A Spatial Characterization of Vegetative Composition, Landscape Structure, and Hydrology of the Yellow River Floodplain, Central Wisconsin

Abstract

The Yellow River floodplain forest, like many floodplain ecosystems, is a unique and diverse landscape both in structure and in composition. Recent research has suggested that although ecological processes operating at local spatial scales are important, they may not be adequate to understand local species abundance and diversity, habitat fragmentation, and other biotic interactions. The objectives for this study were to determine (a) what types of communities exist, (b) how the assemblages are structured, (c) the structure and composition of the plant communities, and (d) to determine the pre- and post- dam hydrologic regime for a section of the Yellow River in central Wisconsin. In 2000, vegetative data were collected at four layers; overstory, intermediate, saplings, and seedlings. Data were collected at pre-determined plot locations along established transects that extended perpendicular from the river margin to the observed edge of the floodplain. Silver maple (Acer saccharinum), northern red oak (Quercus rubra), swamp white oak (Quercus bicolor), red maple (Acer rubrum), and green ash (Fraxinus pennsylvanica) were found to comprise 64% of the co-dominant overstory that contained 18 species. The intermediate layer was comprised of 15 species with one species, black cherry (Prunus serotina), that was rare and did not occur within the overstory. At 30% of the sample plots, no intermediate layer was present. The seedling and sapling layers were comprised of 15 and 7 species, respectively. Regeneration was largely inadequate for most species and sparse for the entire landscape. It appears that the sparse regeneration was due to the high level of the canopy cover. Ninety-six species were found in the herbaceous layer. Thirty of these species occurred in ≥ 5% of the sample plot locations with the most frequent being violet spp. (Viola spp.), bristly greenbrier (Smilax pseudochina), stinging nettle (Urtica dioica), and touch-me-nots (Impatiens spp.) –all light generalists. To elucidated species patterns within the floodplain, two principal component analyses were performed using SAS statistical software. The first principal component analyses of the overstory and intermediate layers indicated that 69% of the variance was accounted for by the first eight components. The second principal component analysis of the sapling and seedling layers indicated that 58% of the variation could be accounted for by the first six principal components. The patterns detected indicate that species with similar flood tolerances appeared to occur within the same plot locations. Despite this possible interpretation of the species patterns, several environmental factors not included in this study are most likely contributing to the structure and therefore it is difficult to determine the driving force behind those patterns. FRAGSTATS*ARC analyses on two coverages, patch types and structural components, indicate that there are three community types, silver maple, oak, and mixed deciduous, that occupy 58% of the floodplain. These communities are mostly dense, mature forests. In 1965, the Dexterville Dam was constructed on the Yellow River located in central Wisconsin. Using the statistical package, Indicators of Hydrologic Alteration (IHA), analyses were conducted to determine if water flow had decreased or increased in frequency and duration from the time the dam was erected. Daily streamflow data, reported in cubic feet / second (cfs), collected from 1944–2000, were acquired from the Babcock water station (number 05402000), located five miles downstream from the dam. A pre-impact and a post-impact period were established and defined as 1944-1965 and 1966-2000, respectively. Results indicate flood events, as defined by the high pulse, occurred at approximately the same time of the year with April being the peak flood month. The median high pulse was 385.1 cfs (10.9 cms) pre-impact and 524.0 cfs (14.8 cms) post-impact, an increase of 138.9 cfs (3.93 cms). Maximum water flow within a 3-day period increased 60.0 cfs (1.7 cms). The frequency of low and high pulses indicated that approximately the same number of days are affected each time period, but the duration and frequency has changed, especially for the low pulses. These events had a duration of 14.3 days pre-dam and 5.0 days post-dam and a frequency that changed from five to two times per year. Additionally, the rise and fall rate increased as well as the number of times per year that the water flow changed from rising to falling or visa versa, which was 17% greater post-impact. Results from this study indicate that water flows have increased since the construction of the dam. This increase in flow could have negative implications for the regeneration and succession of floodplain tree species and on animal species that utilize the floodplain for feeding or nesting grounds. Precipitation records from 1903-2001 were analyzed using a Wilcoxon-Mann- Whitney non-parametric test. The result indicated that there has been a significant increase in the precipitation amount (p=0.0291) for the Yellow River study area. Although precipitation increased, flow characteristics such as the rise and fall rates, reversals, and durations are more likely to have been altered by dam fragmentation. These measurements can have a major influence on the vegetation, sedimentation, and geomorphology in a system. This suggests that the hydrologic changes observed are most likely influenced by both climatic differences as well as the dam modification to the system. Although extensive data on vegetation have been collected within the Yellow River study area, the results indicate that dams have landscape-level effects that are difficult to predict and interpret due to the variations in the hydrologic regime. To better understand the river and its adjoining floodplain, more research is needed that further addresses other features such as topography, ground cover (litter, soil and woody debris), elevation (changes observed throughout the floodplain in association with levees and oxbows), landscape metrics on nearest neighbors (similar patches adjacent to each other), sediment load, and channel morphology in the Yellow River or similar floodplain ecosystems.