The Source

by Martin Doyle

  Van Cleef had intuitively combined two important ideas that would shape—and be debated in—stream restoration for the next 150 years. First, he recognized that the physical shape of a river—its morphology—was a controlling factor for trout, and that a river’s morphology was as malleable as the layout of any road. But second, he intuited the potential for private motives to restore the trout streams. Van Cleef suspected that the value of trout fishing was the experience, and he thought the trout-fishing experience was as valuable as any commodity: if people could be ensured of a good fishing experience, then they would be willing to pay some fee—an access fee—to fish that particular river. His thinking about restoration was shaped by the economic realities of the nineteenth century and his particular role as an industrial attorney. He presumed, based on the economy at the time, that private efforts were needed to restore and preserve streams, and that government would play little if any role.

  In 1868 Van Cleef organized the Willowemoc Club, which began privatizing Catskills fishing areas exclusively for club members. He and his fellow club members purchased thousands of acres of trout-fishing waters in the region. The group would go on to form the Beaverkill Club in 1878 and the Balsam Lake Club in 1886, which collectively amassed more than three thousand acres, including six miles of the upper Beaverkill River.4

  After preserving the land as private property, an approach little different from contemporary efforts by The Nature Conservancy, the clubs then turned to fixing the streams. In the 1870s Van Cleef developed rough sketches and descriptions of construction approaches, whose implementation became a hobby for club members between their fishing outings. They used what was available—logs and boulders—and placed them in particular configurations and places to create enough of the conditions for trout to be attracted to their stream reaches again. They built miniature dams—weirs—across channels to create pools upstream and fast-flowing riffles downstream. They brought in larger rocks to build veins in the river—partially submerged barriers to flow, like hydraulic speed bumps. They tried various shapes and patterns of different structures in their search to make things just right for the now elusive trout.

  These were amateur efforts. Van Cleef and those who followed him in the Catskills were passionate, but they were hobbyists nonetheless. Their work was based largely on intuition and what they observed in streams they thought were in better condition. They bankrolled their own efforts and benefited personally from their relatively modest successes in actually improving trout fishing, which in turn may have increased the membership and influence of their fledgling fishing clubs. But their restoration work could never have been considered financially profitable—or even really profit seeking. Nor would they have the opportunity to test the long-term profitability of this venture—by the time Van Cleef’s protégés in the Catskill fishing clubs had completed much of their work, and could theoretically have sold club memberships for a higher price, the country was in the grip of the Great Depression.

  Wading in rivers can be a clumsy endeavor. Submerged logs, shifting sand, and swift flow can humble any novice fisherman. Rubber waders add bulk, making a stream or river that much more challenging to navigate.

  The mark of a veteran fluvial geomorphologist in the field is dexterity when walking in rivers. Doug Thompson, clad in chest waders, skips and scoots down into Connecticut’s Blackledge River, then up on its banks, then down and through pools and riffles with an ease that shows he has spent an inordinate amount of time tromping through streams. He is as comfortable in waders as he is in his basement flume lab. When Thompson started teaching at Connecticut College in the late 1990s, his flume was the workhorse for his research, but he also studied some local streams. His goal was to build an enormous dataset of the characteristics of natural pools and riffles in New England. The Blackledge was convenient, and so he started studying the pools there, dragging along his students to measure all things hydraulic. The Blackledge was where his career started, and also where it took an unusual turn.

  Standing knee deep in the Blackledge River, Thompson points out a few of the logs half buried in the gravel of the riverbed. The logs are largely innocuous except for one thing: they are perfectly perpendicular to flow. Then he points downstream a bit toward the bank of the river. There, a line of half-submerged rocks protrudes from the bank into the main flow of the river, and they are slightly larger and more angular than the rocks in the river. As we move on down the river, Thompson points to another log, this one oriented at a perfect 45-degree angle to flow. When he first noticed these startlingly precise eccentricities a decade earlier, he realized that “something wasn’t quite right.”

  At the end of one of the logs, Thompson points out a rusty spike that is mostly buried in sand. Not modern rebar or a well-manufactured binding of some kind; a spike. This was the kind of spike that Thompson had literally tripped over early in his research; and after seeing the first spike, he dug around and found a lot more in logs all up and down the Blackledge. Thompson had stumbled onto the work of humans in the middle of nowhere. With a trained eye, the rest of the river comes strangely into focus: a rhythmic series of fingerlike rock piles or logs in angular structures poking out into the stream. Each of these structures was intended to deflect flow to form a small pool and riffle. Thompson had wanted to study how pools and riffles are formed naturally, but there weren’t actually any natural pools and riffles in the Blackledge. This is a fairly common predicament in the heavily modified New England landscape. As he acknowledges, “It turns out working on completely natural pools in New England may not have been the best idea.”

