Engineering Within Ecological Constraints (1996)

Hsieh Wen Shen

Decades ago, hydropower plants were considered the most environmentally sound means of generating electricity because they produce neither smoke nor nuclear waste. Gradually, however, the ecological consequences of dams and other river modifications have become appreciated. In particular, the environmental impact of the Aswan Dam in Egypt raised many concerns with regard to hydropower development. Gradually we learned the need to live in harmony with our environment. Streams are not just conduits for supplying water for human needs; they are also communities of species. This paper discusses the general goal of attempts to improve the ecological properties of rivers, describes alternative specific objectives, and reviews two Cases that provide insights into the potential for collaboration between ecologists and engineers.

The general goal for ecological development in a stream is to achieve a sustainable condition so that human beings can live in harmony with their environment. The World Commission on Environment and Development (known popularly as the Brundtland Commission) defined sustainable development as "development that meets the needs of the present without compromising the ability of future generations to meet their own needs" (World Commission, 1987). Many international and national organizations, including the United Nations, UNESCO, World Bank, and the Earth Council have held meetings to discuss various concepts of sustainable development and sustainability indicators. In general, ecologists treat "sustainability" as preservation of the natural function and status of the ecological system, whereas economists emphasize the maintenance and improvement of human living standards as indices for sustainability.

Two key questions arise:

1. How much should we emphasize human interests in defining ecological goals?
2. How do we define and achieve sustainability in the ecological aspects of stream systems?

No precise answer can be given for either of these questions. When human interests conflict with those of another species, one must judge the appropriate goal and approach for each problem. Sustainability with regard to stream ecology is also difficult to define. Maintenance may be required to sustain certain aspects of streams. The following examples may be instructive.

It is possible to distinguish four alternative objectives for projects intended to improve ecological conditions in streams and rivers.

1. Restore to natural condition or a selected previous condition. Ecological processes are dynamic. Some processes are known now and others may not be known now or ever. Ideally, one wishes to enhance ecological conditions to satisfy future needs, but unfortunately this is difficult to achieve, especially for diverse collections of species. A common approach is to attempt to restore a river basin to a natural or previous condition. It is assumed that if the future hydrological condition is the same as that of the selected previous condition, then the majority of the ecological criteria will be satisfied. It is generally not feasible to return a particular stream in the United States to its condition during pre-Columbian time. Rather one is more likely to achieve success by returning the stream to a more recent condition. This general approach is particularly attractive when directed at diverse biological communities. In this approach, one must investigate the previous conditions such as the frequency and the extent of flooding and the ranges of various flow parameters such as flow depths and flow velocity. The case study on the Kissimmee River is a good example of this approach.
2. Achieve certain specific ecological criteria. If the ecological criteria for certain species can be determined, then an alternative is to restore or maintain the river basin according to their specific needs, such as flow parameters required by different life stages, the magnitude and frequency of flooding, acceptable levels of sediment in the flow, and various water quality limits. A great deal of research is needed to understand the ecological criteria for many ecological elements under various hydrologic conditions. The case study of the Niobrara River (below) is a good example of this approach.
3. Maintain stable streams. The third, less satisfactory alternative is simply to design a stable stream system. This may or may not satisfy various ecological

3. criteria. A dynamic stream may be more suitable in a dynamic ecological evolution.
4. Follow a combination approach. Often, the restoration of a river basin to its natural condition must include considerations of both ecological criteria and the stability of the streams such that the three preceding approaches must be combined.

Regardless of the particular objective, successful enhancement of a stream will usually require integrated efforts by specialists from many fields. Public involvement is also critical. The following case studies demonstrate the importance of cross-disciplinary collaboration and public involvement.

The channelization project in the Lower Kissimmee River Basin, Florida, resulted in severe losses of river floodplain wetlands and waterfowl populations throughout the river valley. An earthen channel, the C-38 canal, is the main element of this project. Immediately after project completion in 1971, a strong effort was initiated to restore this basin to its prechannelization status. This large Kissimmee River restoration project, when it is completed, will be a milestone in our journey toward ecological harmony because the only major goal of the project is ecological enhancement. In 1986 engineers and scientists from the University of California, Berkeley, working with the South Florida Water Management District, developed a set of restoration options. Ecological goals for the restoration plans were formulated. Alternative restoration plans were evaluated for their potential to satisfy the ecological goals. Analyses were based on a combination of field data, physical modeling, and numerical modeling. Finally, backfilling of certain reaches of canal C-38 was recommended. Details of this study are provided in Shen et al. (1994).

