Back to search

Michigan Technological University

USING RANDOM FOREST MODELING TO PREDICT EARTHWORM DISTRIBUTION IN THE OTTAWA NATIONAL FOREST

Abstract

dc:description.abstract

Our study applies the machine learning method, Random Forest (RF), to understand distribution patterns and predictive powers of environmental variables determining earthworm occurrence in northern hardwood forests of the Great Lakes region. In our study we found earthworm species: Dendrobaena octaedra (Savigny), Lumbricus rubellus (Hoffmeister), Lumbricus terrestris (L.), Aporrectodea rosea (Saigny), Aporrectodea calignosa (Saigny), and Aporrectodea tuberculata (Eisen). Presence/absence data of L. terrestris were used in predictive distribution modeling for the Ottawa National Forest in the Upper Peninsula of Michigan based on the following Geographic Information Systems (GIS) variables: forest cover type, soil texture, soil pH, and distance from roads. Random Forest results were successful in producing models with high predictive accuracies and stable environmental variables when predicting L. terrestris occurrence. Deciduous cover type contributed the most to the outcome of the RF models, followed by soil texture, distance from roads and soil pH. The effectiveness of this approach in modeling earthworm distribution could be the first step in leading a large-scale predictive modeling effort to determine earthworm distribution for all of the Great Lakes region and other northern hardwood forest ecosystems. Having this insight would advance forest management efforts and regional studies addressing earthworm ecological effects.

Degree

thesis:*
Name thesis:degree_name
Master of Science in Applied Ecology (MS)
Level thesis:degree_level
Campus Access Master's Thesis
Discipline thesis:degree_discipline
College of Forest Resources and Environmental Science

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Cleaver, Megan
Committee members dc:contributor.committeemember
  • Ann Maclean
  • Kate Heckman
Contributors dc:contributor
  • Blair Orr
  • Erik Lilleskov

Subjects

dc:subject × 9

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
Michigan Technological University
Base URL
digitalcommons.mtu.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Cleaver, Megan. USING RANDOM FOREST MODELING TO PREDICT EARTHWORM DISTRIBUTION IN THE OTTAWA NATIONAL FOREST. Campus Access Master's Thesis thesis, https://doi.org/10.37099/mtu.dc.etdr/716