The Impacts of Dams and Other Impoundments on Riverine and Riparian Ecosystems

Dams are engineered structures with a single purpose of holding water back; although there are many reasons why we want to do that. Dam building began alongside the Agricultural Revolution approximately eight thousand years ago for irrigation (Joyce, S. 1997) and today irrigation is responsible for approximately half of all dams registered with the World Register of Dams (, retrieved 01 March 2012). Other purposes for dams include hydropower, water supply, flood control, navigation, recreation and fish breeding (13). Dams are classified according to size, structure, or purpose.

Unfortunately there is no global accounting of dams and impoundments, however the World Register of Dams through the International Commission of Large Dams or ICOLD, has 92 country members who can register dams measuring over 15 m in height (13). According to this registry, there are approximately 38,000 large dams globally, many of which are devoted to a single purpose with a growing trend of multipurpose dams (we humans are so efficient). Within the United States, there is the National Inventory of Dams, hereafter NID, maintained by the Army Corps of Engineers (10) with an accounting of almost 80,000 dams in the United States. NID consists of all dams with a height greater than or equal to 25 ft or 8 m and at least 15 acre-ft of storage, dams with a significant or high hazard in the event of failure, or 50 acre-ft of storage and greater than 6 ft or 2 m in height. There are many dams and impoundments in the United States not accounted for, as they do not meet these requirements and their cumulative storage is small when compared to the larger structures (Graf, 1999).

As mentioned above, dams have many purposes, the most important being water storage for irrigation. According to ICOLDstatistics, approximately 40% of the world’s food is irrigated with water from a reservoir and with the combination of an increasing population and climate change; water storage for irrigation is an immeasurable benefit of dams. Freshwater is scarce on our planet, accounting for only 3% of the water, with only a fraction of that available to humans, see Figure 1 below. In addition to this scarcity there is an unequal geographic distribution of freshwater AND natural seasonal and climatic variations; and the absolute necessity of dams becomes obvious.

Figure 1. Detailing water distribution on our planet, from 10

There is a catch, of course. While dams provide immeasurable benefits for humans, they also provide immeasurable harm to the ecosystems containing them. Dams fragment rivers, upstream from downstream and rivers from floodplains; in some cases irreversibly harming the ecosystem. The purpose of this paper is to provide an overview of the impacts dams have on these ecosystems and some of the restoration efforts available to mitigate these impacts. I caution the reader to keep in mind this is an extremely intricate, complex and dynamic topic, dams provide so many benefits and cost us so much at the same time.

The reasons for utilizing this media of a wiki are that it allows for a nonlinear, constantly evolving conversation between the creator of this page (me, and you the reader. Overall, wikis allow for self-education and collaboration between many people who are interested in this topic. Time and time again in this paper the reader will notice highlighted words, please follow the links to gain more information regarding the topic. More philosophically, the integrated nature of wikis somewhat mimics the integrated nature of watersheds.

Natural Flow
Before I can detail the many ways dams harm ecosystems, it is helpful to know what a healthy, natural riverine and riparian system looks like. Rivers are dynamic, constantly changing spatially along their course from mountain headwaters to ocean deltas and constantly changing temporally; seasonally with annual precipitation patterns and over longer time frames with the ever-changing climate. In 1997 a paradigm shifting paper was published (Poff, et al, 1997) titled The Natural Flow Regime: a paradigm for river conservation and restoration, detailing what a natural, healthy river ecosystem looks like. According to these workers, five components make up the flow regime of every river: magnitude, frequency, duration, timing, and rate of change. Magnitude is the rivers discharge or volume of water flowing past a point per unit time. Frequency is how often a specific magnitude or more occurs. Duration refers to the time period associated with a certain flow, for example, how long a specific discharge is maintained. Timing can also be stated as predictability, or how regular specific discharges can be. The last component is rate of change, or flashiness; meaning how quickly does a river change from one flow rate to another. Combining these five components allows a river’s ecosystem to be defined and there are an infinite number of combinations. Figure 2, taken from Poff, et al, 1997, details flow histories of four very different rivers. Fig. 2a, Augusta Creek, MI, displays a river with large groundwater inputs. Fig. 2b, Satilla River, GA, shows a system with large variability and low predictability. Fig. 2c, Upper Colorado River, CO, also show large fluxes in magnitude, but a high seasonal predictability. Fig. 2d, South Fork of the McKenzie River, OR, displays a system with large magnitude flux during the rainy season and a large groundwater input. These five components describe the flow regime; the driving force behind the physical structure of a river environment and physical structure is essentially habitat. Just like the five components mentioned above, the physical structure and habitat of river ecosystems are also dynamic and the many species that live in and depend on these environments have evolved to take advantage of this dynamism.
Figure 2. 3D hydrographs detailing some of the natural variability that exists in all river systems. From Poff et al, 1997.

