Using Insights from the New England and Scandinavian landscapes to solve the “mystery” of the Mima Mounds (Thurston County, Washington). 

Keywords: Mima Mounds, natural disturbance, microtopography, landscape history.

   Ever since Charles Wilkes led the United States Exploring Expedition through “the level prairie” near what is today Olympia, Washington, the Mima Mounds have provoked curiosity, speculation, debate, and disagreement, all without giving rise to a broadly accepted solution to the question of their origin (Wilkes, 1845).  Hypotheses have ranged from Wilkes’s improbable suggestion of construction by natives of the region, to the more widely accepted fossorial rodents (Dalquest and Scheffer, 1942), to the more recent sorting by seismic vibration (Berg, 1989).  Dozens of other theories have been advanced over the years; a summary and discussion of which can be found in Washburn (1988).  The excessive passions the mounds have aroused over the years, in the form of a torrent of letters in response to any mention of the subject, drove one editor, F. L. Campbell of the Scientific Monthly, to declare in 1948 the discussion “concluded” and to request his readers stop writing the journal and to instead correspond directly with the authors involved (F. L. Campbell, 1948).

    This paper takes a different, less speculative approach.  It proposes that a solution may be found by examining a well known but overlooked cause of mound formation and then extending the potential geographic, temporal, and biotic contexts of mound formation to include forested landscapes.  It engages in what D. R. Foster calls a process of “thoughtful consideration of nature’s history” (2000).  By doing so, a hitherto rejected causal process emerges as a simple and elegant solution (Farnsworth, 1906, M. R. Campbell, 1906).  The Mima Mounds are now regarded, perhaps improperly, as the “type mounds” for many different mounded landscapes.  This paper confines its discuss to the type mounds themselves, and not others that may or may not have resulted from similar processes.

    About the only thing most researchers have been able to agree on is the location and rough physical description of the appearance of the Mima Mounds.  Their physical characteristics, dimensions, range, and density remain well described by Bretz (1913).  They are located in the southwest corner of Thurston County, Washington, on the gravely outwash plain of the Vashon Glacier.  See detail of map, illustration 1.  They originally covered about 30,000 acres before being regraded and leveled for farming, roads, and housing development.  Their density ranges from 17-30 per acre, for a what was once a total of perhaps 900,000 mounds.  They are round or elliptical in plan, with a profile of a roughly spherical section.  Their height ranges from a barely noticeable rise to 7 feet.  See illustration 2, Bretz diagram. The mounds are relatively uniform in shape and apparently random in distribution.  See illustrations 3 and 4, aerial photographs.  They resemble other mounds elsewhere in North America and around the world.   The mounds are now mostly covered by an assemblage of [fire-tolerant] prairie grasses and forbs ?? and the prairie is bordered by Western hemlock - Douglas fir second-growth forest (source for plants??).  The mounds are composed of unstratified soil; within many mounds there is a layer of darker soil with organic matter, including charcoal, on top of gravel bed, with dark intrusions extending downward; there are some small boulders mixed in.  See illustrations 5 and 6, cross sections of mounds.  The mounds are unlikely to be older than the retreat of the Vashon glacier, 14,000 years before present, or younger than the earliest carbon 14 date from soil on the mounds, 4500 years before present (Washburn, 1988; other source for Carbon 14 date??).

    This paper proposes to get past the deadlock in comprehending the mounds’ origin by proposing a radical cognitive shift:  abandoning the too-narrow definition of the surrounding Biome as a “prairie,” an assumption no one has questioned since Wilkes first employed the term in his Report (1845).  Although always described as prairie, this paper proposes that the known time scale, less than 170 years, is too short to generalize the full range of the biome’s potential.  There no reason to believe that this or any other prairie plant community is static, namely that it has always been a prairie in the floristic sense since the Vashon glacier receded.  Research into prairie landscapes elsewhere lead us to question the assumption of prairie biota as a constant since the beginning of the Holocene.  However, at least since Omer Stewart’s work half a century ago, we have suspected that at least the extent of the Great Plains in North America is a result of human-caused fire regimes (Stewart, 1956; Pyne, 1982; Flannery, 2001).  Charcoal in soil samples and recent invasions by trees give reason to believe that the Mima prairie are also human-caused.  In fact, since fire suppression began over 100 years ago, the invasion (colonization?) by Douglas fir challenges the assumption of having always been a prairie.  Arboreal pollen further tests assumptions about the unchanging nature of the prairie landscape.  Yet, all but one of the nearly 100 articles consulted for this paper take the unchanging prairie biota as a given.

