Experiments were conducted to examine the use of spectral reflectance curves for discriminating between plant species across moisture levels. Weed species and soybean were grown at three moisture levels, and spectral reflectance data and leaf water potential were collected every other day after the imposition of moisture stress at 8 wk after planting. Moisture stress did not reduce the ability to discriminate between species. As moisture stress increased, it became easier to distinguish between species, regardless of analysis technique. Signature amplitudes of the top five bands, discrete wavelet transforms, and multiple indices were promising analysis techniques. Discriminant models created from data set of 1 yr and validated on additional data sets provided, on average, approximately 80% accurate classification among weeds and crop. This suggests that these models are relatively robust and could potentially be used across environmental conditions in field scenarios.
