Lower water level to restore degraded freshwater marshes
Overall effectiveness category Unknown effectiveness (limited evidence)
Number of studies: 2
Background information and definitions
This action involves one-off action to lower the water level in degraded marshes, to a depth that should support emergent vegetation. This means that intervention should (a) occur at one point in time, after which the water level is not actively managed, and (b) must affect a marsh that is wetter than normal, but is still recognisable as, or retains substantial characteristics of, the target habitat. Specific techniques to reduce water levels include removing dams downstream, switching off pumps that add water to a focal site, and improving drainage by digging shallow “runnels” or deeper creeks (Wigand et al. 2017).
Caution: This action may have negative effects on habitats elsewhere in the catchment. For example, removing dams could flood marshes, swamps or upland habitats downstream. There may also be conflicts with water needs of human populations that need to be managed.
Related actions: Lower water level to restore/create marshes from other land uses; Backfill canals or trenches; Actively manage water level; Manage water level to control problematic plants; Reprofile/relandscape, may involve raising the ground surface towards or above the water table; Lower water level to complement planting.
Wigand C., Ardito T., Chaffee C., Ferguson W., Paton S., Raposa K., Vandemoer C. & Watson E. (2017) A climate change adaptation strategy for management of coastal marsh systems. Estuaries and Coasts, 40, 682–693.
Supporting evidence from individual studies
A before-and-after study in 1993–1996 of a lakeshore marsh in Ohio, USA (Wilcox & Whillans 1999) reported that over a year of drawdown, the area of emergent vegetation increased. In 1993, two years before drawdown, the marsh was mostly open water with only 10% covered by emergent vegetation stands. In 1996, after approximately one year of drawdown but before any other interventions were carried out, 73% of the marsh was covered by emergent vegetation stands. Colonizing vegetation included several herbaceous wetland species, wind-dispersed woody species, and some upland herbs (not quantified). Methods: In 1995, an embankment was constructed across the mouth of Metzger Marsh to replace a natural barrier beach that had disappeared in the 1970s. The embankment separated the marsh from Lake Erie and caused a decline (drawdown) of the water level in the marsh. Vegetation was surveyed before (1993) and after (1996) drawdown (further details not reported).Study and other actions tested
A replicated, controlled, before-and-after study in 2007–2010 of six freshwater marshes in Manitoba, Canada (Baschuk et al. 2012) found that a partial but long-term drawdown of the water level affected the area covered by some marsh vegetation classes. In marshes where the water level was drawn down, there was significant variation over time in the coverage of five vegetation classes. There was an initial decline followed by recovery in the area of vegetation dominated by sedges Carex spp. (before: 141 ha; after one year: 103 ha; after three years: 127 ha), cattails Typha spp. (111 ha; 69 ha; 98 ha) and horsetails Equisetum spp. (9 ha; 2 ha; 20 ha). Coverage of dead vegetation showed the opposite pattern (15 ha; 80 ha; 13 ha). Coverage of vegetation dominated by bur-reeds Sparganium spp. increased steadily over time (<1 ha; 2 ha; 7 ha). There was no significant change over time in the coverage of other vegetation classes, including trees (see original paper for data). In marshes that were not drawn down, the area of all vegetation classes was stable over time (see original paper for data). Methods: This study focused on six 84–207 ha marshes within one wetland complex. From August 2007, the water level was lowered in three marshes (average depth: 30 cm; maximum depth: 60 cm) using water control structures. This level was maintained over the entire study period. In the other three marshes, the water level remained unnaturally high (average: 67 cm; maximum: 100 cm). The coverage of vegetation types was measured from satellite images, taken in the summer before (2007) and for three years after (2008–2010) drawdown.Study and other actions tested