In most rivers, the ecological environment is the result of long-term adjustments by countless complex factors. Many of these factors and their interrelated processes are extremely difficult, if not impossible, to define. Thus, ecologists stress the need to restore a river basin to natural conditions after the occurrence of man-made changes. Unfortunately, in the Kissimmee River Basin, human activities have changed the upstream conditions so much that the level of Lake Kissimmee cannot be allowed to fluctuate as much as it did before channelization. Thus, it is necessary to establish a set of ecological criteria as the targets for restoration, rather than attempt to recreate prechannelization conditions. Alternative restoration plans must then be rated on the basis of their likelihood to achieve these criteria.

As in many other restoration projects, it is difficult to establish generally accepted ecological criteria because people representing various concerns stress different ecological goals. Conflicting requirements may even be proposed by

the same group of people. Thus, the first task in our study of the Kissimmee River restoration was to search for a set of ecological criteria. These criteria must be ecologically justifiable and feasible, because if they are too restrictive then all feasible alternative measures can be eliminated. After many site visits, the principal investigator gradually satisfied the various interest groups that the study was sincerely searching for the most feasible plan for ecological restoration of the Kissimmee River. A symposium, suggested by the team and organized by the South Florida Water Management District, was held at Orlando in October 1988 to discuss various ecological and engineering concerns. This symposium was designed to focus on the restoration of the Kissimmee River ecosystem as a whole rather than individual species. A set of ecological restoration criteria was established. Their details will be discussed later. In essence, restoration requires that the floodplain receive flow relatively frequently to serve as a wetland. Also, floodwaters should return to the original river system slowly. These ecological criteria must be satisfied while simultaneously meeting independent constraints for flood control, navigation, and sedimentation. The main engineering approaches were to divert flows to the river's original course and its floodplain by blocking the C-38 canal with hydraulic structures such as weirs and earth plugs. Backfilling part of the canal was also considered.

The primary criterion for achieving environmental restoration goals is the reestablishment of prechannelization hydrology. Key characteristics of prechannelization hydrology were discussed in several papers presented at the Kissimmee River Symposium (Loftin et al., 1990). Critical hydrologic determinants of prechannelization ecological integrity were reduced to a form that could be used to compare alternative restoration plans. Key hydrologic criteria are given below.

1. Continuous flow with duration and variability comparable to prechannelization records. Historical data indicate that continuous discharge was a critical factor in the evolution and maintenance of biological communities in the prechannelization Kissimmee River ecosystem. Restoring the integrity of the Kissimmee River ecosystem depends on reestablishing the range of discharges of appropriate duration during representative (e.g., 10-year) postrestoration periods. Sheri et al. (1994) described the monthly discharge variations. The flow discharge characteristics thought to be necessary to restore biological communities that existed before channelization are (a) continuous flow from July through October, (b) highest discharges in September through November and lowest flows in March through May, and (c) a wide range of discharge variability. These features should maintain favorable levels of dissolved oxygen during summer and fall, provide nondisruptive flows for fish species during their spring reproductive period, and restore the temporal and spatial heterogeneity of river channel habitat.
2. Average flow velocities between 0.24 and 0.55 meter per second (mps)

2. when flows are contained within the river channel. Specifically, flow velocities within 60 percent of river channel cross sections must not fall below 0.24 mps for more than three consecutive days during July-September and 10 consecutive days during October-June. These velocities complement the discharge criteria by protecting river biota, because too little flow results in low concentrations of dissolved oxygen and excessive flow could interfere with important biological functions (e.g., feeding and reproduction of fish).
3. Overbank flow along most of the floodplain when discharges exceed 40-57 cubic meters per second. This criterion reinforces and will reestablish important physical, chemical, and biological interactions between the river and its floodplain.
4. Slow flood recession rates. A flood recession rate of less than 0.3 meter per month is required to restore the diversity and functional utility of floodplain/ wetlands, foster sustained river/floodplain interactions, and maintain river quality. Slow drainage is particularly important during biologically significant periods, such as the nesting season for wading birds. Rapid recession rates (e.g., rates that will drain most of the floodplain in less than a week) lead to fish kills and thus are not compatible with ecosystem restoration.
5. Floodplain inundation frequencies comparable to prechannelization hydrology, including seasonal and long-term variability characteristics. Shen et al. (1994) shows the prechannelization inundation frequencies for the floodplain adjacent to the Fort Kissimmee gauging station and provides guidelines for this criterion. For example, during a representative 10-year period, November stages should inundate 100 percent of the floodplain during 4 of the 10 years and 75 percent of the floodplain during 7 of the 10 years. When the entire floodplain is inundated, depths along the periphery of the floodplain should measure between 0.3 and 0.6 meter. These inundation levels will lead to redevelopment of floodplain structure and function and reestablish the floodplain as an integral component of the Kissimmee River ecosystem. Ecologically, the most important features of stage criteria are water level fluctuations that lead to seasonal wet-dry cycles along the Periphery of the floodplain, while the remainder (approximately 75 percent) of the floodplain is exposed to only intermittent drying periods that vary in timing, duration, and spatial extent. As stated in Loftin et al. (1990, p. 26):

