The Source

by Martin Doyle

  All of this restoration work, from Van Cleef to Hubbs, was “in-channel” restoration. The river channel itself was taken as a fixed template within which work was done. Large numbers of stone weirs and log dams might be built, but the channel itself was rarely if ever considered part of the project. But in the middle of the twentieth century, there was a growing desire to restore larger rivers; that urge started to raise the potential of working on the river shape itself, and potentially on large rivers.

  Mark Twain opened his memoir Life on the Mississippi by observing that the Mississippi is the “crookedest river in the world, since in one part of its journey it uses up one thousand three hundred miles to cover up the same ground that the crow would fly over in six hundred and seventy-five.” Meanders like those seen in the Mississippi, with their varied flows, depths, and sediments, are considered by ecologists to be the root of the extremely high biodiversity of rivers. Physical diversity begets biological diversity.12

  Yet there are few reasons for industrial society to tolerate meandering rivers. After decades of being content with removing occasional sandbars or accumulations of logs blocking flow, river engineers of the early twentieth century set their eyes on using the increased scale of dredging and bulldozing technology available to correct the inefficiency of meanders, just as industrial foresters had removed meanders on smaller streams and rivers of New England. While dredges and bulldozers were straightening rivers, they also deepened and narrowed them to concentrate flow into as narrow a corridor as possible. On top of all of this, they removed any debris that was in the channels: logs, roots, trees, and so on. All of these activities together—straightening, deepening, narrowing, clearing—collectively fell under the ominous term channelization: the conversion of wide, shallow, wandering, snag-filled rivers into straight, deep, narrow, cleaned-out gutters.

  Along the Lower Mississippi River, channelization was most dramatic between 1932 and 1955, when the Corps of Engineers executed its new flood control mandate by shortening the Mississippi by 150 miles. Beyond the Mississippi, between the passage of the 1936 Flood Control Act and the Clean Water Act in 1972, over 11,000 miles of rivers were channelized as part of Corps of Engineers projects. Meanwhile the Soil Conservation Service channelized over 21,400 miles on its own, often in smaller headwater rivers. One nationwide estimate put total stream and river modification at 200,000 miles, or over 7 percent of the total stream miles in the United States. In states with lowland streams and rivers more prone to flooding, channelization was even more intense: over 26 percent, or 3,123 miles, of streams and rivers in Illinois were channelized in the mid-twentieth century. And the work wasn’t done when the bulldozers left: all of these efforts had to be maintained because the natural tendency of rivers to meander required constantly re-dredging the sandbars that were deposited and protecting the riverbanks that eroded as rivers sought their natural wiggly state.13

  But for society, all of this initial work and chronic maintenance was worthwhile because channelization worked. It reduced the severity and duration of floods locally, it facilitated navigation, and it increased available land for farming. Perhaps more subtly, channelization was a river management practice that played to the strengths of engineers. Rivers could be taken from complex, unruly tangles of swamps and floodplains and converted into straight, linear, trapezoidal forms. Channelization made rivers rational.

  These benefits, however, came with enormous impacts and losses to ecosystems. From destroying fish habitats to eroding banks, channelization inflicted ecological havoc. Particularly alarming was the seeming permanence of the channelization. It was widely recognized that pollution damaged the ecology of rivers, but rivers were seen as “self-cleansing”: given enough time, or given enough distance downstream, they would break down their pollution. Channelization, however, was a more permanent harm. As noted at a fisheries symposium dedicated to addressing the ills of channelization:

  When a stream is polluted, its ecology is greatly changed. Like a man with a serious acute or chronic illness, its activities and functions are altered, often drastically, but there is always the hope of recovery. When a stream is channelized, it is permanently disabled.14

