Rebuilding nature's filters: the reclamation of streams

by Bob Petersen

In the 1960s, at great expense, Danish engineers "channelized" the Skjerna River, removing the vegetation that grew alongside it and deepening and straightening the river bed, in order to speed drainage of its adjacent wetlands and claim them for agriculture. The Skjerna's catchment area was the second largest in the country, and hundreds of farms were enlarged as a result.

Thirty years later, however, authorities admit the project was a mistake. The Danish Parliament has voted to restore the Skjerna to "its former channel and ability to overflow its banks", at a cost of some US$10 million. In a stroke of unusual irony, the same engineering firm that originally channelized the river has been assigned to undo its own earlier work.

Denmark isn't the only country where the policies of past decades are being reversed. Throughout the world, small streams and rivers have for years been the object of "improvement" efforts similar to the first Skjerna project. Such projects unquestionably increased the amount of land available for farming. But, it is now realized, they also destroyed many functions natural streams and stream-side vegetation perform that benefit mankind - functions we can ill afford to neglect.

One of these is the ability of streams to absorb urban, industrial and farm waste, something they have been doing for centuries. Not only do streams carry waste away but, if they are not overloaded and their physical structure is not destroyed, they gradually reduce the concentration of nutrients, organic compounds, metals, and toxic pollutants such as pesticides, herbicides and excess chemical fertilizers that run off from farm fields. In some places where the ability of streams to do this has been compromised through channelization, too much nitrate is leaving farms and entering the ground water, where the resulting high nitrate levels pose a health hazard to drinking water.

In addition to this so-called "self-cleaning" ability, a healthy stream provides habitat for fish living in the channel, as well as wildlife and medicinal plants living along its margin. In arid and semiarid areas, the shading canopy of streamside vegetation can effectively reduce water temperatures, thus minimizing moisture loss through evaporation.

Attention throughout Europe is increasingly focused on reversing the loss of these important stream functions, via practical, cost-effective ways that don't disrupt the productivity of surrounding farmlands. (Further catalyzing this trend is the realization that while farmland and farm products were in high demand in the 1960s, too much land was eventually converted to agricultural use, driving down farm prices.) Numerous restoration projects are under way. In Austria hydraulic engineers who were still channelizing streams only a few years ago now say such efforts are a thing of the past. Swedish engineers admit the hydraulic systems they developed to drain water from the landscape were too efficient, taking too much water from the catchment area at too fast a rate. An adequate solution to the resulting problems can only be implemented at landscape level, involving entire catchment areas, and including not only the channels themselves, but streamside vegetation zones as well. A complete change in the structure of the landscape is required, from a patchwork of fields to a mixture of fields and clean-flowing streams protected by a streamside canopy of vegetation.

This necessitates a plan that can take into consideration vast differences in local topography, economic resources, and the personal preferences of individual landowners. As far as possible, such a plan must also be simple, cheap and flexible.

Building Block model

One such approach is to see restoration measures as a series of Building Blocks, which can be combined in different ways over time according to landowner preference, landscape form and available funds. Installed together, these blocks can improve natural habitat, restore self-cleaning capacity, and reduce the loss of important nutrients from the landscape. While some measures are more important or costly than others, all are effective. They can be used individually or together.

The Building Block model can be described as a series of steps in three groups. The first group consists of three measures having to do with reestablishment of streamside vegetation. The second comprises three steps dealing with reducing the steepness of channel banks and changing channel shape. The third includes additional measures that help retain water in the catchment. The restoration measures in question are relatively independent, and, except for ( I ). do not necessarily have to be taken in the order given:

(1) Buffer strips - The first and most important restoration measure is to set aside a minimum 10-metre-wide strip of land between adjacent fields and the water, on both sides of the stream channel, along the entire length of the stream. These "buffer strips" reduce (or buffer) nutrient and sediment loss from the landscape, while also providing bird and plant habitat. They are the chief determinant of small stream physical structure, because their presence protects stream banks and streamside vegetation. They are the foundation for other restoration measures. The cost of setting aside buffer strips depends on land values. In Sweden, good agricultural land runs between US$3 000 and $5 000 per hectare. Assuming 10 metres on each side of the water, two ha/km of stream will have to be purchased for each kilometre of stream channel. Since most farmland is fairly flat, with a drainage density of roughly one km of stream/km² of landscape, setting aside two ha of land for stream restoration requires setting aside 0.2 per cent of the catchment. Actually, most land near streams is difficult for farmers to use, either because it is too wet or too close to steep channel banks. In addition, the cost of setting aside buffer strips has a hidden discount. Channels, in most farm areas, are about five metres wide at the top. That is 0.5 ha, or 5 000 m2 of land/km of stream, that is not being used. The total width of the buffer area will be 25 m, not 20 m, representing an immediate 25 per cent discount on the purchase.

(2) Revegetation - Buffer strips which have been under field crops should be planted first with fast-growing woody shrubs (in Europe, alder. Willow or aspen). Without replanting, buffer strips will still re-seed naturally, but the process would take up to seven years to achieve complete effects. A mixture of native riparian species is recommended for replanting. Exotic species should he avoided. The effect of replanting will be to eventually restore a landscape more in keeping with local architecture and other cultural appearances. Nor should the importance of the natural plant life that will recolonize the strip be underestimated. Vegetation will stabilize stream banks and inhibit erosion, while its canopy will prevent evaporation. Growth of indigenous plants will be stimulated, including medicinal or edible species. For example, sweet flag (Acorus calamus) is a common aquatic plant used in Europe, Asia and North America for its beneficial effects on the stomach, relief of chronic dyspepsia and as a sedative.

