Miller Woods oak savanna
(within Indiana Dunes National Lakeshore)

Location and owner/manager:

Lake County within Gary. National Park Service .

From the west at Interstate 90 (East-West Tollway) exit 17: Go east on Highway 12/20. At about a mile the routes split up – stay on Highway 12 (East Dunes Highway) to the left. Go another mile east on Highway 12 to Lake Street and turn left. Go north on Lake Street two thirds of a mile and turn left into a small parking area at the Miller Woods trailhead. The trailhead is immediately north of the Paul Douglas Center for Environmental Education pedestrian bridge across Lake Street.

From the east at Interstate 94 exit 19: Go north on Highway 249 one mile to the where the highway bends east and turn left on Midwest Steel Highway. Go north on Midwest Steel Highway and immediately exit to westbound Highway 12 (which involves a left turn from the south). Go west on Highway 12 five miles to Lake Street and turn right. Go north on Lake Street two thirds of a mile and turn left into a small parking area at the Miller Woods trailhead. The trailhead is immediately north of the Paul Douglas Center for Environmental Education pedestrian bridge across Lake Street.

From the southeast at Interstate 80/94 exit 15: Go north on Ripley Street two thirds of a mile to Highway 20 (Melton Road) and turn left. Go west on Highway 20 one and one third of a mile to Lake Street and turn right. Go north on Lake Street three quarters of a mile and turn left into a small parking area at the Miller Woods trailhead. The trailhead is immediately north of the Paul Douglas Center for Environmental Education pedestrian bridge across Lake Street.

Note: The Miller commuter rail station is about a tenth of a mile west of the intersection of Highway 12 and Lake Street. Miller Woods is a little less than a mile from here via Lake Street.

Free. No camping (but exceptionally nice camping is available elsewhere within Indiana Dunes National Lakeshore). Pets on leash.


In the beginning there was a bit of beach grass. The wind blew, grains of sand collected around the beach grass and eventually great towering dunes developed along with a new branch of science. So begins the story of ecology.

Ecology – at least the part pertaining to plants – began along the beach in Denmark. Eugenius Warming developed the idea of plant communities during his travels abroad and then studied the plant communities of the beach dunes in his native land. His research showed a distinct progression of plant communities from which the concept of plant succession was first developed.

The scientific community at first barely took notice of plant ecology nor was it entirely accepted as much of a science. Evolution and natural selection when introduced in the mid-19th century sparked a scientific revolution - the debut of ecology and plant succession a few decades later evoked what may be described metaphorically as a collective yawn. However several botanists at the University of Chicago were inspired. In particular the graduate student Henry Cowles began to study the dunes around Lake Michigan and then other plant communities in the Chicago area. His studies of the Lake Michigan dunes culminated in a doctoral dissertation published in 1899 that expanded upon Warming’s earlier work. A few years later he would show that plant succession far from being a unique feature confined to the continually disturbed ecology of a dune complex was an almost universal feature of all landforms with plants. About the same time Frederic Clements conducted field surveys in Nebraska and independently arrived at a concept of plant succession that differed from that of Warming and Cowles. Clements viewed plant societies as broad abstractions moving in the direction of a final association that was optimized for the conditions of a given climate. Cowles on the first page of the introduction to his dissertation made it clear that he viewed all plant associations to be loose – ‘a panorama never twice alike’ (Cowles, Henry Chandler. 1899. The ecological relations of the vegetation on the sand dunes of Lake Michigan. Chicago: The University of Chicago Press).

Miller Woods – which today is usually referred to generically as a Black Oak savanna (and sometimes as a barrens) although there are also wetlands here – was scrutinized by Cowles and is a fine example of a mature plant community within a south shore Lake Michigan dune complex. It is in a heavily industrialized corridor with rail road tracks to the south, a steel mill to the west and for many years was an unofficial dumping ground. The surrounding development may partly account for its survival for in a large active dune complex it might very well have been buried by a series of wandering dunes. There are now no large wandering dunes along this part of Lake Michigan save for the tallest dune of all – Mount Baldy – which the National Park Service continually struggles to keep in its place. Miller Woods is relatively open with many plants usually associated with the prairie thriving in an understory that gets only partial shade.

