timberwolf. Black bears were nearly extinguished by the 1920s and 30s, but made a comeback; they are now considered common. Some furbearing mammals having a high population growth (muskrat and mink) survived the logging era, but those with a strongly cyclic and low recruitment rate (marten, fisher, lynx, and beaver) became scarce (Gates, Clarke, and Harris 1983).
Lumbering had opened the forest, improving range conditions for deer. After an initial increase, deer populations plummeted to a mini mum in 1890 due to rampant fires and uncontrolled hunting. The advent of fire control in the 1930s produced an explosion in deer herd numbers. Overbrowsing stunted reproduction of white pine, cedar, hemlock, and yellow birch, favorite foods of deer. As the forest opened, the range of many vertebrates extended northward, improving conditions for the ruffed grouse, sharp-tailed grouse, and prairie chicken (Gates, Clarke, and Harris 1983). The thirteen-lined ground squirrel greatly increased in numbers as hardwoods became established.
Harper (1918) and Gleason (1923) gave vivid accounts of the plant associations during the early years of the Station. On the dry uplands, Harper recognized three communities dependent upon the moisture holding capacity of the soil: (1) hardwoods and hemlock on the mesic sites, (2) white and red pine on intermediate sites, and (3) jack pine on the driest sites. Harper recognized red and white pine as the dominant tree species before the appearance of the lumbermen and hypothesized that, given enough time for the accumulation of humus, the pine forests on the sandy upland might be succeeded by hard woods. Although Harper recognized a second type of succession in our region resulting from fire, he felt that the hardwoods were rarely burned, as Kilburn concluded some 39 years later. Harper (1918, p. 32) summarized the impact of lumbering on the vegetation as follows: the 'pernicious activities' of the lumbermen a few decades ago removed the greater part of these two valuable pines (red and white), and the ground formerly occupied by them has now a low scrubby growth of birch and aspen, which is burned too often for the white pine to reestablish itself, though the red pine is making some headway.
Gleason's eloquent account in 1923 of his botanical observations in northern Michigan recognized the unique nature of the transition forest. He described this so-called tension zone forest as an army of deciduous trees from the south meeting an army of coniferous trees from the north. Gleason blamed the lumbermen for the demise of this unique forest community (p. 273): The armies of both belligerent forces have been sadly decimated by a third and much more powerful force-man, armed with the axe, the circular saw, and the railroad.
Gleason recounts that after lumbering around Douglas Lake, the land was a "jungle of brush heaps . . . dense thickets of aspen, birch, and pin cherry with an almost continuous ground cover of bracken fern." He saw extensive white pine regeneration and growth where a seed source remained and blamed careless fire for the failure of white pine to replace aspen. These fires were frequent; he counted 12 fires in one day during a drive of seven miles. Of the 3,000 acres the Station owned in 1923, Gleason knew of only two small areas that had escaped fire and saw the old aspen trees being replaced by what he called "a thrifty growth of healthy pines" (p.275).
Gleason recognized that aspen germination was favored by fire. He felt that the pine, if protected from fire, could replace the aspen since the poplars are relatively short-lived trees. He speculated that the demise of many aspen after 25 years could be due to the poverty of the sandy soils, crowding, or water stress. The work of Barnes (1966) and his students illustrated the fire adaptedness of this "phoenix tree" in terms of germination characteristics, its stimulated suckering following fire and rapid growth rates. Perhaps what Gleason did not know at the time was that Hypoxylon canker is responsible for the relatively rapid death of certain genetically susceptible clones. We now know that many aspen stands on Station property are approaching 80 years of age and may persist even longer.
In 1960, Kilburn summarized the effect of settlement on the origin al vegetation of the Biological Station as follows: Between 1850 and 1900, pine and hemlock were logged on the upland sandy soils. For 40 years after the original lumbering, fires were frequent, eliminating most of the small pine and practically all of the hemlock. With fire control in the '20s, second growth forests of bigtooth aspen dominated. The logging removed red pine to the advantage of jack pine, which in the face of fire protection gave way to oaks. Kilburn (1960) felt that, given natural restocking of seedlings, the pines would return, with a slower return of hemlock due to the scarcity of seed. In the pine oak forest type, fire merely damaged the oaks, so large oaks dominated in 1957, but Kilburn hypothesized that pines would take at least 200 years to return.
