Figure 6. Plant biomass/tube as affected by soil treatment and plant species. A differ-
ent lower-case letter within the four soil treatments or the two plant species implies a
statistically significant difference within soil treatments or plant species, respectively.
ments are significantly different from the Palmer-
With the exception of shoot weight, the perennial
ton treatment. For this hypothesis, none of the
ryegrass had higher biomass than tall fescue.
biomass categories were significant at the 5 %
Because of the consistent patterns among soil
level. If the statistical criterion is relaxed to the
treatments (Fig. 6), we also evaluated these treat-
20% level, then all null hypotheses are rejected for
ment means using "contrasts" (Steel and Torrie
all four plant properties (Fig. 6).
1980). For example, we tested the following
Overall there were distinct patterns for plant
hypotheses:
growth. The perennial ryegrass grew better than
Hypothesis 1
the tall fescue. Growth of plants in the contami-
nated Palmerton treatment was poorest, and
H0: (T1 + T3 + T4)/3 T2 = 0
(3)
growth of plants was improved in the DCR-
Hypothesis 2
treated Palmerton soil.
Zn concentrations in plant shoots, crowns, and
H0: (T3 + T4)/2 T2 = 0
(4)
roots were always highest in the Palmerton soil
and lowest in the clean soil; the DCR treatments
where T1 is the clean treatment mean, T2 is the
fell between the latter two treatments (Fig. 7).
Palmerton mean, T3 is the DCR mean, and T4 is the
There were no significant differences in Zn con-
compacted DCR mean.
centrations attributable to plant species. For Zn,
Hypothesis 1 is a test of whether the Palmerton
there was a trend in relative concentrations that
mean is significantly different from the other treat-
fell in the following order: roots > crowns >
ment means. For crown weight, we rejected the
shoots.
null hypothesis at the 5% level of a Type I error.
Element contents in plant compartments (up-
Relaxing the standard to the 10% level, then Hy-
take) are a product of biomass and element con-
pothesis 1 would be rejected for all tested plant
centration. For Zn, there were no significant dif-
properties (Fig. 6). The contaminated Palmerton
ferences in Zn uptake attributable either to soil
treatment was significantly worst than the average
treatment or plant species (Fig. 8). Treatments
of the other treatments.
that were high in Zn concentrations, such as the
Hypothesis 2 is a test of whether the DCR treat-
16
Previous Page
