Physical manipulations of rivers are widely promoted as restoration solutions to the degradation of a variety of ecological properties. Because physical and hydrologic properties of rivers are closely linked, and hydrologic regime is known to be a major driver of ecological function in rivers, this approach may be appropriate. However, the response of rivers to hydrologic manipulations such as dam removal and controlled flooding has largely gone untested. In these studies I measured ecological responses of three different river restoration projects and reviewed recent literature on physical restorations in streams.

Vegetation recovery on two former impoundments that were dewatered following dam removal was rapid. Planting with native vegetation initially contributed to species richness on one site, but not in the second site. Few of the planted species were present on either site four years post-removal and the presence of invasive species increased over this interval. Seed bank analysis suggested that the dominant invasive species Phalaris arundinacea was not present in the seed bank immediately following draw down.

In the small headwater stream, Boulder Creek near Baraboo Wisconsin, the removal of two small dams initiated channel formation and sediment transport similar to that observed in other dam removals in Wisconsin. Phosphorus uptake, expected to be reduced by the dam removal, was not significantly different post-removal or in corresponding months the following year. This was due to the relatively small physical impact of dam removal and the naturally high variability of phosphorus concentrations in the stream.

A water control structure installed in a levee on the Baraboo River allowed the reintroduction of flooding to a previously drained and farmed floodplain. Denitrification and potential denitrification as well as soil moisture, nitrogen, and organic matter content were measured one year before and for two years after the manipulation. While soil moisture increased and soil nutrient concentrations were consistent throughout the study period, there was no evidence that denitrification rates increased. In some areas they may have decreased. Spatial and temporal variability of denitrification were both high making it difficult to extrapolate measurements to the whole floodplain.

A survey of published literature on the biological responses of rivers to physical restoration showed that these types of restorations can achieve restoration goals in some situations. It is unclear from most of these studies if rehabilitated conditions are persistent. While conceptually it is reasonable to hypothesize that intentional physical or hydrologic alterations to rivers may rehabilitate biological parameters, there is not yet good evidence that current river restoration techniques can consistently produce this response.