The responses of invasive exotic plant species (IES) to silvicultural treatments one growing season after timber harvesting were examined in the Cumberland Plateau region of Southeastern Kentucky. Treatments included a commercial deferment harvest and unharvested control applied to five watersheds within University of Kentucky’s Robinson Forest. The effects of harvesting were compared between treatments and between preharvest and postharvest samplings. The spatial distribution and abundance of targeted IES throughout forest sites and trail systems were calculated from several sampling schemes. Additional analyses were performed to quantify forest disturbances derived from harvest activities to determine the relationships between soil, light levels, and other environmental characteristics and IES cover. Logistic and multivariate analysis techniques were used to analyze differences in IES distribution between pre-harvest and post-harvest units to relate post-harvest IES to microsite conditions. Microsite conditions within the forest and along the trail system proved important for explaining the presence and distribution of IES. Timber harvesting caused a significant increase in both Ailanthus altissima and Microstegium vimineum within harvested areas. However, many other identified IES did not initially respond to disturbances. Throughout the treatment units, species were influenced by disturbance type and intensity, as well as proximity to reclaimed surface mined land.

Sediment is a prevalent non-point source pollutant associated with forest operations. Roads and skid trail surfaces have erosion rates that are greater than the harvest area. Forestry best management practices (BMPs) have been developed to minimize erosion on skid trails, but few projects have compared the effectiveness of different BMPs for bladed skid trails in the mountains. This project evaluated soil erosion rates from bladed skid trails in the Ridge and Valley physiographic region of Virginia following an operational timber harvest. Skid trails were assigned into six blocks where each block had similar slopes and soils. All BMP treatments had waterbars, which are considered the minimum acceptable BMP closure treatment. Each block contained four different skid trail closure BMP treatments (waterbar only (Control), slash-covered (Slash), seeded (Seed), and seeded with fertilizer and mulch (Mulch)). The 24 treatment units were isolated with waterbars and installed following the Virginia Department of Forestry (VDOF) BMP guidelines. The randomized complete block design had three slope class ranges: gentle (0%-10%), moderate (11%-20%), and steep (21%-30%). Stormwater runoff from skid trails was directed at downslope waterbars and eroded material was trapped in silt fences at each treatment area. Depth and area of eroded soil collected in silt fences was measured monthly to quantify total erosion volume for the skid trail area and converted to a per acre basis. Volumes were converted to mass using soil bulk density within the trapped sediment. Control treatments had an average erosion rate of 6.8 tons ac-1 yr-1, with rates up to 73.5 tons ac-1 yr-1 following installation and during extreme rainfall events. Seed treatments recorded an average erosion rate of 2.6 tons ac-1 yr-1, with rates reaching 27.2 tons ac-1 yr-1. Adding grass seed provided ground cover, but not consistently over time. Due to high rates of ground cover, the Mulch treatments averaged 0.5 tons ac-1 yr-1 with an extreme of 3.8 tons ac-1 yr-1. Slash treatments were found to reduce erosion rates to an average of 0.4 tons ac-1 yr-1, with the highest rate being 1.8 tons ac-1 yr-1. Site characteristics on experimental units were collected quarterly in order to model erosion rates with commonly used erosion models for forestland (USLE-Forest, RUSLE2, WEPP:Road). Direct erosion estimates were compared to erosion model predictions produced by USLE-Forest, RUSLE2, and WEPP:Road in order to partially confirm the relationship between sediment trap data and the models. Using multiple analyses it was determined that USLE-Forest and RUSLE2 predicted mean values that are more similar to the actual measured rates, RUSLE2 and WEPP:Road have better linear relationships to the measured rates than does USLE-Forest, and USLE-Forest was the most statistically similar to the measured data using a nonparametric Steel-Dwass Multiple Comparisons Test. All models performed inadequately when attempting to predict Control or Slash treatments; while all models performed the best at predicting Mulch treatments. / Master of Science