Fig. 2 illustrates the many magnitudes, but in each hydrograph there is a peak discharge, or range of peak discharges, in other words flooding. These relatively large volumes of water are responsible for moving or redistributing large volumes of sediment, nutrients and organic material through the system and can be described by the Flood Pulse Concept (Junk, et al, 1989). Also shown by the hydrographs in Fig. 2 is that dynamism can be predictable (Fig. 2c) or unpredictable (Fig.2b). Every river has a unique flow regime so it follows that impacts due to dams will also be unique; based on the river itself and the purpose of the dam.

Fragment: Upstream from Downstream
A dam creates a disconnect on a river, segmenting once continuous reaches into upstream and downstream components. The impacts are also divided; upstream impacts are quite different than downstream.

Immediately upstream a reservoir, or man-made lake, is created. The most obvious impact is habitat inundation and loss; what was terrestrial becomes aquatic. The impact to local biota can be devastating, especially if the new reservoir is in a headwaters, mountainous region, where speciation is common in valleys (Nilsson and Berggren, 2005). Riverine and riparian zones are frequently important migratory pathways for many species and reservoirs usually disrupt these paths (, Environmental Impact of Dams, retrieved 30 January 2012). This has been well studied in regards to many migratory fish.

Rarely is the flooded area cleared of vegetation prior to flooding and this rotting vegetation can lead to oxygen depletion in the bottom of the reservoir (Baxter, 1977). Deoxygenated water is suspected to be the reason behind the more than 120,000 dead fish found downstream of the Yacyreta Dam on South America’s Uatuma River in 1994 (14). Soluble material, usually held in the soil, leaches into the reservoir waters and can change the water chemistry markedly, depending on the amount of soluble material and the retention time of the water (Baxter, 1977). This leaching combined with the degassing of rotting vegetation leads to a release of greenhouse gases; CO2 and CH4 (Carbon Dioxide and Methane) (Nilsson and Berggren, 2005) contributing to climate change. The toxic metal mercury, formerly bound in the soil, can be leached and changed into methylmercury, a toxin that bio-accumulates in fish (Nilsson and Berggren, 2005). The leaching can also release an increased amount of nutrients, Phosphorus and Nitrogen, leading to a temporary increase in aquatic productivity that cannot be sustained (Nilsson and Berggren, 2005).

Reservoirs usually display a different shoreline than that of natural lakes; longer and less circular with either a dendritic effect if many tributaries are present, or long and narrow if the river banks were high (Baxter, 1977), as shown in Fig. 3 below. Lake Nasser, created by Aswan High Dam Reservoir, is located in Egypt and Sudan and is 500 km long (
The image below shows Lake Nasser’s many dendritic arms, known as khors, that developed as the reservoir flooded the Nile River’s tributaries. The water’s metallic white color, called sunglint, is an optical effect created by the Sun’s light reflecting off the water’s surface and into the camera lens.
Figure 3. Lake Nasser in Egypt, one of the largest reservoirs in the world, created by the Aswan High Dam on the Nile River. (, retrieved 05 February 2012).

The shoreline area of lakes and rivers, known as the littoral zone, is usually an area of high biodiversity and productivity, as are all transition zones. In rivers, the littoral zone biota have evolved to live in conditions found on the banks of rivers; long dry periods that experience periodic flooding (Baxter, 1977). Conditions for a reservoir’s littoral zone are dependent on the purpose of the dam and range from rapid fluctuations in water depth, as in the case of a hydropower reservoir; to long periods of flooding with periodic dry episodes (Baxter, 1977) for a recreational reservoir. The unnatural cyclicity of drawdown and flooding keeps the biotic communities in an immature state as they are constantly trying to re-establish themselves (Baxter, 1977).

In addition to an unnatural shoreline, the longitudinal cross- section of reservoirs is quite different from that of lakes. In a natural lake the deepest areas are usually found near the center; while in reservoirs the deepest area is found adjacent to the dam itself. This can lead to surface currents expending their energy either downward or toward the dam, instead of in the shallow areas (Baxter, 1977).