Henry Cowles’ study of the Lake Michigan dunes although published in the 19th century is well worth examining for those interested in grassland ecology. Many plants of the Great Plains make appearances at various points in the plant succession he outlines in ways that you might not entirely expect. It is also a well-written classic piece of research – a model for how to present significant ideas simply. I will present a lengthy summary of it here before attempting to tie it in with what a visitor might encounter on a visit to Miller Woods.

The most important pioneer plant of the Lake Michigan sand dunes identified by Cowles has a common name of Marram Grass and a generic name of Ammophila which appropriately enough means ‘sand lover’ in Greek. At the time of Cowles’ research American and European Marram Grasses were considered the same species. Today American Marram grass Ammophila breviligulata is distinguished from European Marram grass Ammophila arenaria by its shorter ligule - hence its specific epithet ‘breviligulata’ - and a taller inflorescence. From a land management standpoint European Marram Grass is also considered a weedy invasive in many North American localities. Both species work along the same general principles. The leaves of Marram Grass have a tubular shape. The edges curl inward which serves to minimize the drying effects of the wind and sun. Thick rhizomes grow vigorously both vertically and horizontally forming strong intertwining mats. The leaf nodes if covered by shifting sand can grow roots and sometimes they grow roots that are suspended in air and collect blowing sand – in effect the beach grows around the beach grass. The growth habit of Marram Grass and a steady supply of windblown sand together create a process that enables an embryonic sand dune to grow outward and upward.

Cowles makes several important points here regarding embryonic sand dunes. First, where the wind often shifts directions as it does along Lake Michigan sand and wind can not by themselves form steep towering dunes – ‘one wind destroys the products of another wind’ (Cowles, p176). They can form small mounds but these small mounds are destroyed when the wind shifts even a little bit. For the small mounds to become large dunes an obstacle to the wind is required that increases in size. A dune forming plant provides this requirement of a growing obstacle. A dune forming plant also must be a perennial as otherwise the dune would be destroyed at the end of one or two growing seasons. Once an embryonic sand dune forms its ‘growth increment is often greater during the later stages of a continuous wind current’ (p176). The growing sand dune itself becomes an even greater obstacle to the wind than the dune forming plant that often gets buried.

Marram Grass is the most important dune forming plant but not the only one by any means. Along Lake Michigan’s northern beaches Montana or Griffith’s Wheatgrass Elymus albicans (formerly Agropyrum dasystachyum) has a similar growth habit as Marram Grass and is almost as common. Two other grasses - Canada Wildrye Elymus canadensis and Prairie Sandreed Calamovilfa longifolia (formerly Calamagrostis longifolia) - are also minor dune forming plants. Several willows and the Sand Cherry Prunus pumila are nearly as important dune formers as the grasses. Finally a tree – the Plains Cottonwood Populus deltoides ssp. monilifera – is responsible for forming the tallest dunes along Lake Michigan. Here as a side note to the dune building process around Lake Michigan Cowles distinguishes between the ‘impoverished’ flora of the south end of the lake and the ‘luxuriant’ flora of the north. The north is a bit cooler and the plant societies there more diverse but in Cowles’ opinion more than just climate accounts for the difference. The land around the north end of the lake is more heavily forested resulting in more driftwood along the beach which in turn provides more protection for the germinating seeds of a wider variety of plants (p172).

The size and shape of the Lake Michigan sand dunes reflect the growth habit of the plants around which the dunes form. Montana Wheatgrass tends to form low spreading dunes because the roots of the plant interlock and spread laterally but the plant can not grow upwards as quickly as most of the other dune forming plants. Prunus dunes are often small and steep – the plant can not spread but grows quickly upward. Salix (willow) dunes are larger than the dunes that form around the grasses because willows have an even greater power of upward growth. Cottonwood dunes are the highest and steepest but do not spread. The plasticity of Cottonwoods is remarkable in Cowles’ opinion – a tree of the river bottoms is also at home in the dunes. The Cottonwood is the most vigorous dune forming plant but has trouble reproducing due to a lack of vegetative propagation and poor seed germination. Cowles (p185) generalizes the role of plants in dune formation as follows: ‘Summing up this matter, then, the area of a beach dune is determined chiefly by the amount of the vegetative propagation of its tenants, the steepness is determined by the rapidity of this propagation, and the altitude by the power of the dune plants to increase in height.’