Sediment is one of the leading non-point source pollutants in the U.S and has detrimental effects on biological communities such as aquatic communities; human use such as recreation; and natural processes such as flood water storage. For silvicultural operations, the majority of sediment is produced from erosion on highly disturbed areas, such as skid trails, haul roads, and log landings. Erosion from silvicultural activities not only has the potential to introduce sediment into waterways but can also decrease site productivity through the removal of topsoil. In order to minimize erosion from silvicultural operations, forestry Best Management Practices (BMPs) have been developed, but efficacies of various BMP options are not well documented. This study evaluated five closure and cover BMPs for the control of erosion on bladed skid trails through both field based measurements with sediment traps and soil erosion modeling. The erosion models used were the Universal Soil Loss Equation for Forestry (USLE – Forest), the Revised Universal Soil Loss Equation version 2 (RUSLE2), and the Water Erosion Prediction Project for Forest Roads (WEPP – Forest Roads). Erosion model predictions were also regressed against field based results to determine accuracy. The bladed skid trail BMP treatments evaluated were: 1) water bar only (Control); 2) water bar and grass seed (Seed); 3) water bar, grass seed, and straw mulch (Mulch); 4) water bar and piled hardwood slash (Hardwood Slash); and 5) water bar and piled pine slash (Pine Slash). Field based results show that the Control treatment was the most erosive (137.7 tonnes/ha/yr), followed by the Seed treatment (31.5 tonnes/ha/yr), Hardwood Slash treatment (8.9 tonnes/ha/yr), Pine Slash treatment (5.9 tonnes/ha/yr), and finally the Mulch treatment was the most effective erosion control technique (3.0 tonnes/ha/yr). Model accuracy results show that RUSLE2 performed the best overall. Both USLE – Forest and WEPP – Forest Roads under predicted values on the Control treatment, where erosion rates were very high. WEPP – Forest Roads under predicted these values the most. All models generally show that the Control was the most erosive followed by the Seed, Hardwood Slash, Pine Slash, and Mulch treatments. / Master of Science

Forestry best management practices (BMPs) programs were developed by individual states in response to the Clean Water Act in order to protect water quality during and after timber harvests. Our research goals are to compare BMP implementation at stream crossings by region and road type in Virginia and to quantify effectiveness of BMPs by developing hypothetical upgrades and determining upgrade costs. Stream crossings (75 truck, 79 skidder) sampled for BMP implementation were on operational harvests conducted in 2016, from the Mountains, Piedmont, and Coastal Plain of Virginia. Erosion rates of stream crossing approaches were modeled using the Universal Soil Loss Equation modified for forest lands (USLE-Forest) and Water Erosion Prediction Project (WEPP) methodologies. Implementation ratings (BMP-, BMP-standard, BMP+) were developed to characterize crossings with respect to state implementation standards. Costs for upgrading crossings to a higher BMP category were estimated by adjusting cover percentages and approach lengths. Sixty-three percent of stream crossings were classified as BMP-standard, with an average erosion rate of 7.6 Mg/ha/yr; 25% of crossings were classified as BMP+, with an average erosion rate of 1.7 Mg/ha/yr; and 12% of crossings were classified as BMP-, with an average erosion rate of 26.2 Mg/ha/yr. Potential erosion rates decreased with increasing BMP implementation (p <0.0001). Average BMP implementation audit scores for stream crossings were 88% on skid trails and 82% on truck roads. To upgrade from a BMP- to BMP-standard, the cost-benefit ratio of dollars to tons of sediment prevented averaged $166.62/Mg for skid trails and $2274.22/Mg for truck roads. Enhancement to the BMP+ level is not economically efficient and BMP implementation at stream crossings reaches maximum efficiency at the BMP-standard level. / M. S. / Timber harvesting can accelerate erosion processes and result in the discharge of large quantities of sediment into nearby water resources if proper management is not used during and after harvests. Most of sediment entering streams is generated from forest roads and trails, particularly at stream crossings. This study includes 154 crossings (75 truck, 78 skidder) randomly selected from the Mountains, Piedmont, and Coastal Plain regions of Virginia. Evaluations were conducted on lands with silvicultural operations completed in 2016 that had truck road crossings and/or skidder crossings over any stream. Information was gathered at each crossing to model the erosion rates from both approaches to the stream by using the Universal Soil Loss Equation modified for forest lands and the Water Erosion Prediction Project. Implementation ratings (BMP-, BMP-standard, BMP+) were developed to characterize crossings with respect to state implementation standards, and evaluate other factors, including stream bank stability, ground cover, and evidence of sedimentation in the stream. The Virginia Department of Forestry BMP audit was used to score the crossings. Costs for upgrading the crossing to a higher BMP category were estimated by adjusting cover and approach lengths, and then using previous research data and existing road cost models. Potential erosion rates decreased with increasing BMP implementation (p <0.0001). Average BMP implementation audit scores for stream crossings were 88% on skid trails and 82% on truck roads. This research contributes to the evidence of BMP effectiveness and provides transparency to the citizens of Virginia regarding sustainable forestry practices.