The creation of a reservoir increases the surface area of water and leads to an increase in evaporation. This water is now lost to downstream ecosystems and processes. Approximately 10% of water stored in Lake Nasser is lost each year to evaporation (14). Evaporative losses on the Colorado River removes approximately one third of the river's flow, this has led to salinization of the water, further stressing the ecosystem (14).

Rivers carry sediment and when moving water hits standing water the sediment will drop out, forming a delta with the coarsest materials dropping out of suspension first, followed by the fines. If there is a low gradient, the depositing sediment can plug the delta, forcing deposition to occur further upstream. The sediment is trapped behind the dam and reduces the available storage capacity of the reservoir (Joyce, 1997). Nutrients adsorb to sediment and become trapped behind the dam with the sediment leading to eutrophication of the waters, further exacerbating the anoxia, or deoxygenation mentioned previously.

The downstream impacts of a dam are reverse of the upstream; whereas the area just upstream of the dam is flooded, downstream no longer experiences flooding, a natural variability in the flow regime necessary for a healthy ecosystem (Junk, et al, 1989). The reservoir traps sediment and nutrients, plugging the upstream channel while decreasing storage capacity of the reservoir and the downstream waters become sediment starved (Baxter, 1977).

The clear water leaving a reservoir will try to regain equilibrium with regard to its sediment load by eroding the bed and banks downstream of the dam with many varied consequences. Just downstream of the dam this can lead to bed armoring, a process whereby all the finer sediment grains are picked up by the starved water, leaving only the coarse grains such as pebbles and cobbles. This coarsening of the riverbed diminishes habitable area for river species that shelter in the interstitial spaces between grains (Poff, et al, 1997). As degradation of the riverbed and banks continues, the channel may incise leading to downstream tributary headcutting as they attempt to maintain base level with the main channel (Poff, et al, 1997). A consequence of tributary headcutting is increased sediment in the main channel, which no longer has the discharge necessary to move the sediment and aggradation occurs, further hindering flow (Poff, et al, 1997). Another consequence of reduced discharge is channel simplification; a process of decreasing sinuosity and a reduction in mid channel bars and islands (Ligon, et al, 1995), as shown below in Fig. 4. Fig. 4a is a stretch of the McKenzie River in Oregon, USA showing multiple sinuous channels wrapping around islands. Fig. 4b is a small reach of the McKenzie River at two times. The dark blue details the channels in 1967, four years after a dam was built upstream, with islands shown in green. Shown in light blue is the same river reach in 1990, with only one channel and no islands. The river no longer has the high discharges necessary to move sediment, especially coarse gravels. One result was a decrease in salmon spawning and juvenile habitat (Ligon, et al, 1995) dependent on gravel bars. Reducing the physical diversity of a river channel reduces the life it can support.
sinuosity.gif sinuosity2.gif
Figures 4 a & b. The McKenzie River in Oregon, USA. Taken from Ligon, et al, 1995.

Just as high flows are necessary to transport and redistribute sediment and nutrients, they are also responsible for importing woody debris into the channel and transporting them downstream. Large woody debris piles serve to dissipate energy in their immediate vicinity and further trap moving materials creating habitat (Naiman and Decamps, 1997). Should the river still manage to reach the ocean, loss of sediment will result in smaller deltas or possibly even coastal erosion (Nilsson and Berggren, 2000). Southern California has a costly program of dredging sediment from offshore to supply their beaches with sand that was once supplied by rivers (14).

Fragment: River from Floodplain
Large discharges not only serve in channel processes, but also keep a river connected to its floodplain. Within civilization, floods are usually thought of as destructive, the most well known exception being Ancient Egypt. High overbank discharges, or floods, have many benefits to the riparian zone, including fertilization of river and floodplain, aquifer recharge, reviving floodplain topography, dissipating energy, habitat creation and plant germination and seed dispersal. Riparian zones are the transition between aquatic and terrestrial ecosystems; as a transition zone, they experience high biodiversity (Naiman and Decamps, 1997).

