A suite of 78 samples of riverbed gravel obtained at 39 sites along a 140 km reach of Peace River was used in a statistical study of the behaviour of mean grain size downstream. Whilst the exponential decline described by Sternberg was found in the overall data, significant additional 'structure' was also present. Although efforts were made to ensure that a homogeneous set of samples was obtained in the field, large site scale variability in the data prevented the complete elucidation of the additional pattern, which is probably due to injection of new material at tributary junctions. The paper illustrates sampling criteria that must be met for adequate description of grain size variations along a gravel river and identifies constraints that may limit its achievement.

You do not currently have access to this article.