Tributary connectivity, confluence aggradation and network biodiversity

In fluvial networks, some confluences are associated with tributary-driven aggradation where coarse sediment is stored, downstream sediment connectivity is interrupted and substantial hydraulic and morphological heterogeneity is generated. To the extent that biological diversity is supported by physical diversity, it has been proposed that the distribution and frequency of tributary-driven aggradation is important for the magnitude and spatial structure of river biodiversity. Relevant ideas are formulated within the Link Discontinuity Concept and the Network Dynamics Hypothesis, but many of the issues raised by these conceptual models have not been systematically evaluated. This paper first tests an automated method for predicting the likelihood of tributary-driven aggradation in three large drainage networks in the Rocky Mountain foothills, Canada. The method correctly identified approximately 75% of significant tributary confluences and 97% of insignificant confluences. The method is then used to evaluate two hypotheses of the Network Dynamics Hypothesis: that linear-shaped basins are more likely to show a longitudinal, downstream decline in tributary-driven aggradation; and that larger and more compact basins contain more confluences with a high probability of impact. The use of a predictive model that included a measure of tributary basin sediment delivery, rather than symmetry ratio alone, mediated the outcomes somewhat, but as anticipated, the number of significant confluences increased with basin size and basin shape was a strong control of the number and distribution of significant confluences. Doubling basin area led to a 1.9-fold increase in the number of significant confluences and compact basins contained approximately twice as many significant confluences per unit channel length as linear basins. In compact basins, significant confluences were more widely distributed, whereas in linear basins they were concentrated in proximal reaches. Interesting outstanding issues include the possibility of using spatially-distributed sediment routing models to predict tributary-driven confluence aggradation and the need to gather ecological data sufficient to properly test for increases in local and network-scale biodiversity associated with significant confluences and their network-scale controls.