    Local fossil arboreal pollen and layers of charcoal from soil cores suggest that the Mima Prairie was likely to have been a different community and that one reason for its current prairie is periodic burning.  See pollen profiles in illustrations 7 and 18.  The sequence of pollen shows succession of boreal conifers (Picea and Pinus), followed by Douglas fir (Pseudotsuga menziesii), Western Hemlock (Tsuga plicata), western red cedars (Cupressaceae), and opportunistic hardwoods such as Alders (Alnus), [common name?? (Betula), and poplars (Populus).  At all times is a significant quantity of non-arboreal pollen such as and prairie grasses (Gramineae), bracken (Pteridium), and camas (Camas) (Boyd and Leopold, 1999, 142), which increase after the Mount Mazama eruption of 5800 years BP (cite).  From the pollen sequence it is easy to imagine huge, widely spaced trees with grasses, bracken, and camas growing between them.

    Furthermore, recent conservation management discussions center on what to do about the “encroachment” of trees onto the prairie, now that fire is suppressed, suggesting if nothing else, that the “prairie” landscape is still quite hospitable to “invasion” by forest species (Washburn, 1956; Kruckeberg, 1991).  Notice encroaching trees in illustration 9.

    By removing the conceptual barrier of requiring the mound-building processes to take place in a “prairie” biome, and limiting potential processes to those geologic events typical of prairies, or species endemic to prairie landscapes, we can widen the scope of research to include other processes, especially those well known to cause mound building, even those occur in a forested landscape.  As it turns out, there is a well known and well documented processes by which mounds are built in a forested landscape:  the upheaval of root balls when wind knocks down and uproots trees (Foster and Boose, 1992; Wessels, 1997; Boose, Chamberlain, & Foster, 2001).  The phenomenon of wind-thrown trees as a geologic agent is especially well studied in New England and Scandinavia (Troedsson, and Lyford, 1973).  The sizes of mounds created in these regions are in rough proportion to the size of the trees uprooted since the root ball is in proportion to the rest of the tree and most recorded instances are in secondary forests, none of the root balls or resulting mounds is anywhere near the scale of the Mima mounds.  As the roots decay the size of the mound decreases and becomes more regular.  Erosion further reduces the size of mounds by washing away fine particles.  See illustration 10 and 11 of windthrown stumps and resulting mounds in a New England forest.

    Such observations of analogous land forms in forested regions raise two questions:  could there have been trees large enough to cause Mima-scale mounds, up to 7 feet high and 40 feet ?? in diameter?  Could there have been winds in the region strong enough to topple and uproot such trees?  The answers to both questions lie within easy reach.  As noted above, pollen samples from soil cores reveal the Western Hemlock- Douglas Fir forest community typical of western Washington.  Specimens of trees with diameters easily exceeding 3 m [set sizes to metric or English but not both] exist not only in the temperate rainforest of the Olympic Peninsula, but also in protected areas in Thurston County [check location of State forest; cite Jerry Franklin’s booklet].

    Historical records of storms, especially the 29 January 1921 “Olympic Blowdown” x? miles west of  the Mima Mounds, [and the Lincoln Day ?? blowdown event in the 1960s??] demonstrate winds of adequate force (and over comparable or greater area) have visited the region twice in the twentieth century and that even the largest old-growth specimens of these trees are vulnerable to being blown down.

    Photographs of damage from the 1921 storm show rootballs several times taller, but of lesser diameter than the Mima mounds.  This is as one would expect if the woody components of the rootballs were subject decay and the soil to erosion.  See illustration 12 of uprooted stump on Olympic peninsula.  If the Mima mounds were the result of a blowdown of an old growth Hemlock-Fir forest at some time between the last glaciation and the present, one would expect to find the soil within the mounds to be disturbed to a greater extent than the surrounding soil and even after several thousand years, one would expect some at least some traces of the biomass that a forested landscape would have once contained.  Geologic analysis of the mounds and surrounding soil have found both unstratified structure and organic material (Washburn, 1988).  Furthermore, decayed roots would help explain the dark “rootlike intrusions” that have not only puzzled researchers since Bretz described them (Bretz, 1913); but which have also leant credibility to the fossorial rodent hypothesis.  (Even Scheffer 1947 called them “mound roots” according to Pewe, 1948).  See illustration 13 of dark, root-like intrusion.

