Main Article Content

Variation in foliar water content and hyperspectral reflectance of Pinus patula trees infested by Sirex noctilio


O Mutanga
R Ismail

Abstract

The remote detection and quantification of symptoms associated with declining forest health is critical for the introduction of proper pest monitoring and control measures. Sirex noctilio, the Eurasian wood wasp, is one of the major pests responsible for declining forest health in pine forests located in KwaZulu-Natal, South Africa. Researchers have shown  that stress induced by S. noctilio causes a rapid decrease in foliar water content, with the foliage of the tree changing from a dark green to a reddish brown hue. This study examined if the variation in foliar water content due to S. noctilio infestation can be remotely detected. Foliar water content and in situ hyperspectral measurements were obtained from Pinus patula trees experiencing varying levels of stress induced by S. noctilio. Subsequently, foliar water content was correlated to selected spectral variables consisting of known water absorption features, spectral  indices and continuum-removed absorption features. Results showed that the variations in foliar water content across the varying levels of S. noctilio infestation were strongly linked to the variation in hyperspectral reflectance. Except for water absorption features located at 970 nm and 1200 nm, there was a strong correlation between the majority of the spectral variables and foliar water content. Of the spectral variables tested, the water index (WI) provided the strongest linear correlation (r = 0.84) with foliar water content. Ultimately, results obtained from this study provide the foundation for the detection and monitoring of S. noctilio infestations at a landscape level using airborne or spaceborne hyperspectral platforms.

Keywords: absorption features; continuum-removed absorption features; foliar water content; hyperspectral; Sirex noctilio; spectral indices

Southern Forests 2010, 72(1): 1–7

Journal Identifiers


eISSN: 2070-2639
print ISSN: 2070-2620