Master of Science Thesis submitted 1997

Effects of Microtopography on the Natural Regeneration of Wetlands Restored on Agricultural Fields.

Dwight McKinney


The vegetation of two swamp forests along Durham Creek in Beaufort County, North Carolina is described in relation to elevational, hydrologic, and edaphic gradients. Regression relationships between study sites and a downstream gaging station were used to hindcast 25 years of surface flooding frequencies. Elevations of individual plant microsites were used in conjunction with stage probabilities to examine individual species responses to varying surface flooding regimes. Although the sites have similar surface flooding regimes, particularly when considered on an annual basis, differences in evapotranspiration, base flow, and the average depth of the water table during the growing season created considerably different conditions in the two sites. Depth to mottling, flooding frequency, elevation, and several soil chemical properties were all significantly correlated. The correlation between decreased water table fluctuations and higher availability of Ca and Mg in one site may be the result of reduced leaching losses. Direct elevational gradient analysis combined with plot ordinations derived from Detrended Correspondence Analysis and Canonical Correspondence Analysis suggested that structural differences in vegetation between the sites are primarily the result of variations in growing season flooding frequency, subsurface drainage, % base saturation, exchangeable acidity, and disturbance history.

The results of this study suggest that plant species response to flooding is affected by the combined action of many forces including hydrology and soil fertility and that species may exhibit varying responses to flooding depending on soil nutrient availability. Vast differences in flooding regime occurring over a few centimeters in some bottomlands make precise prediction, planning, or design of species-site interactions formidable at best. The common practice of linearly extrapolating surface flooding frequencies from an established gage to some point of interest may not reflect subtle yet critical differences between sites in evapotranspiration, water table depth, base flow, and microtopographic influences on drainage.