Between 1911 and 1954, Frank C. Gates studied forest succession in the Douglas Lake region. From 1923 to 1954, he and his plant ecology classes made yearly surveys of areas that had experienced fires. He clearcut and burned three adjacent plots in 1936, 1948, and 1954; a fourth burn was added by UMBS researchers in August 1980. In 1938, Gates, W. F. Ramsdell, and L. R. Schoemann set up seven fifty-year forest plots to initiate a study of succession on different soil types (Duncan and Varner 1938). Every plant over one meter in height was mapped in these 0.1 acre plots. Subplots were sampled for herbaceous understory and woody species under one meter in height were noted as well. Under the initiative of Ramsdell in 1934 and S. H. Spurr in 1956, a set of permanent plots was established and monitored to study the effects of thinning and other management on many tree species and forested communities. These permanent plots: aspen sets, natural burns, experimental burns, fifty year plots, and forestry treatment plots, are the jewels in the crown of the Biological Station. They allow us to approach the study of succession using three methods: the analysis of (1) stands of different ages sampled at one point in time, (2) a stand or series of stands measured through time, and (3) communities that are manipulated to experimentally test certain hypotheses about succession. I would like to use several studies which utilized these UMBS permanent plots to illustrate patterns of vegetation change in our region since 1936.
Scheiner and Teeri (1981) analyzed Gates' data that lay dormant for many years following his death. They found successional trends, based on species occurrence, similar to those reported earlier by Whitford (1901) and Gates (1930). Bigtooth aspen dominated in the early years following fire, becoming replaced by white pine and red maple, with a gradual increase in the diversity of species. Scheiner and Teeri hypothesized that this course of succession may be very different from that which produced the primeval woodland or that which occurred immediately following this destruction.
Cooper (1981) took another approach to the analysis of some of the permanent plot data. He concluded that once significant aspen mortality begins to occur, as it has in the seventy-year-old forests at UMBS, aspen standing crop peaks and productivity begins to decline. Standing crop and net annual production of pine and oak-maple woodland, however, increase. Cooper hypothesized that within the next two decades production increases will be confined entirely to pine and oak maple. The species dominance will shift. Where pines are well stocked, they will become dominant. In areas where pines are rare, oaks and maples will reign for an unpredictable period of time before pines finally assume dominance. In areas where pines, oaks, and ma pies are scarce, we are likely to see small patches of aspen dominance.
Sharik, Heinen, and Fenster, with the assistance of recent plant ecology classes, have recensused the burn plots and asked mechanistic questions about white pine regeneration. For all tree species, they observed a constant recruitment into the lower strata and an increase in recruitment over time into the upper levels of the canopy. They observed that it took about 15 years for the 1936 burn to be invaded by white pine, compared to only three to five years for initial establishment in the 1948 and 1956 burns. Interestingly, the 1936 and 1948 burns were invaded in about the same calendar years, 1951 53. Aging of the white pines in the surrounding forest indicated that they were probably reaching reproductive maturity at about this time. Their studies support earlier predictions that logging or comparable forms of disturbance and fire must occur in moderately-aged stands of reproducing white pine for this species to persist. More frequent fires result in reversion to aspen, while less frequent fire may lead to dominance by hardwoods, at least on better sites, a trend seen during the past ten years by Sakai, Roberts, and Jolls (1984). The pines of Hartwick Pines State Park, Grayling, just south of Douglas Lake, are be coming Hartwick Maples, and are an example of this reversion to hardwood without disturbance.
CHANGES IN BIRD POPULATIONS SINCE 1928
The Biological Station is unique in having permanent census areas for birds set up by Sewall Pettingill and Douglas James in 1947 and a small mammal trapping grid established by John Douglas in the late '70s. Pettingill (1974) recounts some of the changes in local birdlife during the maturation of the second-growth forest. In 1928, Common Nighthawks, Hermit Thrushes, Rufous sided Towhees, Black-billed Cuckoos, and Brown Thrashers were common. Today, they are rare and can be found only in cleared areas. The loss of open areas also reduced or eliminated the Eastern Kingbird and Mourning Dove. However, the maturation of pines and shade trees has attracted the Pine Warbler and the Warbling Vireo. Due to other changes, such as the use of pesticides, loss of nest sites, and fluctuating water levels, Pettingill noted a sharp reduction in almost all diurnal birds of prey, the Eastern Bluebird, and shorebirds.