  As ecologists began focusing on channelization, they were able to quantify its staggering effects. Along the Obion–Forked Deer river system outside Memphis, Tennessee, the channelization of over 241 miles was estimated to have reduced aquatic habitat by 95 percent and waterfowl hunting by 86 percent. An ecological study of a river in Iowa, where channelization reduced the length of the river from 63 miles to 34 miles, concluded with a pithy remark typical of natural scientists of the mid-twentieth century: “Results seem to clearly indicate that conditions in the channelized section were not favorable for stable populations of larger game fishes.” And the impacts of such channelization on ecosystems did seem to be permanent. Fish communities had been decimated in rivers channelized over seventy years earlier from North Carolina to Missouri to Idaho, and the mammals and waterfowl that previously used the rivers and floodplains had not returned.15

  Channelization was also not particularly cheap. Since the close of the Great Depression, the federal government had taken on a larger and larger portion of river work. Channelization was often funded by the federal government under the auspices of a range of agencies and improvement programs. As criticisms of channelization began to increase, Congress grew more attentive, leading to a 1973 congressional hearing titled “What Federally Funded Draglines Do to Our Nation’s Streams.” During these hearings, the Assistant Secretary of the Interior for Fish, Wildlife, and Parks stated, “Stream channel alteration under the banner of channel ‘improvement’ for navigation, flood reduction, and agricultural drainage is undoubtedly one of the more, if not the most, destructive water development or management practices from the viewpoint of renewable natural resources.”16

  The damage and destruction of channelization was increasingly seen as excessive; the appeal of the benefits associated with channelization began to wane when the environmental costs were factored in. A series of articles on channelization published in the 1970s included such titles as “Our Ruined Rivers,” “The Stream That Used to Be,” “Channelization: Short Cut to Nowhere,” and “How to Kill a River by Improving It.”17

  And so the pendulum swung back, and the movement to fix channelized streams and large rivers began. At first, the mind-set was to reduce the impacts of channelization by modifying the construction practices. The Federal Highway Administration recommended minimal reduction in channel length and replanting of vegetation. It also recommended replacing or supplementing gravel and large boulders in rivers that had already been channelized to approximate what might have existed before channelization. It was presumed that channelization would continue to occur, but that it could be done in a less harmful way; impacts could be mitigated.18

  There was also growing interest in using the construction activities themselves to build habitat as part of the channelization. Channelization would continue to occur, but the hope was that inclusion of habitat structures could remedy many of its associated problems. By the 1980s, the Corps of Engineers was itself experimenting with modifying its engineered river control structures in efforts to improve fish habitats along the channelized Mississippi and Missouri Rivers. The Corps looked for ways to keep the benefits of channelization and river engineering while also restoring some elements of fish habitat. One method involved notching the dikes along rivers—cutting gaps into river-training infrastructure. Dikes forced flow into uniform channels, so by strategically notching dikes, the Corps hoped to gain some fish habitat while retaining the overall functionality of the river engineering for barge traffic and flood control. Between 1974 and 1980 more than 1,300 notches were cut into dikes along the channelized Missouri River alone, and thousands more were cut into dikes on the Mississippi, Iowa, and Illinois Rivers.19

  In the smaller channelized tributaries upstream, habitat structures likewise began to be built, many of them a
lmost identical to those recommended by Van Cleef and the Michigan School decades earlier. In common with large downstream rivers, and with the streams and rivers restored by Van Cleef and Hubbs decades earlier, the restorative efforts were made via structures down in the bed of the river, while the river channel was left alone; the straightened, channelized morphology was still assumed to be a fixed characteristic of modern rivers.

  For conservationists, however, these efforts worked as well as a Band-Aid would in treating cardiac arrest. What rivers needed was major surgery: if straightened and dredged rivers were ecologically moribund, but natural meandering rivers were ecologically robust, then re-meandering should reverse the ecological damage.

