11
harvests within a multiple harvest treatment for each genotype were
averaged or summed (depending on the parameter) into a single
observation before statistical analyses were conducted. The means for
most of the response variables from the single harvests are shown in
tables contained in the Appendix.
Analysis of variance was performed by using the general linear
model procedure of the Statistical Analysis System (SAS Institute
Inc., 1982). When the F-test for treatment effects showed
significance at the 5% level, this analysis was followed by Duncans
multiple range test (DMRT) to compare genotypes. Orthogonal
polynomial procedures were used to determine the nature of the
response over harvest frequencies. Statistical analyses were computed
separately for the tall and short genotypes. The DMRT notations for
tall genotypes are the usual a, b, and c letters, while the notations
for the two short genotypes are letters x and y. Analysis of variance
was not performed on values estimating methane potential because of
limited observations.
Results and Discussion
Oven Dry Biomass Yields
Differences in dry biomass yield did not occur among the four
"tall" elephantgrasses (PI 300086, Merkeron, N-43, and N-51) in either
year (Tables 1 and 2). In 1986, yields for the tall-growing L79-1002
energycane were inferior to the tall elephantgrasses, but the
following season differences did not occur. For all tall genotypes in
both years, the biomass yields decreased as the frequency of harvest