In Kentucky, Best Management Practices (BMPs) were developed to protect the integrity of water resources during and after silvicultural operations, but their effectiveness is largely unknown. The study objectives were to (1) quantify the extent to which current BMPs minimized logging-related changes in stream water quality, channel morphology, and the content of woody debris and (2) quantify the extent to which modified BMPs influenced these characteristics. Six watersheds in eastern Kentucky were commercially logged in 2008-2009. In each watershed, one of three treatments was implemented (each replicated once). Treatment 1 (BMP) was based on current Kentucky BMPs. In treatments 2 (BMP+) and 3 (BMP++), current BMPs were modified with additional protective measures. Before the harvest, sampling sites were established in each watershed. Channel morphology and woody debris data were collected at each site in 2006 and again 2014. Data for water chemistry were retrieved from ongoing sampling records. BMP watersheds were compared to one another and to unharvested control watersheds in each year independently. Data within each individual watershed were also compared across years. The results indicate that current BMPs were effective in minimizing changes in the watershed characteristics examined here, and the benefits of modified BMPs were highly variable.

As an efficient filter of airborne Hg compounds, forests retain a significant portion of the Hg deposited to the land. Forested watersheds have been identified as major sources of low-concentration Hg compounds to surrounding streams and lakes. Mercury export from forests is highly variable in both space and time. It is difficult to use field surveys alone to capture the spatiotemporal variation inherent in this variable. Mathematical models are required for improved representation. The objective of this Thesis is to develop and test a monthly dynamic model that can be used to estimate seasonal Hg export from forested watersheds to low-ordered forest streams. The fully developed model consists of four model components: (i) a forest hydrology component, to simulate variation in soil temperature, soil moisture, and stream discharge for input to the other model components; (ii) a forest nutrient cycling and biomass growth component, to simulate forest growth and litter production; (iii) a forest litter decomposition component, to simulate seasonal production of dissolved organic carbon (DOC); and (iv) a monthly DOC and Hg export component to simulate the translocation of DOC and Hg from forested watersheds to low-ordered streams. The Hg-export component incorporates an Hg-to-DOC binding coefficient estimated from a one-time stream survey of Hg and DOC concentrations. Simulations of in-stream Hg concentrations show two main trends: (i) an annual trend, associated with the seasonal (monthly) dynamics of forest litter production, decomposition, and DOC production and export, and (ii) a multiple-year trend, associated with forest harvesting and re-growth patterns of regenerating forests. This study demonstrates that (i) wetland- and conifer-dominated watersheds release a greater amount of Hg to aquatic ecosystems than upland- and deciduous species-dominated watersheds, and (ii) forests nearing maturity, export more Hg than young forests.

No description available.