Several species, however, are new to the region: the Mockingbird, Brewer's Blackbird, and the Dickcissel. Still others such as the chimney swift, purple martin, robin, and blue jay have taken advantage of the new habitats created by humans (Gates, Clarke, and Harris 1983).
Mark Roberts, formerly of Duke University, now at University of New Brunswick, used much of the permanent plot data and extensive site reconnaissance as part of his dissertation (1983) to evaluate the successional trends of this region and to offer some insights into successional mechanisms. He concluded that succession is influenced not only by soil type, time, disturbance, and what was there, but also by biotic forces, such as the life history properties of species (their mode of growth, reproduction, longevity) and even browsing and insect defoliation. The extant forces shaping our forests are as heterogeneous as the forces that shaped them: glaciation, soils, lumbering, fire, human settlement, and the composition of the ancestral forests. On drier sites, we can expect young stands of bigtooth aspen, black cherry, red maple, red oak, and serviceberry to give way to red and white pine, red oak, and paper birch. Some semblance of the presettlement forest will return; however, with fire suppression, we will not see the jack pine return to its original dominance on the poorest sites. Due to a lack of seed from its fire sensitivity, we can expect a distinct scarcity of hem lock compared with presettlement communities.
Special thanks to Burton V. Barnes, C. Louis Borie, George M. Briggs, Arthur W. Cooper, Howard A. Crum, David M. Gates, Frank C. Gates, Sandra J. Planisek, Mark W. Roberts, Ann K. Sakai, Samuel M. Scheiner, Terry L. Sharik, Michael Sliva, Robert J. Vande Kopple, Thomas J. van't Hof, James R. Wells, and Gary R. Williams for photographic contributions, technical assistance, editorial comments, and stimulating discussion. The support of the University of Michigan Biological Station Naturalist-Ecologist Training Program, Andrew W. Mellon Foundation, is greatly appreciated.
Andresen, N.A. 1976. Recent diatoms from Douglas Lake, Cheboygan County, Michigan. Dissertation, University of Michigan, Ann Arbor.
Barnes, B.V. 1966. The clonal growth habit of American aspens. Ecology 47: 439-47.
Benninghoff, W.S. 1960. Pollen spectra from bryophytic polsters, Inverness Mud Lake Bog, Cheboygan County, Michigan. Papers of the Michigan Academy of Science, Arts and Letters 45: 41 60.
Clements, F.E. 1916. Plant succession: an analysis of the development of vegetation. Carnegie Institute of Washington Publ. No.242. Washington, DC: Carnegie Institute of Washington.
Clements, F.E. 1936. Nature and structure of climax. Journal of Ecology 24: 252 84.
Cooper, A.W. 1981. Above-ground biomass accumulation and net primary production during the first 70 years of succession in Populus grandidentata stands on poor sites in northern Michigan. p.339-60. In: D.C. West, H.H. Shugart and D.B. Botkin, eds. Forest Succession: Concepts and Applications. New York: Springer Verlag.
Davis, M.B. 1981. Quarternary history and the stability of forest communities. p. 132 53. In: D.C. West, H.H. Shugart and D.B. Botkin, eds. Forest Succession: Concepts and Applications. New York: Springer Verlag.
Deevey, E.S., Jr. 1949. Biogeography of the Pleistocene. Geological Society of America Bulletin 60: 1315-1416.
Futyma, R.P. 1982. Postglacial vegetation of eastern upper Michigan. Dissertation, University of Michigan, Ann Arbor.
Duncan, D.P. and R.W. Varner, 1938. Experiment designed to show the relationship between soil and changes in the vegetational cover. University of Michigan Biological Station, unpubl. ms.
Farrand, W.R., R. Zahner and W.S. Benninghoff. 1969. Cary-Port Huron interstade: evidence from a buried bryophyte bed, Cheboygan County, Michigan. Geological Society of America Special Paper 123 (INQUA Volume): 249-62.
Gates, D.M., C.H.D. Clarke and J.T. Harris. 1983. Wildlife in a changing environment. p.52-80. In: Susan Flader, ed. The Great Lakes Forest: An Environmental and Social History. Minneapolis, Minnesota: The University of Minnesota Press in association with the Forest Historical Society, Inc., Santa Cruz, California.
