Vegetation Changes Within the Flood Relief Stage of Two-Stage Channels Excavated Along a Small Rural Clay River

Abstract

(1) Vegetation changes on the steeply-sloping batters and near-horizontal berms, excavated 0.3 m above dry-weather water level to form the flood relief stage of a two-stage channel along the River Roding near Abridge, Essex, were monitored during 1980-82. (2) The mean number of species per 0.25 m2 was greatest during the first summer after berm excavation. Diversity (H'') was consistently greater on batters compared with berms, and in ungrazed compared with grazed sites. (3) Grass species dominated the vegetation in all areas. With the exception of Agrostis stolonifera L. and Alopecurus geniculatus L., invasion by native species was inhibited by cattle grazing. Most of the agriculturally-productive grasses in the seed-mix performed poorly in ungrazed areas. (4) By the third year there were significant floristic differences between ungrazed batters and berms. Arrhenatherum elatius (L.) Beauv. ex J. and C. Presl and Dactylis glomerata L. were more abundant on the freely drained batters (P < 0.01), while Agrostis stolonifera and Rumex spp. were more abundant on the berms (P < 0.001). Phalaris arundinacea L. was confined to berms, growing prolifically in poorly-drained areas. (5) In ungrazed areas, tall (.gtoreq. 0.76 m) plants, especially reeds, represented a flood hazard by the third summer because more than 25% of the flood channel was filled with vegetation. In contrast, reeds did not grow on ungrazed berms excavated 0.8-0.9 m above dry-weather water level. (6) Creeping grasses are recommended for seeding ungrazed high level berms excavated to one side of an otherwise undisturbed river channel. Two-stage channel construction has considerable potential as an ecologically acceptable method of flood alleviation along small clay rivers, but the amount of land required for berm excavation may limit the scope for this type of river management.