bowtie2 mapping qualities of 0

I have used bowtie2 to map paired-end reads together with unpaired reads to a reference genome (default parameters). I noticed when looking at my alignment in IGV, that at some locations there are a lot of reads with a mapping quality of 0. I know that in BWA a mapping quality of 0 indicates that the read had multiple equally good mapping locations and that BWA chose one alignment at random. My question is, does this also apply to bowtie2 alignments? The bowtie2 documentation does state that that the tool chooses a random alignment in case of multiple equally good alignments, but it doesn’t say anything on the mapping quality it gives those reads.