Reestablishment of ecological integrity requires that all restoration criteria are met simultaneously. A piecemeal restoration program in which some of the established restoration criteria are achieved in one segment of the system, and other criteria are met in another portion, will not accomplish restoration goals and may be of little or no value. Game fish populations, for example, may still be limited if appropriate flow characteristics are restored but production of potential food resources on the floodplain is limited by inadequate inundation, or inaccessibility to river fish because levees or berms block the connectivity (interaction) between the river and floodplain. Alternatively, restoration of

Three alternative restoration plans have been selected by the District for analysis. In the Fixed Weir Plan, 10 weirs would be installed along the canal to divert flows into the river oxbows adjacent to the weirs. In the Level I Backfill Plan, the same canal reaches in which weirs would be installed in the Weir Plan would be completely backfilled between the two junctions with the oxbows. In the Level II Backfill Plan a specific, long, continuous canal reach would be backfilled. New river reaches would be created to maintain a continuous river reach with the same canal backfill reach.

Fixed Weir Plan: The advantage of using weirs is their flexibility in operation. Gated weirs can be opened during floods to decrease the need for flood levees or additional flooding fights.

During high floods the oxbow flow velocities would be between 2 and 3 mps. These velocities would cause erosion and deposition of sediments in the river oxbow reaches and could interrupt navigation. Channel maintenance probably would be needed after major floods.

Approximately 40-50 percent of the oxbow length (revitalized river channels) adjacent to the weirs would have flow velocities within the ecologically acceptable range of 0.24 to 0.55 mps. These oxbows were part of the active river system before the construction of C-38 canal. However, many more reaches would have velocities greater than 0.55 mps (with maxima on the order of 1.5 mps) than below 0.24 mps. About 26 kilometers of banks (revitalized river channels) would be inundated at flow discharges of 40 cubic meters per second at the entrance of the river and 57 cubic meters per second at the downstream end of the river. About half of this inundation length would occur at oxbows. Several existing control structures in the river, together with their levees, can control the recession rates in the lower 40 percent of the pools. However, the stage recession rate after floods, in at least 60 percent of all upper reaches of pools, would far exceed 0.3 meter per month. These rates could even reach 0.3 meter in a six-hour period. Complex flow operation schemes might be devised to control both inflows and outflows from each pool to reduce the recession rate.

Level I Backfill Plan: For this plan, during high floods, the oxbow flow velocities would vary between 1.5 and 2.1 mps. These velocities would cause erosion and deposition in the oxbows. During normal flows, between 18 and 33 percent of the oxbow velocities would be in the range of 0.24 to 0.55 mps. Except for periods of low flows, more flow velocities would be above 0.55 mps (with maxima on the order of 1.5 mps) than below 0.24 mps. About 16 kilometers of

banks (in revitalized river channels) would be inundated for flow discharges of 40 cubic meters per second at the entrance of the river and 57 cubic meters per second at the exit end of the river. All internal control structures together with their levees can only control the recession for the bottom 40 percent of the respective pools. At the upper 60 percent, the stage recession rate after floods would far exceed 0.3 meter per month. It is nearly impossible to design complex flow operation schemes to control both inflows and outflows from each pool to reduce the recession rate because the capacity for flood flow under this plan would be greatly reduced by the earth plugs blocking the canal. The amount of time that given proportions of the floodplain would be inundated under this plan is slightly less than would be inundated under the Fixed Weir Plan, and these values are far from meeting the requirements for ecological restoration.

Level II Backfill Plan: The Level II Backfill Plan should produce flow velocity ranges close to the ecological criteria if the future precipitation regime matches the precanal conditions and the flow regulation scheme can be properly designed. Historical data suggest that the flow properties of the Kissimmee River were similar in 1901 and 1958. In addition there was no detectable change in the river's course. During high floods, the oxbow flow velocities would vary between 1 and 1.3 mps and probably would not interrupt navigation. During normal flows, in 40 to 52 percent of the lengths of the oxbows, the flow velocities would be in the range from 0.24 to 0.55 mps, and in very few oxbows would the velocities exceed 0.55 mps. The flow velocity in all oxbows would be below 0.76 mps.

