The relationships among woody debris, composition of vegetation, topography, and sediment on ten gravel bars along a Montana river are examined in order to test a hypothesis of positive feedback leading to reduced rates of sediment transport. Plant species diversity is related to the area, sediment, and woody debris of bars. Several variables were significantly correlated and significant regressions of vegetative composition on sediment, topography, and debris resulted. The amount of woody debris, however, is not a statistically significant function of topography, sediment, or vegetation. Large amounts of woody debris are located both on low, barren, apparently new gravel bars, and on higher bars, where fine sediment is deeper and more mature stands of vegetation exist. Analyses of systems of linear equations representing hypothetical pathways of causality regarding vegetation, woody debris, topography, and sediment were not statistically significant. We suggest that time, a variable not amenable to measurement in this system, is an important direct control of processes at every step of the causal chain in such a way that its effects cannot be subsumed by any of the other variables. A disequilibrium condition in which the operations of mesoscale ecological processes are superimposed on sedimentological processes with macroscale components may also explain why the correlations are not maintained when ordered into a system.