Thesis Title

Laughing and Learning Through David Quammen's Natural Acts

Date of Graduation

Spring 1998

Degree

Master of Arts in English

Department

English

Committee Chair

Nancy Walker

Subject Categories

English Language and Literature

Abstract

The United States Forest Service (USFS) is charged with managing a large portion of publicly held lands in accordance with the Multiple Use and Sustained Yield Act of 1960 (MUSY, P.L. 86-517, 16U. S.C. 528). The Ava Ranger District of the Mark Twain National Forest in Southwest Missouri is one of the few USFS managed areas that allow Off-Road Vechicle (ORV) use on specially designated trails since the 1960s. Soil erosion is a focal area of concern for environmental groups in the area. Methods for predicting soil erosion include the Universal Soil Loss Equation (USLE), Modified Universal Soil Loss Equation (MUSLE) and Geographic Information System (GIS) models. These are either not developed for use in a forested environment or have limitations which limit their use in managing the soil loss on publicly held lands. In the period from 1991 to 1994, a monitoring program was initiated to determine which of the 125 miles of trails should be closed or rerouted. This program was a reaction to erosion but did not prevent or predict it. Regression analysis indicated no statistical correlation between slope, soil erodibility, or rider use rating. Inconsistencies in the USFS data did not allow the use of these data in a model to predict the amount of soil erosion. To determine if one-time trail transect measurements and a series of physical variables could be used to create a predictive model, 53 sites were selected and evaluated. Trail cross-section depths were measured at 6 inch intervals then averaged. This average depth of trail was the Y-variable in the linear regreession analysis. The physical variables of slope, soil K-factor, trail orientation, depth to bedrock, and hillslope position position were recorded at each site and were entered into the linear regression analysis as the X-variable. There was no significant statistical correlation with average depth of trail and each of the physical variables. Quartile boxplots were developed for each of the variables. No trends in these data could be discerned. Multiple regression analysis indicated a week statistical correlation. Efforts to predict the amount of soil erosion based on one-time transect measurements and the physical variables of slope, soil, K-factor, trail orientation, depth to bedrock, or hillslope position using regression models were unsuccessful in this study. Factors which may have influenced the lack of correlation include the amount of soil compaction, differences in vegetation, slope aspect, trail variability, inaccuracies of the soil survey, hillslope position determination, and clay content of the soil. Human behaviors may also have influenced the results. Type of riding, engine size, tire type, and distance from a trail head. In the absence of a predictive model, a monitoring program such as that implemented from 1991 to 1994, seems to be the best alternative. The designation as a fee use area may provide the funds necessary for agents to once again monitor the trails. Proper taining for any person volunteering for trail monitoring could avoid some of the problems experienced with the data from 1991-1994 monitoring programs.

Copyright

© Laura Lee Sapp

Citation-only

Dissertation/Thesis

Share

COinS