Quantifying flooding regime in floodplain forests to guide river restoration

Abstract

Determining the flooding regime needed to support distinctive floodplain forests is essential for effective river conservation under the ubiquitous human alteration of river flows characteristic of the Anthropocene Era. At over 100 sites throughout the Connecticut River basin, the largest river system in New England, we characterized species composition, valley and channel morphology, and hydrologic regime to define conditions promoting distinct floodplain forest assemblages. Species assemblages were dominated by floodplain-associated trees on surfaces experiencing flood durations between 4.5 and 91 days/year, which were generally well below the stage of the two-year recurrence interval flood, a widely-used benchmark for floodplain restoration. These tree species rarely occurred on surfaces that flooded less than 1 day/year. By contrast abundance of most woody invasive species decreased with flooding. Such flood-prone surfaces were jointly determined by characteristics of the hydrograph (high discharges of long duration) and topography (low gradient and reduced valley constraint), resulting in increased availability of floodplain habitat with increasing watershed area and/or decreasing stream gradient. Downstream mainstem reaches provided the most floodplain habitat, largely associated with low-energy features such as back swamps and point bars, and were dominated by silver maple (Acer saccharinum). However, we were able to identify a number of suitable sites in the upper part of the basin and in large tributaries, often associated with in-channel islands and bars and frequently dominated by sycamore (Platanus occidentalis) and flood disturbance-dependent species. Our results imply that restoring flows by modifying dam operations to benefit floodplain forests on existing surfaces need not conflict with flood protection in some regional settings. These results underscore the need to understand how flow, geomorphology, and species traits interact to produce characteristic patterns of floodplain vegetation, and that these interactions should form the basis of effective river restoration and conservation.

Proceedings of the American elm restoration workshop 2016

“…Regardless of these regional variations, U. americana consistently appears to be most abundant in the better drained, older parts of forested point bars, riverine islands, and active floodplains where flooding is less frequent than in the Salix-and P. deltoides -dominated pioneer habitats but still frequent enough to prevent invasion by upland tree species (Marks et al 2014). That transition zone where U. americana is most abundant occurs where flooding happens about 1 percent of the time (i.e., 4 days/year, on average) (Marks et al 2014).…”

“…The F. pennsylvanica -U. americana dominance will continue until sediments accrete to the point that the flood-intolerant tree species that dominate the surrounding upland forests can colonize (e.g., Acer saccharum Marshall, Tilia americana L., Quercus rubra L., etc.) (Marks et al 2014). Regional variations on this successional sequence are related to differences in the species pool associated with climate and soil pH.…”

“…For example, on the drier western edge of its range, U. americana may be codominant with Acer negundo L. instead of A. saccharinum (Weaver et al 1925, Wiebe andShadick 2011). In riparian forests along medium and higher gradient streams and rivers, the pioneer tree species that U. americana eventually replaces can be Platanus occidentalis L. or on more acidic streams Betula nigra L. (Marks et al 2014, McClelland and Ungar 1970, Oosting 1942. In particular, in the southern part of the U. americana range there are many more floodplain tree species such as Celtis laevigata Willd.…”

“…That transition zone where U. americana is most abundant occurs where flooding happens about 1 percent of the time (i.e., 4 days/year, on average) (Marks et al 2014). …”

“…This fact may be underappreciated because even researchers have sometimes erroneously identified U. americana as U. rubra. Specifically, unlike U. americana, U. rubra appears to be generally restricted to higher floodplain terraces that do not flood every year (Curtis 1959, Marks et al 2014. Moreover, within the Connecticut River watershed, U. rubra was further restricted to those high terraces that had the highest soil pH (Marks, unpublished data).…”