Validation of Landscape-Scale Decision Support Models That Predict Vegetation and Wildlife Dynamics

Stephen R. Shifley, Chadwick D. Rittenhouse, Joshua J. Millspaugh

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

14 Scopus citations

Abstract

Landscape vegetation and wildlife suitability models can be used to gain insights into the theoretical underpinnings of system processes; to make general predictions of change over large areas and long time frames; to make specific predictions for a small area; to address a single species of concern; and to evaluate trade-offs among multiple species. Several models to predict vegetation change for trees and forest stands are well established, and are widely used to predict forest change, and develop management prescriptions. Methods for validating short-term predictions of tree or stand change are documented in numerous sources. These typically compare model predictions with changes observed from remeasured forest inventor y plots. The metrics of interest are usually tree size or growth, tree mortality, and stand change per hectare (density, volume, diameter distribution). Validation of models generally consists of measuring the departure of the observed from predicted vegetation change, and reporting bias and precision of estimates by species, per hectare, and over time. Evaluation of landscape-scale forest vegetation models can be subdivided into three parts: validation of forest structure change in the absence of disturbance; validation of the rate of exogenous disturbance; and evaluation of species dynamics as affected by disturbance type and by ecological land type.

Original languageEnglish
Title of host publicationModels for Planning Wildlife Conservation in Large Landscapes
PublisherElsevier Inc.
Pages415-448
Number of pages34
ISBN (Print)9780123736314
DOIs
StatePublished - 2009

Funding

We thank the U.S. Forest Service and the University of Missouri for their support of our research.

FundersFunder number
U.S. Forest Service-Retired
Southeast Missouri State University

    Fingerprint

    Dive into the research topics of 'Validation of Landscape-Scale Decision Support Models That Predict Vegetation and Wildlife Dynamics'. Together they form a unique fingerprint.

    Cite this