Analysis of variance in mean radial growth during 1992 to 1994 for green ash trees sampled in nine Iowa and Wisconsin cities. Analysis of mean radial growth for the past 10 years (1985 to 1994) and the past 20 years (1975-1994) indicated similar trends (unpublished data). Similar conclusions were obtained when the analysis was conducted using tree diameter as a covariate (continuous) variable.
| Source | df | SS | MS | F | P>F |
|---|---|---|---|---|---|
| Phytoplasmas detecteda | 1 | 293.50 | 293.50 | 22.77 | 0.0001 |
| City | 8 | 90.58 | 11.32 | 0.88 | 0.5409 |
| Phytoplasmas detected X city | 8 | 20.88 | 2.61 | 0.20 | 0.9892 |
| Diameter classb | 3 | 43.21 | 14.40 | 1.12 | 0.3507 |
| Phytoplasmas detected X diameter class | 3 | 4.39 | 1.46 | 0.11 | 0.9521 |
| City x diameter class | 20 | 222.58 | 11.13 | 0.86 | 0.6299 |
| Phytoplasmas X city X diameter class (Error I)c | 52 | 670.30 | 12.89 | 2.78d | 0.0001 |
| Error II (sample variation) | 213 | 988.20 | 4.64 | ||
| Corrected total | 308 | 2333.64 |
↵a Results of DAPI test for the presence of phytoplasmas in root sections.
↵b Tree diameter (at 1.4 m above the ground) classes of 0 to 30 cm, 31 to 50 cm, 51 to 70 cm, and 71 to 100 cm.
↵c In this survey, the number of DAPI-positive and DAPI-negative trees in each diameter class and city was not controlled (n=1 to 22). An analysis of the unweighted means for factors specified in the effects was conducted using the interaction mean square as the error term and locations (cities) as replications.
↵d The variance for individual observations (Error II) was used to test the significance of the interaction effect.