Use of profile analysis of repeated measures in a herbicide trial on blue morning glory (Ipomoea indica (Burm.) Merrill).

摘要

When several dependent variables, evaluated on the same scale, are measured against one or more independent variables using the multivariate approach, progressive differences in response to treatments can be determined. Each subject or a randomly selected group of subjects within each treatment is assessed repeatedly over time. Because the same scale of measurement is used, uniform profile response curves can be produced. This method of analysis has not been used commonly in weed science, but has value in providing a more complete picture of a weed's response to herbicide treatments. As an example, a field trial was undertaken approximately 130 km north of Brisbane, Queensland, to examine the effects of six herbicides (dichlorprop, triclopyr, fluroxypyr, glyphosate, metsulfuron-methyl and 2,4-D amine), some at several rates of application, when applied as overall sprays to blue morning glory (Ipomoea indica). Results from six assessments, spanning seven weeks, were subjected to profile analysis.While all herbicides damaged the weed, substantial regrowth occurred within the assessment period for all treatments except triclopyr and 2,4-D amine. The most cost effective treatment was 2,4-D amine at 400 g 100 litre-1 with the addition of Synetrol oil at 0.2%. Another introduced climber, glycine (Neonotonia wightii), was not affected by 2,4-D amine at the applied rates and could replace blue morning glory as a smotherer of native species if 2,4-D amine were the only herbicide employed. This highlights the need for weed management strategies to take into account the full complement of weeds at managed sites.