    Based on existing pollen data, this paper argues that some time between 5000 and 8000 years before present, a severe wind event blew down part of an old growth forest, uprooting thousands of trees (Washburn, 1988; Kruckeberg, 1991; Boyd and Leopold, 1999).  Since then, a combination of erosion and decay reduced the root balls to their current shape and size and that periodic fires, perhaps human-induced, have kept the forest species from re-colonizing the landscape.  This hypothesis is not new.  In fact, it was first suggested in a letter to Science in 1906 by Iowa physician P. J. Farnsworth.  He had been following an ongoing discussion of similar mounds in Texas and wrote to offer a hypothesis based on his own observations of windblown trees in his family’s farm in Vermont (Farnsworth, 1906).  Since then, only one other writer has cited him, only to dismiss his argument along with 11 others (Campbell, 1906).  Dr. Farnsworth, used the same tools of observation that D. R. Foster, advocated in 2000.  Perhaps it is no accident that they both knew forests in New England, where landscapes have their longest recorded history in the United States; likewise, some of the difficulty in finding an acceptable cause for the Mima Mounds could stem from their being in the part of the country with the shortest recorded history.

    How have so many capable scientists working for over a century and a half failed to consider the uprooting of giant trees as a possible cause?  Three factors emerge above all others:  They shared the assumption that the landscape’s prairie flora remained static and constant for the past 10,000 years, an assumption that limited researchers to considering processes that could occur in a prairie, or to peri-glacial events concurrent with the prairie’s formation.  Second, because the assumed context was a prairie, most investigators came from fields within geology that dealt processes that could form or have an impact upon prairie landscapes. The assumption of a prairie landscape furthermore limited the consideration of biological agents to those which associated with prairies, namely the pocket gophers that so dominated the discussion since Dalquest and Scheffer’s posited hypothesis (Dalquest and Scheffer, 1942).  Having ruled out other biotic communities, no one but Farnsworth thought to consult specialists in forested landscapes.  The third and final reason is contingency: no one familiar with the pit and mound microtopography of New England, Scandinavia, or other forested regions happened to make the connection to the much larger mounds of Thurston County.  The third factor bears out Foster’s assertion that “although we are geographically and culturally separated and frequently do not draw adequately on each other’s literature, scientists in every  can benefit greatly by sharing and learning from each other’s studies in historical-geography” (Foster, 2002).

Further Research:
Questions for future research include:  could a high water table or hard pan below the soil surface limited depth of roots, thereby increasing vulnerability to blowdown?  Analysis of plant material using Henry and Swan’s methods (1974).  Careful re-examination of pollen record and new samples.  Statistical comparison of tree and mound distribution patterns.  Refining environmental chronology.

James R. Karr for advising me to begin with the known.  Christina Anderson, for sharing the initial “aha” moment with me after reading Tom Wessels’s book together.  Britton Pettit, now State forester with the Washington State revenue department, for confirming that the density patterns fell within the range of what would be expected in an old-growth forest in western Washington; State Geologist Pat Pringle for listening attentively and encouraging me to put my thoughts to paper; Jerry Franklin for the stem maps of old-growth forest; Dave Knowblach for photographs, citations, and encouragement; Dana O’Day Senior for lending her geologic insights to my ongoing discussion of the subject; Kim Limond, the reference librarian at Clinton, Iowa, for biographical information about P. J. Farnsworth.  Tom Wessels, Arthur Kruckeberg, and the members of the Columbia History of Science Group for patiently listening while reserving judgement.  All errors, of course, remain my own. 

Copyright 2002-2003
Michael KucherUniversity of Washington
1900 Commerce St
Tacoma WA 98402

    Berg, Andrew (1989) Formation of mima mounds–a seismic hypothesis.  Geology 18, 281-284.

    Boose, E.R., Chamberlain, K.E. & Foster, D.R. (2001) Landscape and regional impacts of hurricanes in New England.  Ecological Monographs, 71, 27-48.