The process that forms a sand dune is one that increases in power as the dune forms. The plants often lose as the sand builds up at an increasing rate and the growing parts of the plant move further from the water table. Another stage in plant succession begins as the ‘dune former’ plants are replaced by ‘dune holders’ and a few annuals. The annuals at this stage of the process germinate early in the spring, grow rapidly and then set typically winged seeds – Cowles thought the best example of such an annual to be Common Bugseed. (The Linnean name of which was Corispermum hyssopifolium when Cowles wrote his paper. This tumbleweed which was once regarded as a single plant widely distributed across much of the northern hemisphere has since been apportioned within a somewhat chaotic genus of species and hybrids). However well adapted these ‘dune holder’ plants might be to growing among shifting sands they often get covered up as the dunes make a slow march inland.

Eventually the dunes in an active sand dune complex move far enough inland that they are subject to less wind. Cowles (p383) describes such a dune complex as ‘a restless maze, advancing in one direction, but with individual portions advancing in all directions.’ The moving dunes slow down and vegetation captures their lee side. Sometimes the plants that capture the lee side of a dune are the same plants that formed the dune sometimes they are plants from farther inland. In Cowles’ opinion (p302) it is not a uniform process nor does it entirely matter: ‘So far as the capture of the dune is concerned it is a matter of no moment whether the vegetation is antecedent or subsequent. All contribute to the common end.’

Regardless of which plants initially capture a wandering dune two distinct plant communities usually develop on the dune’s lee and windward sides respectively. On the lee side of a Lake Michigan dune a mesophytic (‘moderately moist’) flora of trees, shrubs and vines grows into a basswood and maple forest in the north and sometimes an oak forest in the south. On the windward side a wide variety of low shrubby vegetation develops and moves in the direction of becoming a coniferous forest – or as Henry Cowles so eloquently summarizes it ‘…the heath has several origins but only one destiny’ (p367).

A couple of secondary processes are also at work as the dunes move inland and are captured. Sometimes a large dune moves and exposes a fossilized beach. Other times a blast of wind causes a patch of exposed sand known as a blowout to develop. The exposed areas have similar conditions to that of the beach along the Lake Michigan shore but with no breaking waves and less consistent wind. Blowouts (which tend to be closer to the shore than fossilized beaches) serve as protected areas where a wide variety of seeds can germinate – in particular the Cottonwood which creates the tallest sand dunes often gets a start here. Within newly exposed fossilized beaches slow-growing embryonic sand dunes form but with an unusual twist. The grasses that serve to create embryonic sand dunes along the shoreline of Lake Michigan are all grasses that can be found elsewhere in the prairies of the Midwest and Great Plains but seldom as dominants. However the grass that typically forms embryonic sand dunes in open areas vacated by a wandering dune along Lake Michigan is Little Bluestem (Schizachyrium scoparium formerly Andropogon scoparius) which is one of the most widely distributed native grasses within North America. Little Bluestem has a stiff tuft growing habit – the dunes that form around it are often small and steep. Cowles considered these Little Bluestem dunes (or Schizachyrium dunes if you insist) to be similar to the geological roches moutonnees. (Roches moutonnees are mounds formed underneath advancing glaciers where the upstream or stoss side is often a gentle slope of bare rock while the ‘lee’ side is more steep and ragged – examples are in New York City’s Central Park). Since Cowles’ research Schizachyrium scoparium has been studied in much more detail and while it does not appear to be much of a pioneer plant it is a particularly persistent grass well suited to being buried for years by a large sand dune. Its success in areas repeatedly covered and exposed by wandering dunes may be due less to its being an opportunistic invader and more to its being a tough survivor. The roots of Little Bluestem can go dormant for long stretches under drought conditions while the seeds of this grass can germinate decades after being buried in dry soil. In addition to Little Bluestem Cowles considered three other plants to be characteristic of Lake Michigan fossilized beaches: Beach Wormwood (a short lived perennial sagewort with a deep tap root that now goes by the name of Artemisia campestris ssp. caudata), Dune or Sticky Goldenrod (which has the not so simple scientific name of Solidago simplex ssp. randii var. Gillmanii) and Rough Puccoon (Lithospermum caroliniense formerly Lithospermum hirtum). The plants of a Lake Michigan fossilized beach with their light grey green foliage and yellow flowers would appear to be an association that would not be entirely out of place in the ‘sagebrush sea’ of Wyoming. However this association is not as stable as similar plant associations in the west. At the front end of the plant succession of a newly exposed fossilized beach an active sand dune complex is required – these have become increasingly rare as the beachfront property of Lake Michigan has been developed over the years. At the backend of this process as described by Henry Cowles the fossilized beach will eventually be replaced by a heath and then a coniferous forest. A newly exposed fossilized beach has become an increasing rare ecological niche along Lake Michigan and one of its associated plants - Sticky Goldenrod – is now listed as ‘endangered’ at the State of Indiana level.