(3) Side slope reduction - A major source of sediment and phosphorus discharge into streams are small-scale land failures (caveins) along the channel. Normally they require re-channelizing and cleaning every one to 10 years, depending on soil structure and the discharge character of the stream. A simple method of limiting this is to reduce the channel side slope by half from the commonly used 50 per cent ( 1:1 or 45-degree) slope to a 25 per cent ( 1:3 or 22.5-degree) slope. The amount of soil that must be removed and spread on the surrounding land is slightly more than two cubic metres for every metre of channel restored. Given that 10 m have been set aside on each side of the channels the amount of land required to reduce the side slope to 25 per cent will be five metres, an area well within that already set aside. Side slope reduction brings several positive effects. First. it increases the width of the water channel, creating a small flood plain to dissipate hydraulic energy by flooding, rather than undercutting, side banks (a leading cause of erosion and slope failures). Second, it reduces the need for channel maintenance, conserving both money and the integrity of the stream.

(4) Meanders - Re-creating meanders in a stream that formerly had them will restore its most stable and natural physical state. This will simultaneously reduce nutrient loads by increasing the physical complexity and length of the channel. A meandering channel by definition is one with a channel length at least 1.5 times the length of the down-valley distance. Since many stream processes can be quantified on a length basis, a longer channel automatically has greater self-cleaning properties.

(5) Riffle pools - Where a slightly higher gradient is available, the physical complexity of the stream bottom can be increased by placing rocks at intervals, to produce riffle areas. These should be followed by pool areas. The natural dimensions of riffle-pool associations should be set so as to have one pair - a riffle and a pool - at a downstream distance of five to seven times the stream width. That is, for a one-metre-wide stream, a riffle-pool pair should occur every five to seven metres. The riffles should be about three metres long, followed by a two-metre-long pool. Actual spacing is not critical, however, since streams will re-adjust rocky sediments themselves during the first flood. Alternating riffle-pool sequences are an important aspect of restoring the within-channel habitat of small streams. In addition to stimulating re-oxygenation of the water via the increased turbulence in riffle sections, the clean rocky substrate forms prime habitat for many aquatic invertebrates and a feeding ground for fish. The pools provide rest and refuge areas for fish, as well as storage areas for organic material, which is slowly released into the stream.

(6) Ponds - Small ponds constructed within the flood-plain valley are an economical and multi-use restoration measure. Uses include water retention for irrigation, as well as habitat for raising fish or crayfish. The latter are a particular favorite of Swedish farmers. While ponds may be used for intensive aquaculture requiring fertilization, natural or extensive aquaculture is recommended for sustainable use. Ponds are usually sited off-channel, rather than in the stream itself. A typical small pond will retain organic material and nitrogen, which in turn enhance the value of irrigation water drawn from it.

(7) Riparian wetlands and swamp forests - Along many lowland channelized streams are areas
which are seasonally difficult for farmers to plow, due to excess water. These swampy areas were often originally wetlands or swamp forests, and once restored will enhance wildlife conservation as well as water and nutrient retention ability of the area. In many European countries where excess nitrogen threatens to pollute surface waters, there is considerable interest in using wetlands as cost-effective nutrient reduction systems. Studies in Sweden, where there is a need to reduce the nitrogen load to the Baltic Sea by 50 per cent, estimate the cost of reducing each kilogram of nitrogen entering the Baltic would be only US$1 for wetland restoration, compared to $53/kgN for remedial measures in agriculture, and US$15 to $31/kgN for an increase in sewage treatment plant efficiency. Thus, stream-side wetlands and swamp forests can be an efficient solution to a regional problem.

Small-scale technology

The technology required for stream restoration is small-scale and has low maintenance and operating costs. This makes it especially useful for farmers in developing, as well as developed countries. It is based on common sense notions, which anyone living close to the land and depending on its long-term quality can readily comprehend. In Sweden, with its high standard of living and even higher taxation rate, farmers walk a tightrope between effective long-term farming practices and short-term economic constraints. However, in most European countries farmers know that new, simpler land management practices must be adopted. The landscape is currently being used only for short-term, non-sustainable economic gain, and they are concerned at the long-term crises to which this may lead.

When the Building Block model is explained to farmer cooperatives in Sweden, the response is usually positive. "Yes, we have to do something, don't we?" Then the details of how are discussed: "I can build a pond, but there is no place on my land for a swamp forest". This is how the Building Block model was developed - in discussions with farmers, as a "self-serve" system tailored to fit the individual farmer's financial resources and the landscape of his farm. Adopting any of the measures will bring positive benefits.

The orientation of the model and measures described were aimed originally at lowland streams, but the approach can be used to restore any stream and should be seen as a general strategy for both landowners and governments. Other restoration measures specific to regional problems and topography can be added to those already mentioned. The list should not be regarded as exhaustive or final, but as the starting point for many solutions.