Regime Shifts DataBase

Large persistent changes in ecosystem services

Freshwater river systems experience river channel position regime shifts when the main channel of a river abruptly changes its course to a new river channel. Meandering rivers are especially vulnerable to channel changes because of their high sinuosity which supports meander cutoffs. Other river types with a lower sinuosity can also experience river channel position shifts, for example braided rivers which have many small channels. Changes in river channel position can be understood as regime shifts when one considers the timescale at which they occur and the irreversibility of the shift. Large channel shifts (shifting the river position tens of kilometers) occur only approximately once a millennium. Within this timeframe humans settle along the river and establish complex social and economic structures which depend on the river. When a large, abrupt shift in the channel position occurs it results in huge disruption and the need for large-scale reorganization of the economy and societies that depend on the river. Once the river channel has shifted it will very rarely return to its former position, but instead become stabilized in its new position. In this context the alternate regimes are:

Old channel course

In this regime the river flows along a path which it has followed for many decades. People have therefore adapted to and have based their activities on this position of the river channel. For example, the floodplain area is often used by farmers to cultivate crops. Cities are located in the floodplain area, often protected by levees in case of a flood event. The mentality of the local inhabitants is that a channel shift should not be allowed to happen because their well-being is closely linked to the river in its current position. Therefore, large defense infrastructure is sometimes a characteristic of this regime (see also Tisza River case study).

New channel course

In this regime the river has switched its main course to a new path, often tens of kilometers from its previous course. Typically, riverflow is more rapid, and the length of the river is shorter. In the case of a meander cutoff, a so-called oxbow lake is formed. This happens when the meandering necks connect with each other and the abandoned part of the river after the cutoff is disconnected from the riverflow. The size of a cutoff and the resulting oxbow lake varies and depends on the size of the river. 

The main direct driver of the regime shift is strong floods, associated with large rainfall events. Such events cause the river to shift its course because they have enough power to break through a natural river levee or dyke, and to breach defense infrastructure such as a spillway that tries to control the riverflow (see Mississippi Case Study).

Human activities such as artificial channel widening, removal of debris, changes in the river course or cutting of vegetation play an important role in making a river more susceptible to channel shifts. The reason is that people may change the actual channel to improve transport or create a shortcut to decrease travel time and costs. Often, these activities might start very small by (e.g. removal of debris) and then slowly grow (e.g. by digging of a channel to make a shortcut) until a critical threshold has been reached and the river suddenly changes its course. In the case of the Ucayali River in Peru, an inconspicuous ditch of approximately one meter width was slowly but systematically widened which eventually led to a 71 kilometer cutoff after a strong flood event (see Ucayali River case study) with large negative impacts for the people living in the region. The destruction of dykes or levees can also cause a river channel to shift its position. For instance, a dyke at the Yellow River was destroyed to check the advance of the Jin army in China in 1128, leading to a shift in the river's course (see Yellow River Case Study). 

Floods are usually the direct cause of a shift in river channel position. However, the interplay of many different drivers is responsible for making a river vulnerable to changes in its course. The most common process is that over time sediment buildup gradually blocks the riverflow. A river always tries to take the shortest path and the steepest gradient due to gravity. When sediment is deposited on the riverbed, the gradient declines and slows down the current. This in turn leads to more deposition of sediment because of lower discharge until the river is blocked and spills to the side. This is a natural process, occurring approximately once a millennium in large rivers such as the Mississippi River (see Mississippi River case study).

In other cases, it is common that cutoffs occur at the meandering neck in rivers with high channel sinuosity. Erosion processes and scouring lead to channel incision and erosion of the river walls until the river suddenly breaks through the meandering neck, forming a cutoff. The river is then straightened and a so-called oxbow lake forms. This process is especially pronounced in rivers with a high sediment loading, for example the Amazon River in South America. Cutoffs vary in size depending on the size of the river and can occur alone or consecutively within a short period of time. Such shifts can be linked to the theory of self-organized criticality (SOC). According to SOC, cutoffs occur when the sinuosity of the river meander increases to a critical threshold at which the cutoff occurs.

A crevasse splay can also be responsible for a river channel position shift. It occurs when a natural levee, which was formed by sediment loading along the floodplain, suddenly breaks. This process is known as avulsion. This type of river channel position shift is common in river deltas (so-called delta switching) such as the Mississippi River delta or the Yellow River delta. Over a long period of time, a river delta can shift hundreds of kilometers due to repeated shifts in its main river channel and tributaries. Channel cutoffs in river deltas can also lead to the formation of so-called delta lobes through sedimentation which can form superlobes which in turn can cause further channel shifts. 

Large rivers are of vital importance for the ecology, the economy and the society. They provide for example freshwater for inhabitants living in the region and are an important economically, for example as transportation routes, for inland navigation, the supply of fresh water, food and nutrition through small-scale and large-scale crop cultivation, fisheries, water regulation and regulation of soil erosion.When a river channel shifts course, local industries, cities and agriculture along the old channel course are therefore severly affected. Economic activities along the old channel will usually decline, and in some cases seaports have to be moved. However, impacts might also be positive, for example a decrease in flood events and flood levels along the old channel course. This may lead to new economic opportunities such as changes in subsistence and cash crops that generate a higher income. However, they may also experience greater threats of flooding. The impacts might also consist of an increase in flood levels, riverbed aggradation, bank erosion, lateral channel shifts and stranded communities. These negative impacts might in turn lead to migration. 

Large rivers may be very important for regional and global trade. Therefore, a significant change in the position of the main channel usually has significant, mostly negative, impacts on human well-being.

Typically, river channel position regime shifts are irreversible. Without human action the river will almost never return to its original course. Managing river channel position regime shifts is very difficult because of the sudden and nonlinear nature of such shifts.

Options for enhancing resilience

In some situations attempts are made to retard or prevent a channel shift by controlling the riverflow (volume, current) through engineering works such as levees, spillways and weirs. This is usually done to protect housing, infrastructure and floodplain agriculture. However, efforts to modify the riverflow or implement defense infrastructure are usually extremely costly and requires enormous engineering efforts. Moreover, it is possible that the point where the shift is predicted to occur moves upstream or downstream, in which case the defense infrastructure becomes useless. The most iconic example of extensive engineering control structures to prevent a channel shift is the Mississippi River. The main channel is threatened by capture by the Atchafalaya River, and if this were to occur, New Orleans and Baton Rouge would suffer enormous negative economic consequences (see Mississippi River case study).

Options for reducing resilience to encourage restoration or transformation

It is possible to reduce the resilience of the system to encourage it to shift to a new channel position or return to a previous position, for example by removing levees. However, this is usually extremely costly and risky.