Causes of river avulsion: Insights from the late Holocene avulsion history of the Mississippi River, USA

The emphasis on gradient advantages in studies of avulsion is misleading. While gradient advantages are necessary for an avulsion to occur, the late Holocene avulsion history of the Mississippi River in Louisiana suggests that factors such as substrate composition and floodplain channel distributions are more important. Cross-valley to down-valley slope ratios of the modern floodplain range from 16 to 110 and are typically > 30. The slope ratio is 35 at the location of the Mississippi-Atchafalaya diversion (Old River) yet slope ratios are 83 to 110 immediately upvalley of Old River. All values of Mississippi River floodplain slope ratios are significantly larger than values of avulsion threshold calculated by numerical models. Shallow floodplain cores, C-14 dating of organic remains, and geologic mapping show that the Mississippi River has avulsed only four times over the past 5 ky in the southern Lower Mississippi Valley (LMV). Gradient advantages are widespread, yet avulsions are rare. These observations indicate that factors other than gradient advantage control Mississippi River avulsion. Several examples of Mississippi and Red River avulsion by channel reoccupation support the idea that channel distributions and substrate compositions are primary influences on avulsion. Incipient Mississippi River avulsion and development of the Atchafalaya River involved reoccupation of abandoned Mississippi River channels and a Red River crevasse-splay complex. The modern Atchafalaya River also incises buried Mississippi River channel-belt sands. Abandoned channel belts and crevasse-splay complexes consist of sandy substrates that facilitate scour and the development of channels capable of capturing the Mississippi River. Abandoned channels provide ready-made conduits for Mississippi River flow that can efficiently develop into avulsive channels. Multi-storied sheet sandstones in ancient fluvial deposits may provide additional support for the idea that erodible substrates and flood-plain channel distributions are critical influences on avulsion. These features record episodic reoccupation of channel belts, which at least in some cases, may simply reflect successive avulsions rather than major changes in aggradation rate or extrabasinal factors such as climate.