Two sets of plantings were used in the research. Trees in the first set were originally grown as part of an OSU/OARDC study at the Ohio Division of Forestry nursery at Marietta to investigate nursery practices (seedling density, shading, and age) needed to produce acceptable seedling stock for field planting. Cones had been collected by OSU/OARDC personnel in the fall of 1984 from a natural stand of balsam fir in Canaan Valley, West Virginia, and seed was planted in the nursery in the spring of 1985. Following termination of the study in the spring of 1989, seedlings were lifted from nursery beds and 225 trees were distributed to each of 22 Christmas tree growers in Ohio who had plantations representing a wide range of soil/site conditions. In addition, plantings were made at two OSU/OARDC sites.
Preliminary analyses of data collected in those plantings indicated that trees on 35 of 63 plots had received annual applications of "residual" herbicides (which act, at least in part, through the roots of vegetation) for at least the first four years after trees were planted in combination with periodic mowings. On the other 28 plots, less effective types of weed control were used, including less frequent or no applications of residual herbicides, fall applications of foliar herbicides only, mowing only, etc., and on those plots it was difficult to assess the effects of other soil/site factors on survival, growth, and foliage characteristics of trees.
To provide additional data, plantings receiving "effective weed control" were needed which were of comparable seed source, size of planting stock, and years in the field after planting to that used in the original plantings. For that purpose, trees grown from seed collected in the Canaan Fir Tree Company's planted seed orchard in Canaan Valley and sold to 12 growers in Ohio (some of whom had plantings used in the first set of evaluations) as plugs+1 planting stock in the spring of 1990 were used. The seed orchard had been established in the early 1960s using trees grown from seed collected from natural stands in Canaan Valley. A total of 45 additional plots were established in those plantings, giving a total of 108 plots, 80 of which received "effective" chemical weed control (Figure 1).
Growers receiving trees for each set planted them in one or more blocks in their plantations and maintained them using various cultural practices, including weed control, fertilization, shearing, insect and disease control, and more.
Beginning in the winters of 1996-97 (for the first set of plantings) and 1997-98 (for the second set), a series of evaluations was made in the plantings. Since trees received by cooperators often had been planted over a relatively large area, two or more 10-tree plots having relatively uniform soil and other site factors within the plot were located in each plantation. Measurements and evaluations made on each plot included: (1) topographic factors; (2) a physical description of the soil and collection of soil samples from the Ap and B2 horizons; (3) an evaluation of cultural practices (weed competition, weed control, and fertilization); (4) survival of trees; (5) measurements of total tree heights, four-year heights (height before shearing was started), and annual growth for the last three years; (6) needle lengths and retention; (7) needle color; and (8) collection of a composite foliage sample from trees (uniform length pieces from the upper one-third of the crown). In the laboratory, soil samples and foliage samples were prepared and analyzed, including: (1) soil textures using hydrometer analyses (Bouyoucos 1962); (2) chemical properties of soils (by the OARDC Research-Extension Analytical Laboratory - REAL); and (3) weight per piece of foliage samples (following oven drying). In September 1997 (for the first set) and 1998 (for the second set), foliage samples were collected from each plot, oven dried, and needles were ground and analyzed by REAL for nutrient content of foliage.
Means, standard deviations, and range in values were determined for all variables for plots with and without "effective" chemical weed control. Statistical correlation analyses were run to test relationships between tree characteristics (dependent variables) and all of the site factors (independent variables) measured. Based on those correlations, multiple regression analyses were made in an attempt to develop mathematical models that would relate various combinations of independent variables (topography, soils, foliar nutrients, and management) to various tree characteristics (survival, growth, and foliage characteristics).