Gates, F.C. 1930. Aspen association in northern lower Michigan. Botanical Gazette 90: 233-59.
Gates, F.C. 1942. The bogs of northern lower Michigan. Ecological Monographs 12: 214-54.
Gleason, H.A.1923. Botanical observations in northern Michigan. Journal of the New York Botanical Garden 24: 273 83.
Gold, A., J. Gannon and M.W. Paddock, Jr. 1976?. Douglas Lake profile. A report for the National Science Foundation, LSP76.84559.4 p. Available from: The University of Michigan Biological Station, Pellston, Michigan 49769.
Harper, R.M. 1918. The plant population of northern lower Michigan and its environment. Bulletin of the Torrey Botanical Club 45: 23-42.
Inglis, J.G. 1898. Handbook for Travelers: Northern Michigan. Petoskey, Michigan: George E. Sprang. Grand Rapids, Michigan: Black Letter Press Reprint, 1973.
Kilburn, P.D.1957. Historical development and structure of the aspen, jack pine and oak vegetation types on sandy soils in northern lower Michigan. Dissertation, University of Michigan, Ann Arbor.
Kilburn, P.D. 1960. Effect of settlement on the vegetation of the University of Michigan Biological Station. Papers of the Michigan Academy of Science, Arts and Letters (1959) 45:77-81.
Miller, N.G. and W.S. Benninghoff. 1969. Plant fossils from a Cary.Port Huron interstade deposit and their paleoecological interpretation. Geological Society of America Special Paper 123 (INQUA Volume): 225.48.
O'Neil, W.S.1977. This Land and Man: A Historical Look at the Use of Land and Natural Resources in the Inland Water Route Region of Northern Lower Michigan. 21 p. Informational Report No.3. Available from: The University of Michigan Biological Station, Pellston, Michigan 49769.
Pettingill, O.S. 1974. Ornithology at the University of Michigan Biological Station and the Birds of the Region. Special Publication No.1.118 p. Available from Kalamazoo Nature Center,7000 North Westnedge, Kalamazoo, Michigan 49007.
Roberts, M.R.1983. Variability and mechanisms of secondary forest succession in the aspen ecosystems of northern lower Michigan. Dissertation, Duke University, Durham, North Carolina.
Sakai, A.K., M.R. Roberts and C.L. Jolls.1984. Successional changes in a mature aspen forest in northern lower Michigan: 1974.1981. American Midland Naturalist. in press.
Sakai, A.K. and J.H. Sulak. 1984. Four decades of secondary succession in two lowland permanent plots in northern lower Michigan. American Midland Naturalist. in press.
Scheiner, S.M. and J.A. Teeri.1981. A 53-year record of forest succession following fire in northern lower Michigan. Michigan Botanist 20: 3-14.
Schorger, A.W. 1955. The Passenger Pigeon, Its Natural History and Extinction. Madison, Wisconsin: University of Wisconsin Press.
Sears, P.B. 1942. Forest sequences in the north central states. Botanical Gazette 103: 751-61.
Spurr, S.H. and J.H. Zumberge. 1956. Late Pleistocene features of Cheboygan and Emmet counties, Michigan. American Journal of Science 254: 96.109.
Voss, E.G.1956. A history of floristics in the Douglas Lake region (Emmet and Cheboygan counties), Michigan, with an account of rejected records. Denison University, Journal of the Scientific Laboratories 44: 16-75.
West, D.C., H.H. Shugart and D.B. Botkin, eds. 1981. Introduction. p. 1-6. In: Forest Succession: Concepts and Applications. New York: Springer Verlag.
Whitford, H.N.1901. The genetic development of the forests of northern Michigan: a study in physiographic ecology. Botanical Gazette 31: 289.325.
Whittaker, R.H.1953. A consideration of climax theory: the climax as a population and pattern. Ecological Monographs 23: 41:78.
Whittaker, R.H.1973. Climax concepts and recognition. Handbook of Vegetation Science 8: 137.54.
Wilson, l.T. and J.E.T. Potzger.1943. Pollen study of sediments from Douglas Lake, Cheboygan County and Middle Fish Lake, Montmorency County, Michigan. Indiana Academy of Science Proceedings 52: 87-92.