Physical and vegetative characteristics of a relocated stream reach, constructed wetland, and riparian buffer, Upper Saucon Township, Lehigh County, Pennsylvania, 2000-04

Abstract

The U.S. Geological Survey, in cooperation with the Pennsylvania Department of Transportation, Engineering District 5-0, investigated physical and vegetative changes within a relocated stream reach, constructed wetland, and riparian buffer from September 2000 to October 2004. This report presents an evaluation of data collected using methods from multiple sources that have been adapted into a consistent approach. This approach is intended to satisfy a need for consistent collection of different types of data with the goal of transferring technology and findings to similar projects. Survey data indicate that adjustment of the upstream part of the relocated stream reach slowed over the monitoring period, but the downstream channel remains unstable as evidenced by excessive deposition. Upstream migration of a nick point has slowed or stopped altogether as of the 2003 assessment when this feature came in contact with the upstream-most part of the channel that is lined with riprap. Documented streambed erosion in the upstream cross sections, along with deposition downstream, has resulted in an overall decrease in slope of the stream channel over the monitoring period. Most streambed erosion took place prior to the 2002 assessment when annual mean streamflows were less than those in the final 2 years of monitoring. An abundance of fine sediment dominates the substrate of the relocated channel. Annual fluctuations of large particles within each cross section demonstrates the capacity of the relocated channel to transport the entire range of sediment. The substrate within the 0.28-acre constructed wetland (a mixture of soil from an off-site naturally occurring wetland and woodchips) supported a hydrophytic-vegetation community throughout the investigation. Eleocharis obtusa (spike rush), an obligate-wetland herb, was the most prevalent species, having a maximum areal cover of 90 percent in fall 2001 and a minimum of 23 percent in fall 2004. Drought-like conditions in water year 2002 (cumulative precipitation was 28.11 inches) allowed species like Panicum dichotomiflorum (witch grass), Salix sp. (willow), Leersia oryzoides (rice cutgrass), and Echinocloa crusgalli (barnyard grass) to become established by fall 2002. Above-average precipitation in water years 2003 and 2004 (58.55 and 53.17 inches, respectively) coincided with increased areal cover by E. obtusa in fall 2003 (56 percent) and decreased areal cover in fall 2004 (23 percent). Pond-like conditions that probably persisted throughout the 2004 growing season favored aquatic species like Alisma subcordatum (water plantain) to the detriment of many emergent species, including E. obtusa. Despite the pond-like conditions, L. oryzoides, an obligate-wetland grass, increased in areal cover (from 12 to 34 percent) between the 2003 and 2004 growing seasons because it was established in the higher elevations and the peripheral areas of the constructed wetland that were less prone to persistent inundation. Canopy development by trees and shrubs in the riparian buffer was initially (fall 2000) poor (39.7 percent), resulting in more available sunlight for the herbaceous understory than in any other growing season. As a result, areal cover of herbaceous species and trees and shrubs less than 1-meter tall was 108 percent in fall 2000 with Lolium perenne (perennial rye), Polygonum persicaria (lady's thumb), and Setaria faberi (foxtail) collectively contributing nearly half the cover (59.2 percent). Because of increases in canopy cover by trees and shrubs (39.7 percent in fall 2000 to 127 percent in fall 2004), herbaceous cover decreased to 76 percent by the fall of 2001 and varied between 72 and 77 percent for the rest of the study period. Tree density in the riparian buffer ranged from 3,078 and 4,130 plants per acre (fall 2000 and 2003, respectively) over the study period but essentially remained constant after fall 2001; computations reported each fall between fall 2001 and fall 2004 are within 10 percent of one another. When the study ended in fall 2004, Acer negundo (box elder) and Fraxinus pennsylvanica (green ash) were the most populous tree species (1,526 and 1,084 plants per acre, respectively) followed by Quercus bicolor (swamp white oak; 720 plants per acre). A. negundo , F. pennsylvanica , and Q. bicolor also contributed the greatest areal cover in fall 2004 (31.2, 24.0, and 18.5 percent,respectively).