Date of Award
12-1-2018
Degree Name
Doctor of Philosophy
Department
Agricultural Sciences
First Advisor
Groninger, John
Second Advisor
Ruffner, Charles
Abstract
Forest disturbance occurs on a wide gradient of selectiveness and creates new growth opportunities for adapted species. Across the spectrum of disturbance, anthropogenic disturbance influences community assembly in the Midwest more than other mechanisms but is its role in shaping and maintaining ecosystems is inadequately considered in most discussions on the historic range of variability (HRV). Forest resiliency is threatened by unprecedented agents of ecosystem change such as invasive species and reduced regeneration potential of native species. Historic anthropogenic disturbance largely resulted in forest conditions which commonly contained high value attributes like heterogeneity across habitat types and landscape diversity, yet also produced forests of undesirable traits due to high grading for timber and overgrazing by domesticated stock. In order to maintain historical representative forests and improve the degraded forests, active forest management is necessary to continue historic disturbance patterns and combat new threats. Forest transition theory is used here to describe the impacts of human settlement and development activities on forest ecosystems across the Middle Mississippi River Valley. To date, researchers have identified the need for information related to changes of forest attributes such as species composition and stand structure, improved descriptions of short- and medium-term dynamics within the context of the long-term transition, and the integration of biophysical drivers of forest change through time. In Midwestern U.S.A., forest dynamics were influenced by frequent, low intensity disturbance events that mediate forest composition and stand structure by selecting for disturbance regimes that create oak woodland and interspersed prairies and meadows. The onset of Euro-American settlement was accompanied by detailed land-use records with information related to forest attributes, agricultural activities, and parcel ownership patterns. We aggregated multiple sources of historic forest conditions into a geodatabase in order to document changes over the past 200 years in Elsah Township, Illinois, where the pre-settlement (1820) forest, once dominated by oak and hickory species, has largely shifted to a maple dominated system with a declining oak-hickory component, heavily influenced by an invasive shrub species, bush honeysuckle. Using on ordinary kriging interpolation, forest density was estimated at 8.7 stems per acre on average with a mean basal area of 14.6 square feet per acre prior to settlement. Conservation practices of the early 1900s, including fire suppression and erosion control resulted in changes to forest structure with density increases to 127 trees per acre with a basal area of 175.8 square feet per acre. The high degree of topographic variability near the Mississippi River influenced forest cover changes as slopes with low angles were the first to be converted from forest cover to other land uses (circa 1850). Forest re-initiation occurred in areas with steeper slope due to a lack of human activities. Forest cover declined to the lowest point in 1927 and has been rebounding steadily throughout this century. Of the original 15,252 forested acres, 11.6% remained covered throughout the past 200 years and coincided with slopes with an average of 39.1 degrees. These data can provide a spatially explicit and historically accurate tool to guide land management decisions including restoration treatment, disturbance regime management, and land use preservation activities in similarly heterogeneous environments. Forest communities along the bluffs of the Mississippi River differ in species composition and stand structure associated with specific topographic positions of floodplain, transition talus slope, bluff top, and upland. In order to assess current stand characteristics and ecosystem trajectory, we measured all woody stems in 316 fixed radius plots (79 plots per topographic position) with a plot area of 25 m2. Alpha (defined as within system diversity) and Beta (defined as between system diversity) diversity and diameter distributions were determined for seedling, shrub layer, and overstory stems. Stem density increased from 21.4 stems ha-1 in 1820 to 613 stems ha-1 in 1936 followed by reduction to 314 stems ha-1 in 2017. Average stand diameter decreased from 40.9 cm in 1820 to 25.3 cm in 2017 (for upland stems greater than 7.5 cm) while basal area increased from 3.3 m2 ha-1 in 1820 to 40.4 m2 ha-1 in 2017. Alpha diversity was highest in the upland overstory and in the river island shrub layer. Beta diversity in the overstory was highest (0.67) between the bluff and the upland while lowest (0.08) between the bluff and the river island. Importantly, mesophytic species are no longer restricted to watercourses and valleys as reported in historical accounts and confirmed by the spatial analysis of original witness tree records. Currently, bush honeysuckle, an invasive species, dominates the shrub layer on most non-hydric sites of the talus slope, upland, and particularly across the bluff top where it is an indicator. Across all forest sites in the study, we found evidence of a community shift to less diversity and more mesophytic species over the past 80 years. Hill prairie vegetation on the limestone bluffs of the central Mississippi River Valley represents a significant portion of the remaining prairie, savanna, and woodland systems of the Midwest and should be appropriately managed with prescribed fire and woody stem reduction efforts. We examined the structure, composition, and temporal community patterns of the forest-prairie gradient by employing hierarchical cluster analysis and non-metric multi-dimensional scaling in combination with indicator species analysis and dendrochronological methods. Results suggest that four general community types exist across the forest-prairie gradient: Group 1 consists of the woodland community structure with significant indicator values for the density of Juniperus virginiana (indicator value 58.4, p = 0.0002), Carya glabra (45, 0.0022), Quercus stellata (23.7, 0.0424), and Lonicera maackii (74.2, 0.0002) and a high basal area (BA) of J. virgniana (21.4, 0.0276) and L. maackii (47.9, 0.0054). The first year of L. maackii presence was 1964 with the primary wave of invasion beginning around 1990. Group 2 contains bare soil coverage in the subplot (40.4, 0.0002) as the one indicator at a significant level. The species with the highest BA in Group 2 include Acer saccharum (9.08 m2 ha-1), Q. velutina (5.89 m2 ha-1), and Q. muehlenbergii (5.32 m2 ha-1). Group 3 typifies the hill prairie community with the sole indicator of grass coverage in the subplots (39.7, 0.0196). Group 4 represents the stage of forest development following the cessation of disturbance events and the trajectory advancing towards a mesophytic forest and contains 14 significant indicators.
Access
This dissertation is Open Access and may be downloaded by anyone.