It is well established in plants (as well as mammalian systems) that there is facultative (makred by H3k27me3) and constitutive heterochromatin (marked by H3k9me3). In the paper attached you will notice (in maize), that DNA methylation in plants goes hand in hand with K9 histone methylation. So it is not absolute that all hetereochromatin will be marked by DNA methylation. Secondly, one must also be careful of the efficiency of your meDIP Antibody and pulldown washes/elution. One must have a few sensitive positive controls spanning genome wide locations to test for efficiency of the pull down before tiling it. If your positive controls are working and you still don't see methylation at specific locations, you can query a few of these candidate genes by k9 versus k27 marks (with a positive control for each at other locations). I would imagine that your meDIP is able to differentiate the two types of chromatin. 

I'm agree with dr. Sen, It's necessary the use of positive controls. I want only add the following consideration, the most MeDIP crucial experimental factors are the quality and cross-reactivity of the 5-methylcytosine antibody used in the procedure.

Good lack for your experiments!

Thank you Subhojit and Loredana. I agree that better definition for heterochromatin is the H3 marks. Still I'm bothered that I don't see higher methylation around the centromere, where repetitive elements are abundant. Could it be too low coverage of reads to the genome?