  Initially flummoxed, Thompson started asking around to see who was responsible for the riverine construction work. The first answer was the local department of transportation, which had moved the stream around when building the adjacent highway in the 1950s. But the highway department had only been replicating patterns first put in place decades before. The spikes were the handiwork of the Civilian Conservation Corps (CCC).

  At the turn of the twentieth century, Teddy Roosevelt and his chief of the Forest Service, Gifford Pinchot, turned the federal government’s attention toward environmental conservation. The initially agrarian economy of the United States had begun growing tremendously thanks to mining and manufacturing, both of which were decimating the environment to the benefit of private industry. Roosevelt and Pinchot grew increasingly critical of this development and began envisioning ways to preserve some of the nation’s natural resources.

  Pinchot was a quintessential progressive; he believed societal problems could be fixed only by science, careful planning, and professionals doing the work. Stream restoration efforts by earlier groups like Van Cleef’s lacked all these things; they embodied the quaint inadequacies of conservation left to amateurs: piecemeal, untested, and transient. Most of all, the approach was unscientific. That this pre–Progressive Era restoration was also being done by a group of vacationing industrialists on private club property only added insult to injury for Pinchot. He saw efforts like those of the Willowemoc Club and other amateur conservationists as inappropriate to the scale of the problem, which called for federal government leadership. Roosevelt and Pinchot’s initiatives would create sprawling new mega-agencies like the Forest Service, the Bureau of Reclamation, and the Federal Power Commission, all of them focused on using the more conservation-minded potential of government in opposing the seemingly overwhelming force of markets for natural resource extraction.

  The catalyzing opportunity to make stream restoration progressive, as opposed to an elite hobby, arose in Michigan when the so-called Michigan School began its work. Formally known as the Institute for Fisheries Research at the University of Michigan, this unit was formed under the biologist Carl Hubbs in 1930. Hubbs was a transplant to Michigan from California. As a Stanford graduate, he began his burgeoning scientific career studying the fishes of the Los Angeles River—which were, unsurprisingly, in steep decline. Hubbs moved to the Midwest in 1917 to be
curator at the Field Museum in Chicago and then in 1920 moved to Michigan, where he was again a curator but also took the opportunity to complete a PhD in 1927. Hubbs served as director of the Institute for Fisheries Research from 1930 to 1935 and was in the vanguard of stream fisheries scientists. He ended his career back in California as director of the Scripps Institute of Oceanography, where he would be inducted into the elite National Academy of Sciences.5

  In Michigan in the early 1930s, Hubbs had begun assembling an army of students who turned their attention to trout. For years, the decline of trout had been countered by stocking—rearing them in hatcheries and then dumping them by the hundreds or thousands into streams. This approach effectively accepted that trout could not survive to reproduce in nature in its human-modified form. Hubbs wanted to rely less on stocking and more on improving habitat conditions so that trout could naturally reproduce. In 1932 Hubbs and his students put together some of the earliest formalized, scientifically based guidelines for restoring streams. They were published in a nondescript pamphlet titled Methods for the Improvement of Michigan Trout Streams. This little pamphlet eventually became a cornerstone document that placed habitat improvement—restoration—on an equal footing with fish hatcheries and stocking programs for trout stream management across the United States.6

  What Hubbs and the Michigan School actually did—what they built to restore the landscape—was largely the same as what Van Cleef had done in the Beaverkill half a century earlier. Most important, they began by constraining their efforts to working within the existing stream channel: they thought the overall shape of the channel was a set feature of the landscape, but they considered the details of the streambed as malleable—able to be designed for specific purposes. From that conceptual starting point, they constructed small dams, weirs, boulders, and logs in particular shapes and spacing. Like Van Cleef and his compatriots, Hubbs and his students were attempting to create hydraulic habitat within the confines of the existing river. But what they did next was entirely different from the haphazard work of the Beaverkill crowd, for Hubbs set out to meticulously document what happened in the streams after the structures were placed. This approach was classic Hubbs; he was a notoriously thorough and careful scientist, and for himself and his students, he intended to handle restoring streams as methodically as any scientific experiment.

  The Michigan School conducted many such studies over the following two decades, documenting the changes in number, size, and feeding habits of the trout in restored streams and unrestored streams. Over the second quarter of the twentieth century, this group published dozens of articles in the primary fisheries science journals. These studies were the basis for development of longer-term evaluations and guidelines, and as students matriculated from the Michigan School and populated government agency positions elsewhere, their ideas about the utility of trout stream restoration as a technology or management technique spread nationally as well. Trout stream managers who had picked up the stream restoration bug as far away as North Carolina’s Pisgah National Forest described the process as having an experimental period, when principles were developed and put into practice, followed by a period in which those principles were adapted and “modified somewhat as a result of additional information and the stress of changing conditions.” The entire restoration process was conducted with a classic scientific sense of trial and error.7

  The irony of the Michigan School’s restoration efforts was that, despite all their monitoring and reevaluating and bringing in the best and brightest graduate students, few actual improvements to in-stream trout habitats were made during the mid-twentieth century; an observer would be hard-pressed to distinguish between a stream restored by Hubbs and a stream restored by Van Cleef half a century earlier. It would also be hard to distinguish between a restored stream and an unrestored stream by looking at the number of trout in each stream. If Hubbs’s meticulous research and technical savvy had shown anything, it was that restoration as it was practiced up to the mid-twentieth century did not actually work.