When overbank flooding occurs, the riparian zone is inundated and the river distributes sediment and nutrients to the floodplain. The Nile River in Egypt is famous for annual flooding, distributing sediment and nutrients along the floodplains, which allowed for productive agriculture. Since completion of the High Aswan Dam and subsequent sediment capture behind the dam, Egyptian agriculture now depends on artificial fertilizers in addition to losing approximately 1mm of new soil each year (14). While the river is fertilizing the floodplain, the floodplain is also fertilizing the river. High flows scour the riparian zone soils, carrying organic matter and nutrients from the floodplain into the main channel, perhaps providing necessary food for some organisms (Baxter, 1977).

The high discharges that produce overbank flooding also help recharge the aquifer locally, maintaining the connection between surface and groundwater. On the floodplains of the Murray River in Australia this disconnect has led to salinization of floodplain soils as the saline groundwater is no longer diluted with recharge from overbank flooding (Nilsson and Berggren, 2000). Should the river incise from lack of sediment, the water table will also lower near the river (14). Another consequence due to lack of flooding is a decrease in floodplain topography. Transition areas, like riparian zones, are areas of increased biodiversity and small topographical differences, such as 1 or 2 m, often result in increased biodiversity (Naiman and Decamps, 1997).

Inundated floodplains provide nursery habitat for many fish species living in the river and the occurrence of a high discharge signals many fish species to move forward in their life cycle (Poff, et al, 1997) either mating, spawning, or hatching. Some native fish species have evolved to take advantage of inundated floodplain as rich feeding grounds; cyclic flooding occurs at times of increased growth (Poff, et al, 1997).

Just as many of the species that live in the river have evolved to take advantage of the variability provided by a natural flow regime, so have the biota living on the floodplain (Poff, et al, 1997). Many of the trees need periodic floods for seed dispersal and germination (Nilsson and Berggren, 2000). The lack of flooding and/or unnatural timing should high flows still occur, weaken these communities, opening the door for exotic species to move in. In the Southwestern United States, salt cedar, a non-native species has colonized many riparian zones as native species are weakened by a lack of overbank flooding (Stromberg, et al, 2007). Salt Cedar was introduced to the United States for the purposes of bank stabilization, to control soil erosion, and as an ornamental plant, however it is now the third most prevalent woody riparian species in western US (Stromberg, et al, 2007). Heavily altered flow cycles, many times a direct consequence of dams, are not only associated with weakening native communities in the western US, but also with strengthening the salt cedar populations (Stromberg, et al, 2007).

Quantifying Impact
Across the United States there are approximately 80,000 large dams, with no accurate accounting of structures smaller than 2m, inventoried by the US Army Corps of Engineers response to the federal Dam Safety Act of 1972, known as National Inventory of Dams (NID) (10).
Figure 5. Locations of large dams in the United States. Taken from GWSP Digital Water Atlas (2008). Map 41: Dams and Capacity of Artificial Reservoirs (V1.0). Available online at

Figure 5 shows the geographic locations of large dams across the United States and at first glance, the unequal distribution is apparent, with the eastern portion showing a greater abundance. This may lead one to believe the same of impact, however the number of dams is simply that; a number and does not necessarily detail or quantify impact. A better measure of impact upon the system is potential reservoir storage (Graf, 1999), with larger reservoirs resulting in greater impacts. To truly evaluate the impact, the locations need to be put in the context of watersheds. The USGS has divided the US into 21 water resource regions based on large river basins or smaller basins with similar characteristics, shown below.
Figure 6. Water-resource regions of the contiguous United States. Taken from, retrieved 07 February 2012.

Combining these 2 data sets provides a relative measure of impact and demonstrates an unequal distribution of fragmentation and impact across the contiguous United States (Graf, 1999).

Figure 7a. Storage per area (km3/km2).

Figure 7b. Storage per mean annual runoff.

Figure 7c. Storage per person (km3).


Figures 7a-c. Taken from Graf, 1999. Geographic distribution of impact.

Fig. 7a details storage capacity per watershed resource region, measuring potential magnitude of potential change in flow or the potential for ecological impact (Graf, 1999) with the greatest fragmentation in California, Texas Gulf and South-Atlantic Gulf regions. Fig. 7b shows storage capacity per mean annual runoff, a measure of change in discharge because of storage, with the Great Plains, Rocky Mountains and Southwest regions showing the highest ratios and therefore the greatest change in river discharge (Graf, 1999). Fig. 7c is storage capacity per person, and again the distribution is unequal, with the northern Great Plains, Rocky Mountains and Southwest experiencing high ratios. These areas export water and water-related services but endure the most of environmental costs (Graf, 1999). The greatest impacts are experienced in the Rocky Mountains, Great Plains and Southwest regions (Graf, 1999), not where the majority of dams are located, but regions that experience less annual precipitation and runoff. Consequently, these are the regions least likely to remove dams, they are necessary for flow regulation.

