According to Dynesius and Nilsson (1994), 77 percent of the 139 largest river systems in North America (north of Mexico), in Europe and in the republics of the former Soviet Union are affected by dams, reservoir operation, interbasin diversions, and irrigation. It is obvious that humans have impacted rivers since the beginning of civilization in the valleys of the Yellow, Indus, Tigris-Euphrates, Ganges, and Nile rivers (Figure 1.2). In fact, the earliest dam was constructed on a tributary of the Nile south of Cairo about 5000 years ago (Schnitter, 1994).

Graf (2001) calculates that 79 percent of American rivers are affected by humans (Table 8.1), and he developed a scale of naturalness that is illustrated by five American rivers (Table 8.2). Much of the fragmentation of rivers in the United States is due to 80 000 dams that have been constructed on these rivers.

Wohl (2001) states that when she moved to Colorado in 1989 she “was impressed by the sparkling water of the mountain rivers, … and assumed that these were natural fully functional rivers.” However, when she “began to read historical accounts of the Colorado Front range and to examine the streams more closely” she “realized how dramatically they had been altered.” She began to “think of them as virtual rivers, which had the appearance of natural rivers but had lost much of a natural river's ecosystem functions.

In Chapter 8, the consequences of human activities on rivers were discussed in passing. Here, some unintended consequences of river modification will be considered. For example, the meander cutoff program on the Mississippi River resulted in a straighter navigation channel but it also resulted in bank instability and the expenditure of vast amounts of money to stabilize the straighter, steeper channel. In addition, the changes of the Platte River, as its hydrology was altered (Chapter 7), and the increased flood stages of the Middle Mississippi River (Chapter 8) as the channel was confined by levees, are all examples of the unintended consequences of human activities.

Even the increase or decrease of gravel in a channel can significantly impact a river. Students of gravel bed rivers are well aware of the protective qualities of a gravel armor. It is perhaps less well recognized that small amounts of gravel in sand-bed rivers can have significant sediment transport and morphological effects. The Missouri and Nile rivers will be used to illustrate other impacts of small amounts of gravel on the morphology and response of sand-bed rivers.

For example, development of an armor significantly reduces bedload transport in a Swiss river. This was demonstrated experimentally by Begin et al. (1980). In a large flume filled with sediment that contained only 1.2 percent of sediment larger than 2 mm, base-level was lowered in order to determine how sediment production changed with channel incision.

Before considering the variability of a single river, it is necessary to consider the different types of rivers that exist (Table 2.1). Once a topic is sufficiently comprehended, it appears logical to develop a classification of its components. A classification can provide a direction for future research, and there have been many attempts to classify rivers (e.g., Schumm, 1963; Mollard, 1973; Kellerhals et al., 1976; Brice, 1981; Mosley, 1987; Rosgen, 1994; Thorne, 1997; Vandenberghe, 2001). Indeed, Goodwin (1999) thinks that there is an atavistic compulsion to classify, and indeed, an individual's survival may depend on an ability to distinguish different river types (deep versus shallow).

Depending upon the perspective of the investigator, a classification of rivers will depend upon the variable of most significance. For example, the classic braided, meandering and straight tripart division of rivers (Leopold and Wolman, 1957) is based upon pattern with boundaries among the three patterns based upon discharge and gradient. Brice (1982, 1983) added an anabranched or anastomosing channel pattern (Figure 2.1) to the triad and distinguished between two types of meandering channels (Table 2.1). The passive equiwidth meandering channel is very stable as compared to the wide-bend point-bar meandering channel (Figure 2.2). This is a very important practical distinction between active and passive meandering channels (Thorne, 1997, p. 188). A highly sinuous equiwidth channel gives the impression of great activity whereas, in fact, it can be relatively stable (Figure 2.3).

Except for the mountain streams discussed in Chapter 2, all of the other channels were alluvial and in regime. Therefore, it is appropriate here to spend some time considering channels that are responding to altered conditions by eroding, depositing, avulsing, and changing pattern – i.e., non-regime channels.

Rivers respond to altered conditions, therefore variability through time is important because a river or a reach may be out of character for a period as it adjusts. For example, Table 2.2 suggests how different types of channels respond to change. A list of 16 responses of channels to change, and the four major variables that influence them are summarized in Table 3.1. Time (history) is included with discharge (increase or decrease), sediment load (increase or decrease), and base level change (up or down), because channels change naturally through time, and time is an index of energy expended or work done. The changes are grouped according to the results of the change (erosion, deposition, pattern change, and metamorphosis). In Table 3.1, the changes that will be affected by the passage of time or by a change of discharge, sediment load or base level are indicated by an X. A brief discussion of each of the responses follows, but two – incision and avulsion – will be considered in detail because of their potential for causing serious problems.

Tectonics is an important determinant of river type (Figure 1.2). It generates the relief that drives the erosional machine; it generates earthquakes, which provide vast amounts of sediment to a river; and it causes avulsion and gradient changes. All of the above influence river type and behavior. Significant effects of tectonics are visible in the Amazon basin and on the Hungarian Plain (Schumm et al., 2000), and as an upstream control its downstream impacts can be significant. The effects of local, active tectonics, which impact limited reaches of a river, are discussed in Chapter 11.

High relief mountains, that are tectonically active and that were or are glaciated produce large quantities of coarse sediment, which in turn, produce wide, high width–depth ratio, braided bedload rivers. In contrast, low relief, stable mountains, especially in humid regions, produce low width–depth ratio, sinuous or straight channels.