  This finding is what most fascinates—and continues to frustrate—Doug Thompson, whose office at Connecticut College is lined with books on turbulence and fluid mechanics as well as file cabinets stuffed with old surveys, maps, and data from a century earlier. Thompson was a well-known rising star in fluvial geomorphology and fluid mechanics circles when he stumbled onto that spike in Blackledge River. After that discovery, he continued publishing papers on turbulence and river processes, but he also started spending more time in the archives of transportation departments and fishing clubs. There he discovered some of Hubbs’s obscure reports, along with bits and pieces of associated data. He immersed himself in the archives of the Michigan School and other decades-old datasets that had been accumulated over the years from far-flung restored streams across the United States. He then did what obsessive scientists do when they get fixated on a particular problem: he reanalyzed the data. In 2006 Thompson published the results of this eco-historical study in a deceptively understated scientific article that concluded, “My statistical results show that the traditional use of in-stream structures for channel restoration design does not ensure demonstrable benefits for fish communities, and their ability to increase fish populations should not be presumed.”8

  Much of the data that Thompson used came from Hubbs, even though Hubbs didn’t know at the time that his techniques were not working—or at least he didn’t know it for certain. In later publications and reports, Hubbs started to hint at his suspicion that not all of the restoration work had been effective at increasing trout populations; but he and his students remained in a river-fixing frenzy. Why? Because stream restoration was filling a desperate need of society: it absorbed labor.

  When the Great Depression hit, the effectiveness of stream restoration was unproven, but the work was perfectly positioned economically, politically, and pragmatically. Just as building levees, dams, and harbors absorbed labor on large rivers, stream restoration absorbed labor on a smaller local scale. Small stream restoration was ideal for the CCC because it required very few technical skills, was done in rural backwaters of the United States, and used local materials like logs and boulders. It was technically primitive because stream restoration as practiced by the Michigan School was in reality a re-creation of the types of structures first thought up by Van Cleef a half century earlier. The Michigan School enabled stream restoration to be included in the CCC program by promoting and justifying the work through studies published in the peer-reviewed pages of fisheries journals. The scientific publications gave these decades-old practices some sense of slide-rule perfection, which provided the grist for technical how-to manuals and pamphlets printed on cream-colored university stationery; stream restoration was thus granted academic credibility with or without any evidence of its success.9

  As methods developed by the Michigan School were eventually adopted by the CCC and the U.S. Forest Service, the Hubbs team used workshops to teach their methods. Hubbs’s assistants moved throughout the United States, leading labor crews in various regions in restoring streams. With CCC backing, stream restoration rapidly expanded well beyond its earlier confines of the Catskills and Michigan, moving into the backwaters of North Carolina, California, and Wyoming. Between 1933 and 1935, over 31,000 structures were constructed on over 400 mountain streams, and by 1936 the CCC had improved almost 5,000 miles of stream. This entire endeavor was a combination of the old and the progressive: the same techniques as Van Cleef, but put to work for Progressive Era public good.10

  The shift from Van Cleef and his private restoration clubs to Hubbs and the Michigan School’s public restoration projects was a major ideological change for stream conservation; it indicated that progressives had won an important battle in the national economy. For Van Cleef and the industrialists of the late nineteenth century, private property and limited access were essential to recovering trout; preservation required privatization. For progressives, making trout streams available to th
e public was part and parcel of conservation. Or, as some Forest Service stream restorers captured it, “The goal established has been kept ever in sight: The most trout fishing for the most trout fishermen, and an equal opportunity for every angler.” There was no actual economic demand for restoration itself, or even for an increase in the trout population. The true results were an incremental increase in the perceived fishing experience of the general public and increased nationwide employment. In fact, the public fishing easements purchased along selected trout streams of the Catskills were intended primarily to provide the Conservation Department with another opportunity to do stream improvement work. Although the actual work being done on streams was not changing too much, the people doing and funding the restoration efforts, along with the places where they worked, were profoundly shaped by the broader forces of the national economy.11

  That all of this stream restoration in the early twentieth century didn’t work, as Doug Thompson would show decades later, made little difference to the work or perceived success of Hubbs. Even though Hubbs never analyzed the data systematically made a summary finding or recommendation, Thompson remarks, “The best indication about what Hubbs thought about restoration success is that once he got to Scripps in California, you don’t hear a damn thing about stream restoration.”