Seed Germination

Technical Note DOER-C6

May 1999

blend must support plant growth and reduce the bioavailability of contaminants. The productivity

of the manufactured soil can be demonstrated by evaluating seed germination and plant growth of

Lycopersicon esculentum (tomato), Catharanthus roseus (vinca), Tagetes patula (marigold), and

Lolium multiflorum Lam. (ryegrass) grown in replicated 10-cm pots under controlled greenhouse

conditions. Temperature in the greenhouse should be maintained at 32.2 5 C during the day and

21.1 5 oC minimum at night. Relative humidity should be maintained as close to 100 percent as

possible, but never less than 50 percent. Emerged seedlings can be counted after 14 and 21 days

to determine seed-germination percentages. Plant seedlings can then be allowed to grow and

develop an additional 4 weeks to evaluate plant growth and appearance. Tomato, vinca, marigold,

and ryegrass (four annual plant species) were selected because they are sensitive to salt, metals,

and nutrient imbalances and represent a wide spectrum of upland plants (Raven, Evert, and Eichorn

1986).

Seed germination is important because the new plant starts as an embryo within the developing

seed. Therefore, the seed occupies a critical position in the life history of the plant. The success

with which the new plant is established is largely determined by the physiological response of the

seed to its environment (e.g., dredged-material salt content). The movement of water from dredged

material to seeds and uptake are essential steps toward seed germination. Some dredged material

with its high bulk density decreases capillary water and vapor movement toward the seed, which

in turn could result in decreased imbibition or physically restrict the swelling of the seed, thus

possibly impeding seed germination (Hagon and Chan 1977). The blend fertility and contaminant

levels usually do not directly affect seed germination, but may adversely affect the plant seedlings

after germination and subsequent plant growth. Therefore, the plant physiological responses to

various blends can be evaluated by using additional end points such as visual observations of leaf

color, size, and shape as well as total aboveground biomass after weeks of growth and exposure

to the blends.

Inferences as to the productivity of a particular blend can be made by statistically comparing percent

seed germination and plant biomass from the different blends to the percent seed germination and

plant biomass observed from the fertile reference soil (Tables 2 and 3; Figure 1). Table 2 shows a

typical manufactured soil screening test experimental design.

Seed Germination. An example of typical results from a greenhouse bench-scale test and

interpretation of data is presented in Figure 1 and Table 3. An evaluation of the statistical analysis

showed that seed germination was influenced by treatment (P = 0.0001), species (P = 0.0001), and

time (P = 0.01). P is the probability used as the criterion for rejecting the null hypothesis (H0). For

an example, if "H0:good germination is influenced by treatment" is a true, it will be erroneously

concluded to be false 0.01 percent of the time when P = 0.0001. Data analysis also revealed

a treatment-species interaction (P = 0.0001). Seed germination in the fertile reference control

(Blend 5) was significantly higher (P < 0.05) than seed germination in Blends 1, 2, 3, and 4

(Table 3). For example, tomato showed a 77-percent seed germination in Blend 2 compared with

83 percent in Blend 5, while marigold showed a 77-percent germination in Blend 2 compared

with 93 percent in Blend 5. However, seed germination in Blend 2 with Toledo Harbor Cell 1

dredged material was significantly higher than Blends 1, 3, and 4 using Toledo Harbor dredged

material as an ingredient. Although Blend 2 showed the best percent germination (64 percent)

overall, ryegrass percent seed germination was higher in Blend 3 than Blend 2 (Table 3). There