1. The extensive British canal system potentially provides a favourable habitat for aquatic plants and, because of its uniformity, it is ideal for studying relationships between disturbance intensity (largely from boat traffic) and communities of aquatic macrophytes.2. The standing crop and species composition of aquatic plants were measured in summer at 396 sites distributed randomly over the canal network. We also quantified the number of rare species and ‘attribute groups’ (groups of species sharing similar suites of biological traits). These data were analysed in relation to standing crop, assumed to indicate a disturbance gradient.3. Consistent with unimodal models and the Intermediate Disturbance Hypothesis, maximum hydrophyte species richness occurred at an intermediate biomass (50–200 g DWm−2). This corresponded to a low frequency of low magnitude disturbance (light boat traffic) on navigable sections, or occasional high magnitude disturbance (the post‐interventionist phase) on sections currently unnavigable and subject to active vegetation management. The frequency of rare species was also the greatest under these conditions, reflecting the availability of regeneration niches.4. Sorting of species into attribute groups revealed that the overall relationship between species richness and standing crop comprised of closely overlapping unimodal responses of nine attribute groups, superimposed on a core vegetation of Potamogeton pectinatus, together with greater representation of filamentous algae, lemnids and elodeids with increasing standing crop (i.e. assumed low disturbance). High species richness was associated with the overlap of functionally different groups of species, rather than with disturbance‐mediated coexistence of functionally similar plants.5. The analysis of a matrix of sites and the representation of plant traits weighted by the biomass of species that displayed them, in relation to different aspects of disturbance, highlighted a shift from traits associated with resilience (turions, unanchored floating or submerged leaves, low body flexibility, budding, small body size), or competitiveness (entire leaves, low reproductive output, high biomass density, large body size) at high standing crop, through to attributes more associated with resistance to disturbance (rhizomes, tubers, streamlining of foliage, low biomass density) at low standing crop.6. Comparison with a stochastic null model of change in species number along a constrained gradient, after correction for variation in sampling effort, indicated that sites towards the tails of the gradient (excluding those with extremely low biomass) supported more species than might be expected from chance alone, while the most species‐rich sites in mid‐gradient generally supported many fewer species than expected.7. We suggest that a disturbance regime that maintains intermediate standing crops would be appropriate for the conservation of species‐rich aquatic vegetation. Precise definition of this regime, under a range of circumstances, requires the study of temporal change at representative sites.