As noted earlier, the initial plantings made in 1989 contained (1) plots which received "residual" herbicide applications for at least the first four years after planting (in combination with periodic mowing) and (2) other combinations of weed control, including occasional use of herbicides and/or mowing. Analyses of data from those plots indicated that in the absence of "effective" chemical weed control it was difficult to assess relationships between performance and appearance of trees and other management, topographic, soil, and foliar nutrient variables (Figure 2).
Because of the type of sampling system used in the study, direct statistical comparisons could not be made between the 80 plots that received "effective" chemical weed control and the 28 plots that did not receive such control. As shown in Table 1, means, standard deviations, and ranges in values for topographic, soil physical, and soil chemical factors sampled on plots with and without "effective" weed control were generally similar. However, values for tree characteristics, plantation management variables, and foliar nutrient levels were very different.
For plots receiving effective chemical weed control, the average rating for vegetation density was 2.0 (light) and 1.6 (excellent/good) for weed-control effectiveness, while for plots without such control, averages were 3.7 (near the dense rating) and 3.6 (fair/poor), respectively. Effects of vegetative competition on growth and foliage characteristics were pronounced. On plots having effective weed control, survival of trees averaged approximately 14 percent higher than on plots that did not receive such control, while total tree heights and four-year heights averaged more than 45 percent higher; in addition all foliage characteristics of trees were eight to 11 percent higher for trees with effective weed control. Similar effects of weed control on survival, growth, and/or foliage characteristics of trees in plantations have been noted in a number of other studies (Brown, 1980 and 2000; Brown et al., 1989; Heiligmann et al., 1985). Also, levels of all elements in the foliage except copper and sodium were higher, with N, P, K, Ca, and Mg being nine to 17 percent higher and micronutrient levels six to 34 percent higher.
Because of the "masking" effects of vegetative competition on plots on other site factors, only those plots that had effective weed control were used in the following discussions of the overall performance of trees and in correlation and regression analyses to look at specific relationships between tree and site factors.