Analysis of Variance, the statistical method used to test for differences in means among several treatment groups.
In brief, you start with a statistical model, such as y = u + Block + Treat + error. ANOVA then partitions the variation in your Y variable among the explanatory terms on the right side of the equation. From these pieces of variation, an F-test can be constructed that tests if treatment means are equal, or in statistical language
Ho: u1 = u2 = u3 = u4
Ha: at least one mean not equal.
SAS Proc Mixed has an abbreviated version of the "ANOVA table", that summarizes the tests performed, as illustrated here.
The F-test, along with its two degrees of freedom, produces a significance probability, or P-value (Pr>F). If the P-value is below your chosen significance level (generally .05), you reject Ho and conclude that at least one of the treatment means differs.