Short term studies have suggested that ground based timber harvesting on wet sites can alter soil properties and inhibit early survival and growth of seedlings. Persistence of such negative effects may translate to losses in forest productivity over a rotation. During the fall and winter of 1989, numerous salvage logging operations were conducted during high soil moisture conditions on wet pine flats in the lower coastal plain of South Carolina following Hurricane Hugo. A long-term experiment (split-plot within an unbalanced randomized complete block design) allowed assessment of long term effects of rutted and compacted primary skid trails and subsequent site preparation on soil properties and loblolly pine (Pinus taeda L.) productivity. The experiment had 12 blocks, four levels of site preparation as the whole plot factor (bedding, disking with bedding, disking, and no site preparation), and two levels of traffic as the subplot factor (primary skid trail, no obvious traffic). After 23 years, bedding and disking with bedding treatments effectively enhanced soil physical properties and stand productivity via promoting greater survival and stocking, but had little effect on the size of individual trees relative to disking and no site preparation treatments. Primary skid trails significantly reduced the size of individual trees, but had no appreciable long term effects on soil properties or stand productivity after 23 years. The study suggests that bedding is the most efficient practice to enhance soil properties, seedling survival, and stand productivity on wet sites. However, site preparation is not necessary for these soils and sites, if strictly intended to restore soil properties and stand productivity in primary skid trails. Reduction in individual tree sizes on primary skid trails emphasizes benefits in minimizing the spatial extent of disturbance. / Master of Science / Heavy equipment traffic associated with ground-based timber harvesting has potential to alter soil properties resulting in lower productivity of the subsequent forest. Various soil tillage techniques have been suggested to offset changes in soil properties and forest productivity on disturbed soils, and to enhance soil properties and forest productivity on undisturbed soils. An experiment was conducted on low-lying Atlantic Coastal Plain sites to compare the effects of four soil tillage treatments (bedding, disking, disking with bedding, no tillage) on soil properties and forest productivity 23 years after treatments were installed. Bedding displaces soil from furrows into continuous, linear mounds called beds, such that bed surfaces are above the original soil surface. Disking is a tillage practice commonly implemented in agriculture. For the disking with bedding treatment, plots were disked followed by bedding. The no tillage treatment served as a control. Additionally, the experiment provided comparison of soil properties and forest productivity between soils heavily disturbed by logging activities and relatively undisturbed soils. Bedding and disking with bedding created favorable, localized soil conditions that promoted greater loblolly pine survival which translated to approximately double the total stand volume per unit area relative to disking and no tillage treatments. Differences in the sizes of individual trees among tillage treatments were minimal. Heavily disturbed soils and soils undisturbed by logging activity were similar in terms of soil properties and loblolly pine volume per unit area; however, individual trees were smaller on heavily disturbed soils. The implications of this study provide practical guidance for forest management decisions. Of treatments compared, bedding is the most efficient to increase total stand productivity on poorly drained sites. Disking with bedding offers no additional benefits, but is more expensive to implement. Results also imply that soils disturbed by logging have potential to recover over time such that long term forest productivity is sustained. However, the reduction in individual tree sizes emphasizes benefits of minimizing soil disturbance during timber harvests.