Restoration Efforts
Dam Removal
A new movement began in the 1990’s; a growing recognition of the mounting environmental costs of dams and a desire to restore ecological integrity to our watersheds by reconnecting upstream with downstream and river with floodplain (Hart and Poff, 2002). Many dams are beginning to show their age (10) and are no longer economically viable (Pohl, 2002). Many dams are subject to a review of potential violations of the Endangered Species Act or the Clean Water Act (Pohl, 2002) and have been found lacking, but many dams are necessary to continue the quality of life now taken for granted in the US. We are experiencing changing social values and placing more emphasis on the value provided by an intact and healthy ecosystem. However, we have also learned from our mistakes and are not rushing into dam removal; it is not as simple as tear down the dam and let the river heal itself. Any extreme change to an ecosystem needs to be assessed prior to action (Hart, et al, 2002). When a dam is removed there are changes in the hydraulic parameters and geomorphic processes, both upstream and downstream of the dam (Hart, et al, 2002). Captured sediment building up in a reservoir for years will suddenly begin moving downstream and can take years to attenuate (Shuman, 1995), with subsequent depositional processes and channel changes (Hart, et al, 2002), which is important to consider because the physical structure of the river controls habitat (Poff, et al, 1997). Upstream will experience an increased connectivity and perhaps colonization by migratory biota (Hart, et al, 2002). There can be significant effects to local wildlife that have adjusted to the unnatural flow regime and reservoir, shown below in Fig. 8. In other words, the dam and reservoir created an environment which the local biota learned to live with.
Figure 8. Spatiotemporal ecosystem changes following dam removal. Taken from Hart, et al, 2002.

There have been many dam removals with various impacts to ecosystems, shown in Table 1.
Table 1. Case Studies of completed or proposed dam removals. Taken from Bednarek, 2001.

Most have experienced positive outcomes, with local biota returning to a more natural state, most notably for sediment and fish movement. Recently, in September of 2011, two dams on the Elwha River, Washington, US, began removal processes, expected to last approximately 3 years. One of the dams is 210 ft tall, the largest dam in the US, on record, to be removed yet. The Elwha River has experienced a loss to its salmon fisheries due to dams blocking salmon from moving upstream, hopefully, with a return to a more natural flow regime, the salmon runs will be restored (
(, retrieved 28 March, 2012). The process can be monitored via live streaming video.

However, dam removal is not always the best option, or is not economically viable. In some cases, mitigation is a viable alternative. For many decades, workers have been studying river ecosystems and while a natural flow regime might not be known or achievable, structural and/or operational changes can be made in the interest of a healthy ecosystem, a concept known as environmental flows.

In 1996, a 7-day experimental controlled flood was released from the Glen Canyon Dam in Grand Canyon, USA. The maximum flow allowed was 1274 m3/s, lower than most pre-dam spring floods, but still high enough to build beach habitat (Patten, et al, 2001). Downstream of Glen Canyon Dam the impacts were typical of an impounded river, stemming from decreased flow rates and lack of sediment. The experimental flood was a practice in adaptive management; testing the release of higher flows without negatively affecting other canyon resources. The test flood did scour and redistribute sediment, as shown below, and positively influenced in-channel sediment deposition patterns.
Figure 9. Before and After Image of experimental flood in the Grand Canyon of the same side bar. As shown, surface area of the side bar increased as well as habitat. Images taken by Mark Schmeeckle of Arizona State University.

Another example of mitigation can be found with the Tennessee Valley Authority’s Reservoir Releases Improvement Plan, RRI. In 1991, TVA decided to practice an adaptive management style with their dam operations downstream of 9 dams to increase minimum flows and improve water quality (Bednarek and Hart, 2005). The results were mixed, with some benefits limited to small areas and only then for a short time, but there was an increase in both average dissolved oxygen and average velocity. Macroinvertebrate family richness increased, while pollution tolerant macro-invertebrates decreased. Overall, there was a shift toward pre-dam conditions (Bednarek and Hart, 2005).

