Hi, I think it is unlikely that you will get much mitochondrial sequence from the sequence tags used to generate the SNPs because of the representational nature of the SNP sequences and their relatively low density.
There might be one or two sequence tags from the mtDNA (17,000 bp), where the two restriction enzymes hit their mark at a distance apart suitable for the Illumina sequencing.
You can check this in silico if you have the whole mtDNA sequence for your species or a closely related species.
DArT is trialing sequencing of partial mtDNA sequencing (CytB and ND4) on the same samples as the SNP data are generated and the results look promising. This will allow the generation of both mtDNA and SNP phylogenies from the same set of samples, if that is useful to you. But it comes at an additional cost.
Hope this helps.
A