Abstract
Drainage rearrangement, involving stream piracy (capture), drainage diversion and/or beheading, may be significant for sediment budgets (including sediment provenance) and biotic distributions, as well as for its more usually considered role in landscape evolution. The processes involved in drainage rearrangement are not as self-evident as its abundant literature indicates. This is especially the case with the commonly invoked stream capture. The key process in stream capture, namely, drainage head retreat, is difficult to envisage as a normal part of drainage net evolution, especially in the light of recent findings on drainage hollow evolution. Stream capture may therefore be a relatively rare event in drainage net evolution. This, and uncertainties with interpretations of supposed elbows of capture, mean that stream capture should not be routinely invoked in interpretations of long-term drainage evolution. Further uncertainties associated with the maintenance of drainage lines during the erosion of significant crustal sections, especially in faulted and folded terrains, diminish the likelihood of many supposed examples of stream capture. It is more likely that examples of drainage rearrangement attributed to stream capture were generated by drainage diversion, but even this may involve special conditions.