As is generally the case, the effect of relief and slope on sediment yields and hence upon the type of river found downstream is modified by other variables, such as lithology and climate (Figure 1.2). Ahnert (1970) collected data for 20 European and US river basins with a range of relief from 89 m to 2869 m. He concluded that there is a direct correlation between relief and rate of erosion and sediment yield. In contrast, for small drainage basins in western, semiarid US the relation is exponential (Figure 5.1).

Floods (Figure 1.2) normally affect almost all of the length of a river but, of course, there are exceptions. For example, the flood may only impact the downstream reaches of a channel or an upstream flood may have its effects dissipated downstream. However, except for extreme events that not only modify the channel, but also the valley, the impacts are ephemeral, being lost as the channel readjusts during floods of lesser magnitude.

Floods produced by local events such as the failure of landslide dams will be discussed in Chapter 15. Here are considered hydrologic events caused by climatic conditions. Wohl (2000b, p. 167) states these conditions as: “A flood may cause dramatic changes along some reaches of a channel and have relatively little effect on other reaches. Similarly, a flood that occurs once every hundred years may create erosional and depositional forms that are completely reworked within 10 years along one channel, but that persists for decades along a neighboring channel.”

Stream channels in eastern Australia decrease in size downstream contrary to hydraulic geometry rules (Nanson and Young, 1981a, b). The reason is that the floods go overbank downstream and the channels convey only part of a flood. Also in Australia, Warner (1987a, b) identified periods of higher and periods of lower average rainfall and flooding. These flood-dominated regimes (FDR) and drought-dominated regimes (DDR) tend to persist from 30 to 50 years. The FDR have mean annual discharges from two to four times greater than DDR.

Accidents (Figure 1.2), as discussed here, are unanticipated events. Most floods are not included (see Chapter 13) because they can be anticipated. For example, 100-year floods do eventually occur. Accidents are earthquakes, log jams, ice jams, avalanches, and volcanic eruptions, which are not normally anticipated and yet they produce dramatic channel changes and catastrophic floods.

Log jams and snags

Trees that fall into a channel can have a marked effect. In small channels, single trees can obstruct flow, thereby causing a back-water effect and sediment deposition. When the trees rot or are otherwise removed, incision into the sedimentary deposit will produce discontinuous terraces (Figure 15.1). Similar results occur as a result of beaver dam construction (Bigler, 2001). In a sinuous channel, the back-water effect of log jams can cause cutoffs. In mountain streams evidence of former jams is a major increase of channel width and the presence of mid-channel bars (Keller and Swanson, 1979).

Even larger rivers in forested regions can be blocked by numerous fallen trees (snags) causing an anastomosing pattern. For example, in 1872 the Willamette River in Oregon frequently had four or five channels. Unlike today's trees, the size of snags in the Willamette River ranged from 30 to 60 m long and up to 2 m in diameter. Clearing of the snags produced a single channel (Sedell and Froggatt, 1984).

It seems very logical that base-level lowering (Figure 1.2) will rejuvenate a drainage network, deliver large amounts of sediment downstream, and cause major channel change. Support for these conclusions comes from experimental studies (Holland and Pickup, 1976; Schumm et al., 1987) and from observations of the headward incision of arroyos, gullies and channelized streams (Schumm et al., 1984). Fisk (1944) concluded that Pleistocene sea-level lowering caused excavation of alluvium and bedrock incision by the Mississippi River for a very long distance upstream. On a smaller scale even a single meander cutoff can cause local steepening and upstream channel degradation (Winkley, 1977). Lane (1955) an eminent river engineer concluded that, when a river is affected by a base-level rise or fall (Figure 16.1), the river will degrade or aggrade to restore its equilibrium profile. After all, the channel must continue to carry its load of sediment with a given water discharge, and this requires a given gradient that must be restored either by deposition (Figure 16.1a) or by erosion (Figure 16.1b).

In contrast to Lane, Leopold and Bull (1979) concluded that base-level changes affect the vertical position of a river only locally and to a minor extent. They argued that not only is stream gradient change important, but that channel pattern, roughness and shape (Rubey, 1952) can also adjust, in order to absorb the effect of base-level change.

In Chapter 5, the effect of tectonics on sediment yields and river systems was considered. Here, the focus is on local effects of deformation of the valley floor (Figure 1.2). This can include not only active tectonics, but subsidence owing to groundwater and petroleum removal, and local uplift as a result of valley incision (Bell, 1999). These examples can be considered as pseudotectonics, but they can have the same effects as active deformation.

Active tectonics can take several forms. Deformation can be along faults or pairs of faults (horst and graben), which should have the same effect on rivers as a monocline, dome, or basin. In addition, an entire valley may be tilted upstream, downstream, or laterally. The possibilities are great, but, in reality, the primary effect of tectonics will be a local steepening or reduction of gradient or lateral (cross-valley) tilting. In order to adjust to these changes, the river must either degrade, aggrade, or change sinuosity. Small streams in the Mississippi River valley show clearly the impact of active tectonics on their morphology (Boyd and Schumm, 1995; Spitz and Schumm, 1997), and indeed, even large rivers, such as the Mississippi, Indus, and the Nile, change patterns in response to active uplift and faulting (Schumm and Galay, 1994; Schumm and Winkley, 1994). In addition to these primary influences, there will be secondary effects, as rivers respond to changed gradient (aggradation or degradation), and there will be tertiary effects, as decreased or increased sediment loads influence reaches downstream of the deformed reach and as aggradation or degradation in the deformed reach progress upstream.