Spectral Identification of Previsual Northern Red Oak (Quercus rubra L.) Foliar Symptoms Related to Oak Decline and Red Oak Borer (Coleoptera: Cerambycidae) Attack

Abstract

A severe oak decline event associated with the outbreak of an insect pest (Enaphalodes rufulus [Haldeman]) affected the Arkansas, Missouri, and Oklahoma Ozarks from the mid-1990s through 2005. Tree mortality was severe and widespread, but patchy in nature. The rugged and remote terrain of the Ozarks led to difficulty in monitoring and assessing the location, severity, and spread of declining stands. In addition, visible symptoms of decline typically occur late in the decline-disease spiral, providing little early warning and leaving few options for mitigation. In situ spectroradiometry using handheld equipment offers an affordable option for characterizing the spectral signatures of plant stress and assessing the feasibility of areawide remote sensing projects without risking a large upfront investment in imagery. Foliage from 30 northern red oak (Quercus rubra L.) trees was sampled using a handheld spectroradiometer, and analysis was conducted using machine learning techniques. Analysis of 602 reflectance and first-derivative variables provided the best prediction (90%) of field-assigned stress rankings. Eight published vegetation stress indices used as surrogate variables yielded approximately 86% accuracy in predicting the correct field-assigned stress ranking. Therefore, use of vegetation stress indices as surrogate variables may be a simpler option for analyzing future imagery of oak stress in the Ozarks while sacrificing little accuracy.

Enaphalodes rufulus, foliar reflectance, insect outbreak, oak decline