An image textural analysis method based on a combination of discrete wavelet transform (DWT) and principle component analysis (PCA) has recently emerged as a promising tool for feature extraction in images in a variety of disciplines. Uncertainty remains on the influence that wavelet type has on the use of this joint DWT-PCA method and on whether the less constraining independent component analysis (ICA) might be more efficient than PCA. In this context, the key objective of this note is to illustrate the effect of wavelet type on the textural analysis of a remotely sensed (QuickBird panchromatic) image of a wetland along the Hudson River in New York State and on the identification of four plant communities (reed, cattail, purple loosestrife, and shrub). The results of calculations involving six different types of wavelets suggest that the DWT-PCA method, unlike other available image analysis methods, is very effective at discriminating shrub from the other three plant communities, with limited influence of wavelet type. The ability to separate among the three remaining community types depends strongly on the wavelet used. By combining results obtained with the Daublets d4 and d12 wavelets, full discrimination among all four plant community types is feasible. For this particular analysis, ICA did not seem to have an advantage over PCA.