Modeling Forest Productivity Using Soil and Other Geographical Variables
Abstract
Site productivity measures the primary productivity potential of forest ecosystems. It is characterized by an interaction of biotic and abiotic factors such as climate, soil, and topography. An accurate site productivity characterization allows for efficient land use allocation, integrated ecosystem planning, and prescribed ecosystem management. Site index, or the height of dominant or co-dominant trees at a reference age, is an important proxy of site productivity and has traditionally been used in many conceptual and simulation models of ecosystem dynamics; however, it assumed that forest management history has no effects on site productivity. Indeed, forest site productivity is dependent on both site and management related factors. Better management of under-utilized woody biomass from forests such as treetops, branches, twigs, bark, and limbs could be potential sources of feedstock to meet the growing demand for biofuels. It is important to find a balance between forest management practices and maintaining forest productivity to protect the ecosystem services we receive from these forests. This study developed predictive models using a geocentric approach by pairing Forest Inventory and Analysis (FIA) plot data with, climate, topography, and soil data across the state of Tennessee. We found that variables derived from digital elevation models were the most influential in site productivity predictions, with elevation being the most important as it was significant in 12 of the 28 models. Average precipitation and effective cation exchange capacity was the next most influential variable in the stepwise regression models, as well as clay percentage. Several site productivity prediction models were developed, a parametric multi-species model captured 48% of the variation with an RMSE of 11.55 feet, as compared to a non-parametric multi-species model that explained 42.17% of the variation with RMSE of 12.03 feet. The parametric black oak model had the highest R2 (coefficient of determination) which was 0.75 with lowest RMSE of 9.58 feet. It is important to accurately characterize site productivity which allows for efficient land use allocation, integrated ecosystem planning, evaluation for ecosystem productivity and diagnosis and prescribed management for a forest ecosystem.
Subject Area
Environmental science|Natural Resource Management
Recommended Citation
Matthew Purucker,
"Modeling Forest Productivity Using Soil and Other Geographical Variables"
(2021).
ETD Collection for Tennessee State University.
Paper AAI28776280.
https://digitalscholarship.tnstate.edu/dissertations/AAI28776280