Dams and reservoirs are a necessity, providing immeasurable benefits and harm at the same time. For 80 centuries we have been building dams for various purposes and they are now an integrated part of our society, an engineering feat we cannot live without. Only within the past half century (of the full 80 centuries) has the damage and harm to ecosystems become obvious to us (I don't want to say we're obtuse, but...), the full extent of which has yet to be realized. With this knowledge comes a promising desire to offset the harm, either via dam removal or mitigation, with the hopes of the writer that the full extent of damage will never be realized. As we face an uncertain climatic future dams will become even more of a necessity than before and as such the need to offset harm also becomes more necessary. Dams are not going away but we can mitigate some of the damage.


1. Baxter, R. M., 1977, ENVIRONMENTAL EFFECTS OF DAMS AND IMPOUNDMENTS: Annual Review of Ecology and Systematics, v. 8, p. 255-283.

2. Bednarek, A. T., 2001, Undamming Rivers: A Review of the Ecological Impacts of Dam Removal, Environmental Management, v. 27, no. 6, p. 803-814.

3. Bednarek, A. T., and Hart, D. D., 2005, Modifying dam operations to restore rivers: Ecological responses to Tennessee River dam mitigation: Ecological Applications, v. 15, no. 3, p. 997-1008.

4. Graf, WL, 1999, Dam nation: A geographic census of American dams and their large-scale hydrologic impacts, Water Resources Research, v. 35, no. 4, p. 1305-1311.

5. GWSP Digital Water Atlas (2008). Map 41: Dams and Capacity of Artificial Reservoirs (V1.0). Available online at

6. Hart, D. D., Johnson, T. E., Bushaw-Newton, K. L., Horwitz, R. J., Bednarek, A. T., Charles, D. F., Kreeger, D. A., Velinsky, D. J., Dam Removal: Challenges and Opportunities for Ecological Research and River Restoration, BioScience, v. 52, no. 8, p. 669-680.

7. Hart, D. D. and Poff, N. L., 2002, A Special Section on Dam Removal and River Restoration, BioScience, v. 52, no. 8, p. 653-655.

[[8.%20http:/|8.]], retrieved 05 February 2012.

9., retrieved on 01 March 2012.

10., retrieved 07 February 2012.

11., retrieved 07 February 2012.

12., retrieved 06 February 2012.

13., retrieved 01 March 2012.

14., Environmental Impact of Dams, retrieved 30 January 2012.


16. Joyce, S. 1997, Is It Worth a Dam? Environmental Health Perspective, v. 105, no. 10, p. 1050- 1055.

17. Junk, W. J., Bayley, P. B., Sparks, R. E., 1989, The flood pulse concept in river-floodplain systems, Proceedings of the International Large River Symposium, ed. D. P. Dodge, Can. Spec. Publ. Fish. Aquat. Sci., v. 106, p. 110-127.

18. Ligon, F. K., Deitrich, W. E., Trush, W. J., 1995, Downstream Ecological Effects of Dams, BioScience, v. 45, no. 3, p. 183-192.

19. Naiman, R. J., Decamps, H., 1997, The Ecology of Interfaces: Riparian Zones, Annual Review Ecology Systems, v. 28, p. 621-658.

20. Nilsson, C. and Berggren, K., 2005, Alterations of Riparian Ecosystems Caused by River Regulation, BioScience, v. 50, no. 9, p. 783-792.

21. Patten, D. T., Harpman, D. A., Voita, M. I., and Randle, T. J., 2001, A managed flood on the Colorado River: Background, objectives, design, and implementation: Ecological Applications, v. 11, no. 3, p. 635-643.

22. Poff, N. L., Allan, J. D., Bain, M. B., Karr, J. R., Prestegaard, K. L., Richter, B. D., Sparks, R. E., Stromberg, J. C., 1997, The Natural Flow Regime: A paradigm for river conservation and restoration, BioScience, v. 47, no. 11, p. 769-784.

23. Pohl, M. M., 2002, Bringing Down our Dams: Trends in American Dam Removal Rationales, Journal of the American Water Resources Association, v. 38, no. 6, p. 1511-1519.

24. Shuman, J. R., 1995, Environmental considerations for assessing dam removal: alternatives for river restoration, Regulated Rivers, v. 11, no. 3-4, p. 249-261.

25. Stromberg, J. C., Lite, S. J., Marler, R., Paradzick, C., Shafroth, P. B., Shorrock, D., White, J. M., White, M. S., 2007, Altered stream-flow regimes and invasive plant species: the Tamarix case, Global Ecology and Biogeography, v. 16, p. 381-393.