The final stage of plant succession on the leeward side of the dunes of Lake Michigan varies with latitude. In the north a river bottom flora of basswood and maples develop while in the south a more open and xeric flora of oak dunes establishes itself. Henry Cowles in 1899 was not completely sure why this was the case – the ‘condition for the origin and development of an oak dune flora are obscure’ (p380). However he noted a few clues. First - remnants of basswood flora were occasionally present within oak dunes midway up the eastern Lake Michigan shore implying that the basswood flora developed there before the oak flora. Second Cowles observed that at one locality on the south shore pines had burned and oaks grew in their place. He concluded that pines were more susceptible to fire than oaks. In a few places he noted that pines grew at the lowest and highest areas of some dunes with oaks in between. From this he speculated that the pines were able to withstand wider extremes of growing conditions but ‘where oaks can be at all they seem to drive out the pines’ (p381). His educated guess in 1899 of the role of fire in the establishment of an oak dune was a good one and remarkably ahead of its time.

There are several types of savanna in the Midwest. Miller Woods is dominated by Black Oak which grows an unusually deep tap root and tolerates nutrient poor sandy soil. (Cowles refers to the obsolete Quercus coccinea tinctoria in his 1899 paper. The taxonomy of North American oaks since then has been considerably revised and the Black Oak that is found along the south shore of Lake Michigan now goes by the scientific name of Quercus velutina – the ‘velvety oak.’). Other Midwest savannas are dominated by Bur or White Oak which usually are to be found growing in a richer wetter soil with more clay. These savannas are often presented as contact areas between the deciduous forest of the east and the grasslands of the mid-continent – the implication here is that the fire tolerant oaks are gradually reclaiming the turf surrendered to the grasses during drier climatic conditions. Another possibility is that in some cases the oaks lead the way for the establishment of an understory of prairie grasses rather than create conditions that drive them out. Over a century later one can reasonably supplement Henry Cowles‘ observations by noting that while fire was likely even more important to the maintenance of the oak dunes than he initially thought widely-spaced Black Oak also have a tendency to reinforce the nutrient poor soil they often grow in.

A fine way to see a Black Oak savanna for yourself is to take a walk at Miller Woods. (There is also Black Oak savanna a few miles east at Howe’s Prairie that was extensively examined by Henry Cowles but it is not quite as accessible as Miller Woods). The Miller Woods trail is an elongated one mile loop with a connecting boardwalk midway that crosses between two interdunal ponds. A walk around the outer loop at a leisurely pace takes about a half hour – much longer if you start looking for some of the rare plants hidden here.

The understory of plants at Miller Woods while having an overall grassy aspect does not really resemble dense tallgrass prairies often found in the Chicago metro area. Henry Cowles described the understory of the oak dunes as a ‘tufted vegetation with intervening patches of naked sand’ (p379). His long list of plants of the Oak Dunes (pp 380-381) has only two grasses – Junegrass (Koeleria cristata now Koeleria macrantha) and Sixweeks Fescue (Festuca tenella now Vulpia octoflora) but a large number of sedges of which he judged Carex pensylvanica to be particularly abundant. In addition to sedges and cool season grasses one can find today many of the warm season grasses of the prairie at Miller Woods although they are not dominants here. Most likely this is because since 1899 Miller Woods unlike most parcels of land along Lake Michigan has been burned on a steady basis either intentionally or accidentally. You are not likely to find in this understory much Marram Grass the pioneer plant of the Lake Michigan shoreline although Lake Michigan is less than a mile away. In the many decades since Cowles’ 1899 paper was published quite a bit of research with respect to nitrogen and nematodes in sand dune ecology has been conducted and it appears that native Marram Grass is quite vulnerable to the organisms that live in all but the most sterile sandy soil. (Marram Grass and other early successional plants may be readily found at nearby West Beach). Many of the prairie forbs at Miller Woods bloom on a slightly delayed schedule compared to the same plants growing elsewhere in the upper Midwest. This is due to a lake effect. The plants at Miller Woods are often just getting started in late spring at about the time when many of the hill prairies of Illinois are in full bloom. The early bloomers at Miller Woods include the cool season grasses, Common Spiderwort (Tradescantia ohiensis), Lupine (Lupinus perennis) , Bird’s Foot Violet (Viola pedata), and Eastern Prickly Pear (Opuntia humifusa). By mid-summer various warm season grasses, sunflowers and asters bloom along with the occasional bee balm or Indian Paintbrush (Castilleja coccinea). In early fall you may see various goldenrods, blue gentians (both Gentiana andrewsii and Gentianopsis crinita) or Nodding Ladies Tresses (Spiranthes cernua). One plant that is often found growing next to the plants of the prairie here and that would be seemingly out of place is Bracken Fern (Pteridium aquilinum latiuscuum). However quite a few prairies and savannas of northwest Indiana also harbor this plant. Within the Indiana Dunes National Lakeshore it is perfectly possible to see a fern growing next to a cactus and indeed there is a remarkably large number of plant families from the desert to the boreal forest often growing here in close proximity.

