I love wild streams—streams with grassy undercuts, logjams, beaver dams, gravelly runs, and headwaters full of young-of-the-year fish. I’ve found myself reflecting on how unfortunate it is to see what used to be a beautiful section of wild stream, now with rocked-up banks and artificial structures—man-made reminders that few places go untouched. Some relish these “improved” streams. They tend to be easy to walk along and cast flies to, yet they’re not natural. And to me, there is nothing more beautiful than the course of nature.
I grew up on a small tributary to a better-known stream in the Driftless area of Southwestern Wisconsin—Driftless because the Wisconsin Glaciation never covered the hills there. As a result, the topography consists of stout little hills and bluffs with valleys shaped over the long years by the rains and the spring fed streams that flow through them.
The original section of property that my parents owned had cattle on it first, and then our horses. It was my dad’s prerogative to get the horses off of it so that the stream banks and vegetation could start recovering and provide better fish habitat. He fished the stream on occasion, but didn’t catch much at all, and we looked longingly on our neighbor’s section of water upstream. Although formerly channelized, their section hadn’t felt the effects of cattle, and even looking over the fence, the difference in bank structure and the quality of pools was remarkable. Heavy grazing on our stretch of water had resulted in beaten-down banks with little to no structure providing cover for fish. In comparison, overhanging knobs of grass and undercut banks provided abundant cover in the upstream section.
Sometime in the mid-to-late 90s we purchased our neighbors’ land when they stopped farming, and I started cutting my teeth on the local trout population using a five-foot spinning rod and some Panther Martin spinners. For several years I’d fish our stream every two to three days—as often as I could while still giving the trout a bit of a break. Of course I always caught more fish on our neighbors’ former section, but it took surprisingly little time for our original stretch to bounce back from being grazed. Annual floods and high water slowly reformed the banks. Improved riparian vegetation—including long, tough grasses and trees—stabilized the bank (preventing catastrophic erosion), while reduced grazing kept the banks from being beaten down, allowing beneficial erosion to undercut the banks. By permitting these natural processes to occur, we were regularly catching fish in our original section of stream within a few short years.
In high school, as I got more serious about trout, I became excited about something called Little Underwater Neighborhood Keepers Encompassing Rheotactic Salmonids (commonly called LUNKER structures). Our high school shop classes made them, and they were supposed to provide excellent structure for fish. Their design is like an open wooden box, placed in a stream bank with rocks and sediment placed on top and the opening facing into the stream underwater. They imitate a natural undercut with the idea that further erosion and stream migration will be minimized.
During this buzz about LUNKER structures, everybody I knew who fished for trout also went on about the terrible effects of beavers on trout and trout streams. They blocked the stream and raised the temperatures, they said, and at the time that was sufficient reason to despise them. It seemed simple logic that beaver dams slowed stream flow, thus warming water temperatures and necessarily harming trout populations. Yet beavers and trout (and salmon) have co-existed for thousands of years. In fact, the near extinction of Atlantic Salmon in the United States has been correlated with extinction of beaver in many eastern watersheds prior to later habitat degradation for which salmon declines are usually blamed. So why did we proclaim that beavers are bad for trout?
It was only later, in college, that I began to think more about the bigger picture with regard to stream conservation. I actually did some reading about beavers and their impact on stream ecology. It turns out that slowed stream flow may warm temperatures slightly, but beaver dams have been shown to reduce levels of eutrophication agents (e.g., sulfates, phosphates, and nitrates) downstream.1 It’s also been shown that beaver dams are not necessarily significant barriers to fish migration.2 Beaver restoration efforts are even linked to increased survival rates of endangered steelhead.
Equipped with more robust knowledge about beaver dams, I began to take a different approach to stream and fish health. I began to ask myself if any given variable on a stream was something that would occur naturally in the absence of (negative) human intervention—be it herons, beaver, erosion, or LUNKER structures. This question radically transformed how I approached stream health.
Rather than focusing on one potentially negative influence, I began to ask what the benefits might be. How might the entire ecosystem benefit, or even depend, on a given natural phenomenon? I found myself asking, what is the difference between “improving” a trout stream through short-term actions versus restoring a watershed over time?
This same dilemma has been at the center of stream management practices—and ecological restoration more broadly—for some time. In 1989, Nathaniel Reed (former Assistant Secretary of the Interior) gave a plenary address at the Wild Trout IV symposium—entitled “From Wild Trout to Wild Ecosystems: Fifteen years of Evolving Stewardship”—in which he reflects on his firm decision (despite opposition) to let the 1974 Waterfalls Canyon Fire in Teton National Park burn itself out. In the address, Reed mentions a conversation he had with Starker Leopold on a high Yellowstone plateau, amidst bone-dry tinder waiting to be ignited. Reed asked Starker what would happen when that tinder burned, and Starker responded that it would depend on how much the public understood about the benefits of wildfires. Reed’s decision to allow the Teton fire to burn itself out was a product of the philosophy that nature is better at managing itself than a single person, or even an educated group of professionals, ever could. However, being at a Wild Trout symposium, Reed was aware that he had to show the benefits of this philosophy to anglers. Implementing catch-and-release policies and restoring riparian habitat in over-grazed areas, he demonstrated, resulted in watershed and habitat restoration for cutthroat trout populations, which in turn benefited grizzly bear populations within the park. Most importantly, he discussed the importance of letting nature run its course. One point resonated with me in particular—being a resident of Wisconsin, a state that had huge issues with habitat degradation and declines in wild fish during the early 20th century. His point was this:
Within some limits a system can and will heal. We have seen it heal and you have helped to achieve it. We do not know the exact point of destruction—but what is important is that healing will occur if we allow it to occur. We need to repress our driving desire to manage, to control. Yellowstone did quite well without any of us for thousands of years. It made it without us. As a matter of fact, the world may have been a far better place before biologists, bureaucrats, gurus, philosophers playing biologists—even before former Assistant Secretaries.
