2.5. Statistical analyses
Prior to analysis, root colonization data were square root arcsine
transformed (Zar, 1999).
Data were tested for equality of variance by Barlett’s test before proceeding.
Spore abundance and percent soilNand P were analyzed using one-way ANOVA with soil origin as the factor.
The MIP bioassay of field soils, and Experiment 1 results were analyzed using ANOVA with crop history nested as a random factor within field.
Experiment 2 was analyzed using two-way ANOVA utilizing a fixed effects model with interactions (inoculum
source×host genotype).
The mycorrhizal responsiveness of maize plants in Experiment 2 was calculated as +AM/−AM mean where
+AM is the total dry weight (roots + shoots) of a mycorrhizal plant and −AMmean is the mean total dry weight of non-mycorrhizal
control plants of the same treatment. All statistical analyses were
performed using JMP version 4.0 software (SAS, 2004).