Laserfiche WebLink
'~ replications (p ~ 0.05), statistical analyses o£ quadrau within treat- <br />• went plots were not warranted. <br />•) Values for cover, density, and biomass for each replication within <br />Hach slope-aspect-mulch-mixture treatment were entered into a standard <br />analysis of variance (AOV) program for statistical analyses. <br />The statistical analyses were complicated by the fact that all <br />IDulch-seeding methods ate not present across all slopes and the fart <br />~ that controls are not replicated. The analysis approach was to compare <br />~ selected treatments that would yield in£ozmatioa about the major ~~per--'- <br />i' mental questions being asked (effect of slope, aspect, seeding method, <br />and mulch). Four separate AOV's were run to compare different groups of <br />treatments. The analyses are outlined on Figures 2.4-1 and 2.4-2. <br />r <br />~ Analyses in Figure 2.4-1 are intended to show aspect, mixture, and <br />mulch relationships on the same slope. Drill-straw is included haze <br />,- <br />bewnse it occurs on both aspects of 4:1 slopes, but it should be recog- <br />nized that straw plots are seeded at a lower rate. <br />~ Analyses in Figure 2.4-2 were made because excelsior mulch and <br />i <br />control treatments were the only treatmeau that occurred on all slopes <br />and north and south aspects. West and east slopes were not analyzed in <br />~•i order to reduce the complexity and because of insufficient replication. <br />i Excelsior ?lulch <br />Level (topsoil) Level (no topsoil) <br />L <br />~~ mesie zeric mesic zeric <br />This analysis vas made to compare the effect of no topsoil on level <br />•~- plot versus topsoiled level and 4:1 plots. <br />Differences between treatment means were coamared by the least <br />signi£icaat difference method (LSD) (Saedecor and Cochran 1967). LSD <br />values for each analysis of variance were calculated by multiplying the <br />standard error of the differences between means by the tabular value of <br />_ Student's t with the degrees o£ freedom of the error mean square. <br />In order to determine trends from 1978 to 1979, 1979 data were <br />entered into the same analysis of variance format as the 1978 data. <br />_ 8 <br />