Several biological surveys have been done over the years at Miller Woods. A survey conducted before the trash were removed at Miller Woods (Wilhelm, G.S. 1990 Special vegetation of the Indiana Dunes National Lakeshore. Research Program Report 90-02. Indiana Dunes National Lakeshore, Porter, Indiana) found impressive botanic diversity here. For example – there are seventeen species of sedges and fifteen species or subspecies of goldenrod. Among warm season grasses the comparison to other Midwest prairies and Cowles’ description of the flora of the oak dunes is particularly striking. Big Bluestem, Little Bluestem, Indiangrass and Switchgrass can all be found here. At most prairies in the Midwest you would be hard pressed to identify more than a dozen species of warm season grasses. At Miller Woods there are fourteen species from the genus Panicum alone.

In addition to the many plants at Miller Woods it is worth noting the presence of one particular insect and two species of squirrels. The rare insect here is the Karner Blue (Lycaeides melissa samuelis) an endangered butterfly that is near its southernmost limit at Miller Woods. Lupinus perennis – the plant that the larva of this particular butterfly prefer - is abundant at Miller Woods and the population of the butterfly itself is healthy but isolated. The two species of squirrels that can be found at Miller Woods are the Gray Squirrel (Sciurus carolinensis) and Red Squirrel (Tamiasciurus hudsonicus). The significance of their presence here at a place where the oaks have been gradually replacing the pines is that Gray Squirrels feed on and incidentally disperse acorns while Red Squirrels do the same for pine cones. (A survey of the mammals of Miller Woods can be found in Whitman, R.L., R.L. Peloquin & R.J. Worth. 1990. Ecology of Miller Woods. Indiana Dunes National Lakeshore. USDI, NPS, Indiana Dunes Report 90-01).

Finally, in wrapping up this discussion of the first major study in American plant ecology and a place that helped inspire it, permit me to slip in a personal observation that perhaps plant succession and the progression of science do not always follow a rational sequence. The plant succession of the Lake Michigan dunes runs counter to the plant succession at many other places at mid-latitudes. Often a parcel of disturbed soil such as a plowed but unplanted field or a roadside ditch is invaded first by weedy annuals then by succeeding waves of perennials each of which create the conditions for their replacement. In the plant succession of the Lake Michigan shoreline the pioneer plants are perennials, the weedy annuals reach their peak in the middle and the initial conditions of bare sandy soil can seemingly reoccur at anytime without the vegetation necessarily going all the way back. The middle stage in particular would seem to be intuitively out of place. Likewise the history of science in the 19th Century with respect to evolution and ecology would appear not to have followed a logical sequence. Natural selection was formulated decades before plant succession was first proposed yet it is by far a more radical and abstract concept. Darwin agonized for years considering the many possible objections to natural selection before finally submitting it as a published theory. The concept of plant succession on the other hand does not require any great leap of imagination rather it involves recognizing the hidden significance of what is ordinary. Small examples of plant succession such as a pond or fallow field eventually becoming a forest should have been obvious to owners of great landed estates since antiquity yet no such owner ever took notice or worked out a generalized concept. That ecosystem of contentious neurons known as the human mind appears to have evolved an instinctive fascination for that which is elusive and abstract.