Laserfiche WebLink
C~ <br />variability determined the number of plots necessary to meet the <br />requirements for obtaining a mean that would detect a 10 percent <br />reduction or increase at a 20 percent probability for cover and <br />biomass of grasses and forbs. Similarly, sufficient numbers of <br />samples were taken to determine the density of the major woody <br />species at a 20 percent probability of detection for a 10 percent <br />change in the mean value. <br />Biomass for herbaceous plants (grasses and forbs) was obtained <br />by harvesting lm2 plots and percent plant cover for each species <br />was determined by use of a one-fourth meter square plot that was divided <br />into 100 square cells 2.5 cm on each side. Basal ground cover was <br />likewise determined following harvesting to determine bare ground <br />• and litter. <br />A composite sample of harvested material from each half-day's <br />work was taken to the laboratory for determining oven-dry weight. <br />In all cases where necessary, data were subjected to analyses <br />of variance for identifying sampling reliability. This is presumed <br />to be biomass and cover for grasses and forbs and for density of <br />woody species. The analysis of variance tables are presented for <br />each vegetation type in Table 2 of the Appendix. <br />Clusters were randomly located by use of a coordinate system of <br />grid lines 0.25 inch apart on an acetate overlay on an aerial photo <br />at 1" to 400'. The grid lines or coordinates run both directions <br />and therefore dissect each other every fourth inch. Coordinates <br />or grid lines running in one direction, up and down, were termed <br />• the "y" coordinates and those inning crosswise were termed the <br />"x" coordinates. Clusters were located to represent equal areas. <br />II.J-3 <br />