  Re-meandering a river required much more expensive tools and more specialized expertise than Van Cleef’s projects on the Beaverkill or the CCC’s efforts to keep unskilled laborers at work during the Depression. The Stream Conservation Handbook in 1974 had recommended that “simple hand tools, a lot of muscle, sweat and ingenuity are all that is required to make a stream-improvement project” and that “stream improvement may be accomplished individually or by seeking help from Boy and Girl Scout troops and 4-H clubs.” And, in a real callback to Van Cleef and the roots of restoration in fishing clubs, the author asked, “How long does it take a couple of fishing buddies to roll a likely looking boulder into midstream to form a fish shelter?” Doing work at the scale of the Lower Missouri River, the Sacramento River, or the Kissimmee River in Florida would require more than a Boy Scout troop or a fishing buddy. The movement from small-scale construction of trout stream habitat toward larger-scale river construction work would involve backhoes, bulldozers, and dredges. The trial-and-error approach used in the past had to evolve into a more engineered approach.20

  Engineers confront increasingly complex projects with more in-depth analysis and planning. Bulldozers have to be preceded by designs, cost estimates, blueprints, construction schedules, and spreadsheets. Most importantly, for the designs, engineers need equations. Engineering is always preceded in some way by science. There are gaps between science and engineering, both ideological and temporal. The ideological gap is that engineers want knowns, specificity, and as much precision as can be wrung out of the available information. Scientists thrive on unknowns, on observing and thinking about phenomena and processes that have yet to be understood. The temporal gap exists because basic science has to precede engineering; observations must be made, systems identified, and uncertainties decreased until science is distilled to its essence: an equation. From an equation, the power of engineering can be unleashed. In the development of gasoline for automobiles, thermodynamics and chemistry had to precede chemical engineering. To reengineer a river required scientific developments in fluvial geomorphology. What was needed was not an intuitive conception of natural rivers, but rather a set of master equations for natural rivers. And in the 1950s, fluvial geomorphologists were in the midst of developing these equations.

  Throughout most of the twentieth century, fluvial geomorphology was confined to the dusty pages of esoteric journals and offices of ivory towers. Geomorphology was a discipline without a home, sitting awkwardly between geology and geography, and geomorphologists were part scientists, part map readers, part scenic narrators. They were what Michael Ondaatje would describe in The English Patient—a book about desert geomorphologists—as those who “walked under the millimeter of haze just above the inked fibers of a map, that pure zone between distances and legend, between nature and storyteller.” Quite simply, they interpreted the landscape.21

  During the 1950s, geomorphology gained scientific legitimacy when it underwent a quantitative revolution led by Luna Leopold. Leopold was in the vanguard of a new breed of geomorphologists. He initially trained as a civil engineer and then went on to earn a PhD in geology at Harvard in 1950. Rather than mapping what he saw—as a geologist would—or describing what he saw—as a traditional geomorphologist would—Leopold’s eclectic training led him to look at rivers through algebra and calculus. A particle of sediment was not only part of a stratigraphic layer built over indescribably long periods of time; it was a ball on an inclined plane. Thus Leopold could apply equations to the sediment particle and predict when and how it would move.

  Professor Luna Leopold, who revolutionized river science by applying equations to rivers, ca. 1978.

  Leopold’s reputation grew rapidly in the sciences, and by age 41 he was chief hydrologist of the U.S. Geological Survey (USGS). While at the agency, he worked with engineers and physicists to learn new ways to analyze data and think about the natural world. He quickly surrounded himself with brilliant young colleagues who spent the 1950s and 1960s tromping around the United States collecting enormous quantities of field data, which they analyzed according to the new quantitative paradigm that Leopold had pioneered. Professional reports and academic journals in geology and hydraulic engineering were soon peppered with articles on “quantitative geomorphology.” Leopold himself analyzed the volumes of data from rivers that had been collected over the past decades, putting a quantitative spin on this remarkably unplowed intellectual terrain. And as this new breed of geomorphologists measured, graphed, and calculated their way through U.S. waterways, Leopold set out to literally rewrite geomorphology—to pen a book on the subject that would change the way the world thought about rivers.