The key word in that paragraph, I think, is “allow.” In the late 19th and early 20th centuries, farming practices on the steep hillsides of Southwestern Wisconsin led to massive erosion problems due to topsoil runoff. This not only damaged the farmlands, but also lowered groundwater levels and caused excessive siltation of streambeds. Praiseworthy efforts by Aldo Leopold, the Soil Erosion Service, and progressive farmers like the Haugen family of Coon Valley sought to fix both the agriculture and the ecological problems caused by farming practices of the time. These efforts resulted in new land-use practices that kept wooded all hillsides with a grade of 40 percent or more, and allowed pasture on slopes with a 20 to 30 percent grade. Contour stripping (a method of farming that plants crops along the natural contour of a hillside) and crop rotation practices were used on slopes with a lesser gradient in order to further prevent topsoil erosion.
This watershed-level management drastically improved crop production, reduced topsoil erosion, improved groundwater levels, and resulted in cleaner, colder water that allowed fish to thrive and naturally reproduce—something they hadn’t been able to do in thousands of miles of historically trout water.
Paradoxically, and maybe even counter-intuitively, erosion is not always bad: as we discovered on our stretch of stream, it is also a natural process that restores nutrients to streams and maintains stream complexity by introducing woody debris and undercutting banks. Although undesired in large amounts, sheer banks of exposed sediment also provide habitat for swallows. While changes in agricultural practice are one part of watershed management, holistic, long-term changes must also be implemented at a localized level as well. Watershed improvement work done now by groups such as Trout Unlimited, in coordination with the Wisconsin Department of Natural Resources tends to be much more localized, though not always holistic. For this reason, it often remains limited to short-term “improvement” rather than long-term restoration. Projects include clearing away brush and box elder (Acer Negundo), placing large limestone riprap (often in conjunction with LUNKER structures) in stream banks for stabilization and to prevent erosion, and even dredging the stream bottom to create deeper pools. Though placing LUNKER structures and rocks on a meandering corner may compensate for bank degradation by cattle, channelization, and over-farming, this is not restoration.
While the comparison may be a bit over-dramatic, the desire to rock up every eroding bank and stream is reminiscent of the drive to eliminate predators from the hillsides with the misconceived notion that we are better at managing nature than nature is, and that somehow by killing the wolves and coyotes that we’ll have more deer. We replace dirt with rocks and think that by killing the wolf we will have more fish to catch. We don’t stop to think of the stream itself. Yet every time we line a bank with rocks, clear out the trees, and replace them with our own plan, the green light radiating from the stream dims a little bit more. A holistic approach to stream restoration and watershed management recognizes the importance of natural processes such as erosion, as well as the role that beaver, for example, play in riparian habitat complexity and alluvial water table levels.
Restoration in the true sense of the word does not place rocks in a bank that would not normally occur there, or install artificial structures that provide cover for fish. In fact, the use of the word “artificial” to describe any aspect of habitat improvement should be cause to re-evaluate the ecological validity of that work. Even now, as restoration efforts attempt to take more natural approaches—using root wads and logs from the trees cleared along the stream—this cannot be seen as anything but an attempt to imitate nature in the short term. Rather than let trees grow to have the bank eroded around them, providing natural root-wads and eventually downed trees for cover, we clear trees out and place the roots and logs the way we think they should go. Rather than let stream flow deepen areas of the stream, we dredge pools, only for them to be filled in by a stream that has its own agenda and ways of working. Worst of all, streams that have miraculously sustained themselves over the years, despite being ignored by back-hoes and diggers, find themselves suddenly the attention of “improvement” by individuals who see an isolated eroding bank, and not a natural process that corrects itself over time.
In his plenary address at Wild Trout IV, Reed went on to say:
Could it be that the advocates for hands-on management really believe that they are equipped with the knowledge to order fires started—as if once they started they could be put out at will; or that they have the Godlike wisdom to supervise the execution of bison and elk that they decide are in their personal view “surplus?” Is that not playing at being God? I prefer the real acts of God, they require patience and discipline…
I prefer stream changes that require patience and discipline, and are truly long-sighted. I prefer watershed restoration over stream improvement. Restoration that sets nature on a course to sustain itself without frequent—if any—continued intervention.
I prefer natural stream changes that provide an element of surprise and discovery upon re-entering a favorite section of water now subtly changed and different. I prefer stream changes that make me appreciate the transitory nature of a favorite hole or run, never knowing when it will change or shift. I don’t enter the outdoors to appreciate the human ability to shape nature to our liking, nor do most of us, I suspect. I enter the outdoors to observe nature at work, and because there are some places where we can still know and see clear evidence that the world is a beautiful place that is able to be itself as long as we allow it.
Featured Image: Fishing a classic Driftless trout stream. Photo by Luke Annear.
Luke Annear was born and raised among trout streams in southwestern Wisconsin, and began fly fishing when he was in middle school. He received his master’s degree in Speech-Language Pathology from UW-Madison and lives in Black River Falls, Wisconsin. During the summer he is a fly fishing guide for carp. Luke blogs occasionally about fly fishing and related subjects at youngmanandthestream.blogspot.
S. A. Muskopf, The effect of beaver (Castor canadensis) dam removal on total phosphorus concentration in Taylor Creek and wetland, South Lake Tahoe, California, Master’s Thesis (Humboldt State University, 2007). ↩
P. S. Kemp et al., “Qualitative and quantitative effects of reintroduced beavers on stream fish,” Fish and Fisheries 13, no. 2 (2012): 158-181. ↩