Upper lake level flood control capacity should be provided by leaving a portion of the Canal C-38 intact in the upper reach. Perhaps the best approach in this plan is to completely backfill part of the river reach. This should satisfy the flood criterion for Lake Kissimmee upstream from the Kissimmee River. Sedimentation problems can be investigated with a careful monitoring system. If significant movement of sediment occurs, appropriate actions should be taken. Construction would be carried out in several stages over several years. The extent of backfilling would be governed by the amount of available funding, available soil, the extent of the vegetation growth, and the possible relaxation of the water level constraint at Lake Kissimmee during flooding.

Field tests should be conducted at different times to monitor growth of vegetation. Knowledge of vegetation growth should be useful both to protect against erosion and to analyze hydraulic resistance. The data collected in the field during future years would determine the exact extent of backfilling.

A certain amount of dredging in the oxbows may be needed. In the Level II Backfill Plan, bank erosion and sediment deposition would be limited. Only a small amount of maintenance dredging in the oxbows may be needed after major flooding.

It is recommended that one type of plan be selected by the District from the

three plans for final engineering design and construction analysis. Certain detailed engineering analyses such as the extent and sequence of the backfilling, as well as the number and location of earth plugs, or weirs, are still needed. The possible effects of each tributary (slough) during flooding should also be investigated. Currently sufficient data are not available to make these analyses.

In accordance with the study team's recommendations, the state of Florida has requested funding from the U.S. Congress. At the direction of Congress, the U.S. Army Corps of Engineers, Jacksonville District, has investigated the team's work and converted it into an engineering feasibility project, which is awaiting further federal funding.

The U.S. Bureau of Reclamation planned to construct a water supply dam on a braided section of the Niobrara River near Norden, Nebraska. The primary objective was to supply irrigation water for local farmers. This plan was stopped by the courts after the U.S. Fish and Wildlife Service pointed out that whooping cranes, an endangered species, stop at that reach of the Niobrara River during their migration seasons. I was asked to develop an engineering design compatible with the requirements of the cranes. Because there are only about 100 whooping cranes, it is difficult to determine their ecological requirements, but they are believed to be similar to those of the sandhill cranes. Sandhill cranes prefer to rest on sandbars that are submerged by less than 0.2 meter of slow-flowing water amid vegetation shorter than 0.9 meter, and away from riverbanks. These conditions are required to enable the birds to forage and evade predators. Based on these requirements, we determined that the braided feature (a river reach with multiple anastomosing channels) was critical to provide shallow, vegetated sandbars at sufficient distance from the riverbanks.

We identified the flushing flow requirements necessary to maintain the braided stream. Certain portions of the stream below the proposed Norden Dam would be unavoidably changed from braided to meandering by the release of flows with relatively small amounts of sediment particles from the reservoir. We planned to maintain the braided characteristic of the river farther downstream by regulating the flow, controlling vegetation growth, and even breaking the ice jams that form annually. We reported our conclusions to both the Bureau of Reclamation engineers and Fish and Wildlife Service biologists. After the conclusion of our study, we were pleased to learn that the Fish and Wildlife Service agreed to withdraw its objection to the construction of the dam if the Bureau of Reclamation would follow our recommended flow release plans. The Bureau of Reclamation agreed to follow our recommendations. In the end the dam was never built because of a lack of funding. Like the Kissimmee Restoration Project, this study indicated that engineers and biologists can work together, and engineers can provide plans to satisfy ecological needs.

These two cases demonstrate the need for integrated efforts between engineers and ecologists to enhance our river management. We should seek to increase communication, conferences, and joint research across various disciplines. Every effort should be made for river engineers and stream ecologists to achieve a basic knowledge of each other's fields.

Loftin, M. K., L. A. Toth, and J. T. B. Obeysekera. 1990. Kissimmee river restoration: Alternative plan evaluation and preliminary design report. West Palm Beach, Fla.: South Florida Water Management District.

Shen, H. W., G. Q. Tabios III, and J. A. Harder. 1994. Kissimmee river restoration study. Journal of Water Resources Planning and Management 120(3) (May/June):330-334.

World Commission on Environment and Development. 1987. Our Common Future. Oxford, England: Oxford University Press.