  Leopold had two co-conspirators in his work, the first being John Miller, a geology professor at Harvard who had spent considerable time with Leopold doing fieldwork in the Rockies and shared his intellectual trajectory. Tragically, Miller contracted bubonic plague while measuring soil erosion in New Mexico and died only a few days after returning to his home in Cambridge. With the book only half done, Leopold pushed forward with another author: Gordon Wolman of Johns Hopkins, known to everyone as “Reds” for his curly, wispy red hair. Wolman benefited from the same Harvard training as Leopold, but while Leopold had spent much time in the Midwest and the Rocky Mountains, Wolman was a self-described “Baltimore man,” intimately familiar with the eastern rivers and the dense human footprint of the twentieth-century Atlantic Seaboard. Most importantly, he shared Leopold’s penchant for the quantitative, and the two of them became fluvial geomorphology’s scribblers in chief.

  The combination of Leopold and Wolman was historic not just for what it contributed to the future of fluvial geomorphology, but for how it embodied the past. Leopold was the son of Aldo Leopold, author of A Sand County Almanac (1949), a book that elegantly mixed natural history and philosophy. In much the same way that Rachel Carson’s Silent Spring embodied the antipollution movement, Sand County Almanac became the bible for the wilderness and conservation movement. Sand County Almanac was unfinished when Aldo Leopold died, and it was Luna who did the yeoman’s work of completing, editing, and publishing the book.

  All the Leopold siblings spent their childhood living in the famous “shack” in rural Sauk County, Wisconsin, the holy temple of conservation. Aldo Leopold’s five children would rank highly in the who’s who of environmental science for the second half of the twentieth century. Four of the five received PhDs, and three became members of the elite National Academy of Sciences. Luna was not only admitted to the National Academy but also awarded the Presidential Medal of Science for his work in rivers. Luna’s intellectual interior, however, was concealed by the gruff exterior of an environmental activist—a man who would leave the Wisconsin shack for a Wyoming ranch retreat where he studied rivers and wrote. After decades with the USGS, Leopold became an infamous presence in the halls of UC Berkeley, where he terrified graduate students with his pointed questions and intolerance for sloppy thinking.

  Reds was the yin to Luna’s yang in almost every way. Luna’s Stetson and scotch were countered by the ever-present bowtie and martini preferred by Reds. While Luna spent his formative years in a shack in rural Wisconsin, Reds spent his in the posh neighborhoods of Baltimore—where his father, Johns Hopkins professor Abel Wolman, founded the discipline o
f sanitation engineering and wastewater engineering. The City of Baltimore renamed its public works building after him, and the highest award given by professional engineering societies for work in water treatment remains the Abel Wolman Award. After a stint working for Leopold at the USGS, Reds returned to the same Hopkins hallways as his father. From this rarefied academic perch, Wolman also gained prominence in science—he was also inducted into the National Academy—as well as produced an entire generation of academic progeny, dozens of fluvial geomorphology PhDs who became the big names of river science in the late twentieth century and revered Reds as a gentle sage at Hopkins. Where Luna was a tall, broad-shouldered, imposing western cowboy scientist, Reds was shorter, slight, genial, gentle, and at ease with anyone and everyone.

  But when they worked together, these men produced a book that was as transformative for river science as their fathers’ works were for conservation and engineering. Indeed, their book, Fluvial Processes in Geomorphology, was an intellectual continuation of their respective fathers’ work—a painstaking concoction of natural science derived from Aldo and engineering derived from Abel. This book, along with the body of work being produced by Luna, Reds, and their academic cohort, took geomorphology from narrative to equation, from descriptive to predictive, from the realm of scientists to that of engineers.22

  The power of this new way of thinking was that it converted natural rivers into a series of equations showing that rivers were somewhat predictable. River forms could be boiled down to a series of lines on graphs. With data gleaned from seemingly innumerable studies of rivers throughout the United States, Europe, and even India, the pool–riffle sequence went from a desirable feature to a scientific rule. In fact, Leopold and Wolman noted that their volumes of data suggested that an entire pool–riffle sequence—or, as they perceived it, a complete sine-wave meander bend—would occur about every six channel widths. That is, if a river was about 20 feet wide, then every 120 feet a person walking down the river should expect to wade through a complete pool–riffle sequence, in a mathematical and predictable rhythm.