    Boyd and Leopold (1999)

    Bretz, J. Harlan (1913) Glaciation of the Puget Sound Region in Washington Geological Society Bulletin 8: 9.

    [Campbell, F. L.] (1948) Remarks on the Mima Mounds [ appended to Troy L. Pewe,’s article in] Scientific Monthly 66, 296.

    Campbell, Marius R. (1906–check date??) Natural mounds  Journal of Geology 14, 715.

    Dalquest, W. W., and Victor B. Scheffer (1942) The origin of the Mima mounds of western Washington, Journal of Geology 50, 81

    Farnsworth, P. J. (1906) On the Origin of the Small Mounds of the Lower Mississippi Valley and Texas  Science 23:589 13 April 1906, 583-584. 

    Flannery, Tim (2001)  The Eternal Frontier:  an Ecological History of North America and its Peoples. Atlantic Monthly Press, New York.

    Foster, D. R.  (2000)  From bobolinks to bears:  interjecting geographic history into ecological studies, environmental interpretation, and conservation planning.  Journal of Biogeography 27, 27-30.

    Foster, D. R.  (2002) Insights from historical geography to ecology and conservation:  lessons from the New England landscape.  Journal of Biogeography 29, 1269-1275.

    Foster D. R. and Boose, 1992.

    Franklin, Jerry, et al. (1981)  Ecological characteristics of old-growth Douglas-fir forests USDA General technical report PNW ; 118

    Henry, J. D. and J. M. A. Swan (1974) Reconstructing forest history from live and dead plant material– an approach to the study of forest succession in southwest New Hampshire.  Ecology 55, 772-783.

    Kruckeberg, Arthur (1991)  A natural history of Puget Sound Country.  University of Washington Press, Seattle.

    Limond, Kim.  (25 Nov 2002) Reference librarian at Clinton, IA personal communication

    Pettit, Britton (2001) (personal communication)

    Pringle, Patrick (2001), geologist with Washington Department of Natural Resources, personal communication.

    Pyne, Stephen (1982) Fire in America:  a cultural history of wildland and rural fire.  University of Washington Press, Seattle.

    Scheffer, Victor (1947) The mystery of the Mima mounds.  Scientific Monthly 65, 283-294.

    Stewart, Omer (1956) Fire as the first great force employed by man, 115-133 in  Man’s role in changing the face of the Earth, ed. William L. Thomas.  University of Chicago Press, Chicago.

    Troedsson, T. and W. H.  Lyford, (1973) Biological disturbance and small-scale spatial variations in a forested soil near Garpenberg, Sweden. Studia Forestalia Suecica, 109, 1-23.

    Washburn, A. L. (1988) Mima Mounds:  an evaluation of proposed origins with special reference to the Puget lowlands.

    Wessels, Tom (1997)  Reading the forested landscape:  a natural history of New England.  Countryman Press, Woodstock, Vermont.

    Wilkes, Charles (1845) Report of the United States Exploring Expedition during the years 1838-1842, vol. 4.

  Appendix 1    Tentative Environmental Chronology

14,000 BP Retreat of Vashon Glacier 

13,000 BP Bretz-like flood (P. Pringle) on outwash plain of Mima Prairie

10,500 invasion of boreal conifers

9,900  first old-growth Douglas fir forest at maximum pollen production

9,800  single greatest charcoal deposit suggests massive fire.  Douglas fir stand perhaps eliminated by a regime of regular burning which shows up as charcoal in pollen cores.

9500-5000  bigger fires once or twice per century with burning in between maintain Oak-savannah prairie. 

6800    Mount Mazama ash later

5700     Osceola Mud Flow? (Pat Pringle)

5000     Douglas fir manage to re-invade prairie and establish themselves.  Result of cooling climate or change in fire-regime?

4500    Trees are quite large, widely spaced with some grassland in between, much as they appear today on edges of Mima prairie.

4400    Blowdown:  Typhoon/hurricane-force wind fells thousands of trees.

4400    erosion of soil from root balls forms base of mounds.

5,100  second largest charcoal deposit as stumps and tinder-dry trees burn, scorching grass, causing more rapid erosion, reducing mound heights. establishment of grass on disturbed landscape slows erosion

4400-1850    regular burning maintains grasses until Europeans halt traditional fire regime

1850-present    recolonization of  “Prairie” by trees Appendix 3

ILLUSTRATIONS can be found at: