Homogenous rivers, homogenous faunas

  1. Peter B. Moyle* and
  2. Jeffrey F. Mount
  1. Center for Watershed Sciences, University of California, 1 Shields Avenue, Davis, CA 95616

Habitat change and invasions of alien species are usually listed as the two biggest causes of biodiversity loss. When the two coincide, the results can be devastating, because altered habitat often favors alien species adapted to the changed conditions. This is particularly true in streams and rivers of North America, where homogenization of fish faunas through extinction of native species and invasions of alien species have been widely recorded (1). Although dams that alter flow regimes have been implicated as a causal factor behind this shift in North America, the degree to which flow regimes have been altered is not widely appreciated. The study by Poff et al. (2) in this issue of PNAS should help to change this problem. They document that flow alteration by dams in the U.S. is pervasive and that the resulting flow regimes tend to converge on one another, regardless of the original flow regimes of the dammed rivers.

Dams homogenize flows mainly by reducing peak flows and increasing minimum flows, creating more uniform flow conditions. They also change the timing and duration of high and low flows. The exact changes to the flow regime vary from region to region but result in rivers with similar flow regimes and reduced variability in flows. From an aquatic organism's perspective, reduced variability in flows is particularly important, because this results in reduced channel complexity, often with stable channels that have reduced capacity to carry high flows and that are incised at low flows. Less complex channels translate into a lowered diversity of structural features preferred by fish and other aquatic organisms (3). This loss of structure, in turn, results in fewer fish of fewer species. Ironically, these relationships do not change with large hydropower dams that have frequent peaking flows, even though they increase flow variability, because …

*To whom correspondence should be addressed. E-mail: pbmoyle{at}ucdavis.edu

« Previous | Next Article »Table of Contents